You are on page 1of 204

Common Pitfalls in the Evaluation

and Management of Headache

Case-Based Learning
Common Pitfalls in the
Evaluation and Management
of Headache
Case-Based Learning
Elizabeth W. Loder
The John R. Graham Headache Center, Brigham and Women’s Faulkner Hospital; and
Department of Neurology, Harvard Medical School, Boston, MA, USA

Rebecca C. Burch
The John R. Graham Headache Center, Brigham and Women’s Faulkner Hospital; and
Department of Neurology, Harvard Medical School, Boston, MA, USA

Paul B. Rizzoli
The John R. Graham Headache Center, Brigham and Women’s Faulkner Hospital; and
Department of Neurology, Harvard Medical School, Boston, MA, USA
University Printing House, Cambridge CB2 8BS, United Kingdom

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in
the pursuit of education, learning and research at the highest
international levels of excellence.
Information on this title:

c Elizabeth W. Loder, Rebecca C. Burch, and Paul B. Rizzoli 2014

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2014
Printed in the United Kingdom by TJ International Ltd, Padstow
A catalog record for this publication is available from the British
Library of Congress Cataloging in Publication data
Loder, Elizabeth, author.
Common pitfalls in the evaluation and management of headache :
case-based learning / Elizabeth W. Loder, Rebecca C. Burch, Paul B.
p. ; cm.
Includes bibliographical references and index.
ISBN 978-1-107-63610-1 (pbk.)
I. Burch, Rebecca C., author. II. Rizzoli, Paul, author. III. Title.
[DNLM: 1. Headache – diagnosis – Case Reports. 2. Headache –
therapy – Case Reports. WL 342]
616.8 491 – dc23 2013049904
ISBN 978-1-107-63610-1 Paperback
Additional resources for this publication at
Cambridge University Press has no responsibility for the persistence
or accuracy of URLs for external or third-party internet websites
referred to in this publication, and does not guarantee that any
content on such websites is, or will remain, accurate or appropriate.

Every effort has been made in preparing this book to provide
accurate and up-to-date information which is in accord with accepted
standards and practice at the time of publication. Although case
histories are drawn from actual cases, every effort has been made to
disguise the identities of the individuals involved. Nevertheless, the
authors, editors, and publishers can make no warranties that the
information contained herein is totally free from error, not least
because clinical standards are constantly changing through research
and regulation. The authors, editors, and publishers therefore
disclaim all liability for direct or consequential damages resulting
from the use of material contained in this book. Readers are strongly
advised to pay careful attention to information provided by the
manufacturer of any drugs or equipment that they plan to use.
We dedicate this book to the memory of Dr. John Ruskin Graham, an
early pioneer in the field of headache medicine and founder of our
Headache Center at the Brigham and Women’s Faulkner Hospital in
Boston. His work and ideals live on at the Graham Headache Center.
We strive to uphold Dr. Graham’s legacy of kindness, understanding,
and expert care for those affected by headache disorders.
Preface page ix
Acknowledgements x

1 Confusing one benign headache with 8 Pitfalls in drug therapy to prevent

another 1 headaches 119
2 Mistaking primary headache for 9 Pitfalls in nonpharmacologic treatment
another condition 20 of headache 145
3 Missing dangerous causes of headache 38 10 Challenges and special situations in
headache management 162
4 Pitfalls in diagnostic testing: imaging
and lumbar puncture 54 11 Medicolegal pitfalls in headache
management 175
5 Pitfalls in diagnostic testing: blood,
urine, and other tests 73
6 When historical or examination findings
are missed or misinterpreted 84 Index 186
7 Errors in management of acute
headache 100


The Illustrated Oxford Dictionary defines a pitfall as This book differs from those in two ways. First, the
“an unsuspected snare, danger, or drawback.” This idea cases in this book focus solely on challenging rather
of dangers that are disguised or difficult to recog- than routine aspects of headache diagnosis and treat-
nize is reflected in the word’s second meaning, which ment. Second, this book was written while the Inter-
is a “trap or snare,” often a covered pit, into which national Classification of Headache Disorders (ICHD)
unwary animals (or doctors?) might fall. Most doc- was being updated for just the third time since it was
tors correctly realize that evaluation and treatment of issued in 1988. As a result, all of the information in
headaches can be challenging. They know that most this book is current with the just-released ICHD-3 beta
headaches are benign but fear missing the occasional version of the classification. The ICHD-3 beta is avail-
serious cause of headache. Some believe that headache able free of charge at the website of the International
treatment is often unsuccessful and unrewarding. Headache Society (; you
To the extent that this aversion to headache may wish to print it for easy reference while reading
exists, we think it is unfortunate. In our experience, this book. We have not reproduced the diagnostic cri-
headache diagnosis is mostly straightforward. Further- teria verbatim but have instead summarized the clini-
more, headache medicine is often fascinating and grat- cal characteristics of each disorder, as specified in the
ifying. There are many unusual, intriguing types of latest version of the classification.
headache as well as new and highly effective treat- Among us, the three authors of this book have a
ments for headache. We find that even “old-fashioned” combined experience of full-time Headache Medicine
treatments, correctly applied, can produce satisfying practice of almost half a century. PR and EL are sea-
improvements for the right patient. As with any medi- soned veterans and bring many years of clinical expe-
cal discipline, though, there are challenging aspects of rience to bear. In contrast, RB is in the early phase
headache medicine. Our purpose in writing this book of her career; she has helped us focus on pitfalls and
is to make readers aware of the common and less obvi- mistakes to which newcomers or non-headache spe-
ous mistakes or pitfalls that we have encountered dur- cialists might be prone. We hope these dual perspec-
ing our years in practice – as well as those we have tives make this a book that is useful to anyone who
observed in the practices of others. might see headache patients, from family practition-
There are many case-based neurology books avail- ers to pain specialists. After all, who among us has so
able to practitioners who are interested in head pain, much experience that s/he cannot learn from the mis-
including several with an exclusive focus on headache. takes and near misses of others?


We offer our sincere thanks to Dr. Martin Samuels, this book. He tolerated with good humor the delay
the legendary and beloved chairman of the Depart- caused by the tardy release of the third version of
ment of Neurology at the Brigham and Women’s and the International Classification of Headache Disor-
Brigham and Women’s Faulkner Hospitals in Boston. ders. We hope he feels the result was worth the wait!
His vision and backing led to the formation of the Divi- Our thanks also go to Kirsten Bot, assistant editor and
sion of Headache within the Brigham Department of publishing assistant, who was so helpful with the final
Neurology – one of the first such divisions to be cre- phases of the book’s production. We also thank our
ated within the Neurology Department of a leading US copy-editor, Anne Kenton, for her keen eye and atten-
academic medical center. The Graham Headache Cen- tion to detail.
ter, and by extension this book, would not exist without
him. Elizabeth Loder, Rebecca Burch, Paul Rizzoli
We also thank Nicholas Dunton, our editor at July 2013
Cambridge University Press, for inviting us to write

Confusing one benign headache

1 with another

Sometimes there is simply no mistaking which pri- Migraine or tension-type headache?

mary headache disorder is causing a patient’s prob-
lem. In these lucky instances, a patient spontaneously Case
provides a history so characteristic of a disorder that
A 34-year-old teacher was referred for consultation
there can be no doubt about the diagnosis. In specialty
regarding recurrent headaches for the last five years.
headache practice, however, the patient who gives the
She was in good health and taking only occasional
doctor such a “silver platter” description of a headache
ibuprofen to treat the headaches. She estimated that
problem is the exception and not the rule. Since the
headaches occurred on average twice a month, lasting
primary, nondangerous headache disorders are clini-
a day or two at a time, and had not recently changed
cal diagnoses with, by definition, normal imaging and
in character. They were bilateral, over her forehead,
laboratory findings, the patient history is critical to
and sometimes accompanied by neck pain. She said
making a diagnosis.
that her job was stressful and wondered whether that
Unfortunately, as seasoned doctors know all too
might be causing the headaches. She missed a day of
well, obtaining an accurate history can be difficult.
work every other month because of headaches, and
This is particularly true for subjective symptoms such
reported she was seeking treatment because her missed
as headache, where patients are struggling to describe
work time had recently become “an issue” with her
things like pain that cannot be directly observed or
employer. Physical and neurologic examinations were
measured by the doctor. Then too, patients also are
normal except for mild tenderness on palpation over
reporting symptoms that may come and go. Studies
the posterior neck and upper trapezius muscles. She
show that patient recall of headache frequency and
had previously been told that she probably had “ten-
features is low and deteriorates rapidly over time. A
sion headaches” and had been referred for physical
missed or delayed diagnosis of a primary headache dis-
therapy but did not find that treatment effective.
order is unlikely to expose patients to life-threatening
harm, but it does have consequences. For one thing,
it may mean the patient does not have the benefit of
How can migraine be reliably distinguished
new, highly effective treatments that work for some from tension-type headache in this patient?
headache problems and not others. Careful evaluation of a full headache history is, in our
In this chapter we deal with cases in which a patient experience, the most useful method of distinguishing
with one of the “big three” common, nondangerous between migraine and tension-type headache. In this
primary headaches – migraine, tension-type, or clus- case, the patient did not spontaneously report char-
ter headache – presented with ambiguous, overlapping acteristic features of migraine such as nausea, vomit-
features and was therefore mistakenly diagnosed. To ing, photo or phonophobia, or worsening with physi-
minimize the chance of confusing these headaches, cal activity. On the other hand, we did not ask!
it is worth becoming familiar with the many ways in Table 1.1 lists migraine features that are contained
which the diagnostic features of common, nondanger- in the diagnostic criteria for the disorder, along with
ous headaches can overlap or be missed. Far and away examples of how these features can be missed or misin-
the most common pitfall is confusion about whether a terpreted. The diagnostic criteria for migraine have not
patient has migraine or tension-type headache, so we’ll changed in the latest version (3-beta) of the Interna-
start there. tional Classification of Headache Disorders (ICHD).

Chapter 1: Confusing one benign headache with another

Table 1.1. Diagnostic features of migraine without aura: common pitfalls

“Silver platter” migraine
features “Not so obvious” migraine history
Duration of 4–72 hours Duration uncertain because patient treats early or falls asleep. Shorter headaches often seen in children
Unilateral (often over the Bilateral, posterior location of pain or prominent complaints of neck pain often lead to a diagnosis of
temple)* tension-type headache, but neck pain occurs in almost three-quarters of migraine attacks
Throbbing* Not all patients who otherwise have clear-cut migraine report throbbing pain; for many patients the
throbbing quality of the pain is obvious only in fully developed, longer duration headaches, so patients
who treat early or fall asleep may not experience this. Be alert for synonymous descriptions such as
“pounding” or “with my heartbeat”
Moderate to severe pain In migraine attacks that are treated or do not progress, pain may never reach severe intensity.
intensity* Differences in pain reporting behaviors and pain perception among patients may affect patient ratings
of pain intensity
Aggravated by or causing Sedentary patients may not have noticed this feature of their headaches
avoidance of physical activity*
Nausea and/or vomiting# Vomiting is prominent in children with migraine, but often lessens as patients get older or headache
frequency increases. Decreased appetite may be present instead
Photophobia and Sensitivity to light or sound may become apparent only in headaches that have a chance to develop
phonophobia# fully; these symptoms may not develop in milder “forme fruste” or treated attacks
* Only two of these four features are required for a diagnosis of migraine.
# Only one of these two features is required for a diagnosis of migraine. At least five attacks meeting criteria are required before migraine
without aura can be diagnosed.

Some diagnostic criteria for migraine are more use- headache. This is true even in patients like the one
ful than others when trying to decide between diag- in our case who present with features such as muscle
noses of migraine and tension-type headache. A sys- tension and neck pain and attribute their headaches
tematic review of diagnostic studies showed that the to stress. Muscle tension and neck pain are common
features most predictive of migraine as opposed to in both migraine and tension-type headache, as is
tension-type headache were nausea, photo and phono- aggravation of headaches by emotional stress or ten-
phobia. When present, a typical history consistent with sion. In fact, migraine patients who have these over-
migraine aura was, unsurprisingly, also highly predic- lap characteristics (particularly neck pain) are most
tive of a migraine diagnosis. likely to receive an incorrect diagnosis of tension-type
In this case, further questioning revealed that
What benign headache disorder might the patient did have some loss of appetite with her
headaches and she became mildly sensitive to light,
account for this patient’s headaches? features which support a diagnosis of migraine instead
Taking the limited history at face value, this patient’s of tension-type headache.
presentation is compatible with a diagnosis of tension-
type headache. It is tempting to think that the bilat-
eral, posterior location of the headache and associated Discussion
muscle tenderness clinch the matter. Migraine is, how- Research in primary care settings shows that most
ever, also in the differential. In fact, while tension-type patients who seek care for troublesome headaches
headache is the most common type of headache in the receive a diagnosis of tension-type headache. This is
general population, it is not the most common type particularly likely to occur when patients report fea-
of headache in patients whose headaches are trouble- tures that are assumed to be highly characteristic of
some enough to seek medical care. A wealth of good tension-type headache – as the patient in our case
quality evidence suggests that once dangerous causes did. For example, many physicians (and patients, too)
of headache have been ruled out, the likelihood is assume that muscle pain or tenderness in the neck or
that patients consulting general physicians for trouble- shoulders is synonymous with tension-type headache.
some headaches have migraine and not tension-type They may also assume the same thing in patients who

Chapter 1: Confusing one benign headache with another

Table 1.2. Headache features and diagnosis of migraine

Specificity of
diagnosis for
Sensitivity of migraine vs. Likelihood ratio for diagnosis of
diagnosis of tension-type migraine (95% confidence
migraine (% of headache (% of interval)
Clinical feature patients) patients) Positive Negative
Nausea 82 96 23.2 (17.7–30.4) 0.19 (0.18–0.20)
Photophobia 79 87 6.0 (5.2–6.8) 0.24 (0.23–0.26)
Phonophobia 67 87 5.2 (4.5–5.9) 0.38 (0.36–0.40)
Exacerbation by physical activity 81 78 3.7 (3.4–4.0) 0.24 (0.23–0.26)
Unilateral 66 78 3.1 (2.8–3.3) 0.43 (0.41–0.45)
Throbbing or pulsating 76 77 3.3 (3.1–3.6) 0.32 (0.30–0.33)
Duration 4–24 hours 57 67 1.7 (1.5–2.0) 0.64 (0.58–0.71)
Duration 24–72 hours 13 91 1.4 (1.0–2.0) 0.96 (0.92–1.0)
Duration less than 4 hours 26 51 0.52 (0.44–0.61) 1.5 (1.3–1.6)

report high levels of psychologic or emotional tension. their lives, rather than the other way around. It is
Faced with a clinical situation like the one described certainly the case, however, that emotional stress is a
above, many physicians might consider sending the commonly mentioned “trigger” of headache in both
patient for physical therapy or prescribing a muscle tension-type and migraine patients. In fact, as demon-
relaxant, both treatments for tension-type headache. strated in Table 1.2, there is considerable overlap of
The assumption that these symptoms are indica- commonly reported triggers between migraine and
tive of tension-type headache probably stems from tension-type headache. This underscores the surpris-
the fact that the term “tension-type headache” sug- ingly low diagnostic value of many triggers and other
gests that “tension” of some sort – perhaps psychologic historical features commonly thought to be pathog-
or muscle – may be the cause of headache. This is a nomonic of one or the other disorder. Having patients
diagnostic pitfall, however, since evidence to support keep a headache diary, such as the one illustrated
these views is lacking. Although patients with tension- in Table 1.3, should help in distinguishing between
type headache do, as a group, have more pericranial tension-type and migraine headaches when the diag-
tenderness than patients with migraine, muscle pain, nosis is uncertain.
especially neck pain, is nonetheless very common in In a large multinational study, over a thousand
migraine patients. patients consulting physicians with a complaint of
Electromyography (EMG) is not useful in distin- headache were asked to keep careful diaries of their
guishing between the two disorders. Over a third of headaches for up to six months. These records were
migraine patients report neck pain with at least some then reviewed by headache experts, and the final
of their attacks. The neck pain can come before, dur- diagnosis of headache type was compared with the
ing, or even after attacks; this variability in its time diagnosis the treating physician had made at the
course makes it unlikely that neck pain is the “cause” of patient’s first visit. When physicians made a diagno-
headache. Both migraine and tension-type headache sis of migraine, this diagnosis was correct in 98% of
patients have lower thresholds for experiencing pain patients. When physicians diagnosed non-migraine
with pressure on muscles than do people without headaches, such as tension-type headache, the diag-
headache; interestingly, the upper trapezius is the most nosis ultimately turned out to be wrong in 82% of
common site of tenderness. patients. The predominant reason for misdiagnosis
Similarly, elevated levels of psychologic and emo- was having missed migraine. The authors of this study
tional distress are common in patients who seek med- concluded that “These findings support the diagnostic
ical care for stubborn headaches, and may in part approach of considering episodic, disabling primary
reflect the impact of poorly controlled headaches on headaches with an otherwise normal physical exam

Chapter 1: Confusing one benign headache with another

Table 1.3. A headache diary that can help distinguish between migraine and tension-type headache
Date Date Date Date Date
Just before the Yes Yes Yes Yes Yes
headache began, OR OR OR OR OR
was there any No No No No No
disturbance of
Just before the Yes Yes Yes Yes Yes
headache began, OR OR OR OR OR
did you have any No No No No No
numbness, or
speech problems?
What did the pain Pounding, throbbing Pounding, throbbing Pounding, throbbing Pounding, throbbing Pounding, throbbing
feel like? OR OR OR OR OR
Steady, tightening, Steady, tightening, Steady, tightening, Steady, tightening, Steady, tightening,
squeezing squeezing squeezing squeezing squeezing
Where was the Right side of the Right side of the Right side of the Right side of the Right side of the
headache located? head head head head head
Left side of the head Left side of the head Left side of the head Left side of the head Left side of the head
Both sides of the Both sides of the Both sides of the Both sides of the Both sides of the
head head head head head
Did you experience Nausea Nausea Nausea Nausea Nausea
nausea or vomiting OR OR OR OR OR
with the Vomiting Vomiting Vomiting Vomiting Vomiting
headache? OR OR OR OR OR
Both Both Both Both Both
Did the headache get Yes Yes Yes Yes Yes
worse with OR OR OR OR OR
physical activity or No No No No No
keep you from
being active?
During the headache, Yes Yes Yes Yes Yes
were you bothered OR OR OR OR OR
by light? No No No No No
During the headache, Yes Yes Yes Yes Yes
were you bothered OR OR OR OR OR
by sound? No No No No No
At its worst, was your Mild Mild Mild Mild Mild
headache pain OR OR OR OR OR
mild, moderate, or Moderate Moderate Moderate Moderate Moderate
severe? OR OR OR OR OR
Severe Severe Severe Severe Severe
How long did your Less than 4 hours Less than 4 hours Less than 4 hours Less than 4 hours Less than 4 hours
headache last? OR OR OR OR OR
4–72 hours 4–72 hours 4–72 hours 4–72 hours 4–72 hours
Longer than 72 Longer than 72 Longer than 72 Longer than 72 Longer than 72
hours or constant hours or constant hours or constant hours or constant hours or constant

to be migraine in the absence of contradictory nosis, as is a careful and probing history. Physicians
evidence.” should avoid placing too much emphasis on histori-
In summary, once a diagnosis of a primary cal features such as pain location, muscle tension, psy-
headache disorder has been made, it is appropriate chologic stress, and headache triggers. In contrast, a
for physicians to think migraine. Prospectively kept history of nausea in conjunction with headaches is
headache diaries are invaluable in making the diag- highly predictive of migraine.

Chapter 1: Confusing one benign headache with another

Table 1.4. The Visual Aura Rating Scale (VARS)

Migraine without aura. Visual symptom characteristic Risk score
Duration 5–60 minutes 3

Tip Develops gradually over 5 or more minutes 2

Scotoma 2
Migraine often presents with features assumed to
be highly characteristic of tension-type headache. Zig-zag line (fortification spectrum) 2
Most patients consulting physicians for trouble- Unilateral (homonymous) 1
some headaches have migraine and not tension-type Maximum VARS score 10
headache. Migraine with aura diagnosis ࣙ5
Adapted from: Eriksen et al. The Visual Aura Rating Scale (VARS)
Migraine with or without aura? for migraine aura diagnosis. Cephalalgia. 2005;10:801–10, with

An 18-year-old woman sought a second opinion ally could see just fine, but shut her eyes tightly because
because she had been told she should not use estrogen- the pain was so bad. Even with further questioning she
containing contraceptives – “the pill” – due to her diag- did not report additional associated symptoms.
nosis of migraine with aura. Instead she was prescribed
a progesterone-only pill but that had caused weight
gain and irregular periods and she recently stopped it. Does this patient have migraine with aura?
She was not interested in an intrauterine device and Aura is a focal neurologic event, which means that it
did not think she would be able to use barrier meth- includes symptoms that can be attributed to dysfunc-
ods reliably. Most of her friends were taking “regu- tion in a particular part of the brain. Aura symptoms
lar birth control pills” continuously and thus did not can be visual, sensory, motor, or mixed. Visual aura
have any withdrawal bleeding, and she wanted to do is by far the most common form of aura. Most peo-
this too. The patient’s headaches occurred only four or ple who have any form of aura will, at least occasion-
five times a year. She wondered if she might be able ally, also have visual aura. Common features of visual
to use estrogen-containing oral contraceptives despite aura are the scotoma, an area of decreased visual acuity
her headaches, since she did not want to get pregnant. or visual loss (not seeing something that is there) or a
positive visual phenomenon (seeing something that is
What important piece of information is still not there) such as a zig-zag line. These areas of visual
loss or distortion are surrounded by areas of normal
missing in this case? vision. Figure 1.1 shows a typical scintillating scotoma
It is not obvious that this patient’s diagnosis of drawn by a patient who has migraine with aura, who
migraine with aura is correct. In fact, all we know perceived that this scotoma was shimmering and pul-
is that she has headaches once every few months. In sating. Often a scotoma will start as a small area in the
order to establish that she has migraine and/or aura, center of the visual field, and then expand and move to
we need to know the details of her headaches and any the periphery of the visual field before fading away.
accompanying features. In this case, the patient con- In migraine with aura, the aura typically precedes
firmed a diagnosis of migraine by describing unilat- the headache; symptoms begin and fade away gradu-
eral, pounding headaches with nausea that last half a ally and do not last longer than an hour. Symptoms
day when untreated and were severe enough to pre- also are unilateral (or, in the case of visual symptoms,
vent her from usual activities, including her usual homonymous – which means they occur in only half of
exercise routine. When asked if she has any warning the visual field). Once the aura begins to fade away, it
signs, or other things that occur in association with her is usually quickly followed by a headache. Sometimes
headaches, she mentioned that sometimes her vision aura can occur without a headache. Table 1.4 repro-
was blurry prior to a headache. She added that occa- duces the Visual Aura Rating Scale, which is a method
sionally “the pain is so bad I can’t see.” When ques- of diagnosing visual aura. This scale assigns points for
tioned closely, she clarified this by saying that she actu- the presence of certain aura symptoms; to diagnose

Chapter 1: Confusing one benign headache with another

Figure 1.1 A typical scintillating migraine scotoma. Note the patient’s attempt to convey the shimmering sensation of movement and
vibration in the crescent-shaped, zig-zag visual phenomenon (fortification spectrum) that is drawn.

aura reliably requires a score of at least 5 of the pos- ommend against the use of estrogen-containing con-
sible 10 points. traceptives in women who are over 35 and have any
type of migraine, or women who have migraine with
aura, regardless of age.
Are estrogen-containing contraceptives
contraindicated in women who have Discussion
migraine with aura? This patient’s headaches met criteria for migraine but
Migraine with aura is associated with an increased the visual events she described were not consistent
risk of ischemic stroke and so is the use of exogenous with a diagnosis of aura. Thus, she has migraine with-
estrogens. The added risk of stroke with each of these out aura and the use of estrogen-containing contra-
things individually is quite small, but it is higher in the ceptives is not contraindicated. General visual blur-
presence of additional risk factors, such as smoking or ring and visual sensitivity, while commonly reported
increasing age. Although it is difficult to place abso- by migraine patients, are not aura. The blurred vision
lute numbers on these risks, authors of these studies described by this patient is better thought of as part
note the increase in risk is likely to be multiplicative of her headache prodrome. Prodromal events occur
rather than additive. Because of these risks, guidelines before a headache but are not focal neurologic events.
from a number of authoritative groups, including the Changes in mood, appetite, or concentration are
American College of Obstetrics and Gynecology, rec- commonly reported migraine prodrome symptoms.

Chapter 1: Confusing one benign headache with another

Distinguishing between migraine aura and migraine nosis, since sumatriptan is effective for the treatment
prodromal or associated symptoms is important of individual attacks of both migraine and cluster
because evidence that migraine with aura increases the headache.
risk of stroke continues to mount. Additional questioning about the features and
duration of individual headaches, as well as the pattern
of attacks over time, can distinguish migraine from
Diagnosis cluster headache. In this case, when questioned, the
Migraine without aura. patient reported that as a child he would occasionally
vomit with his headaches, and that they could last up
to a day. Both of these features are more consistent with
Tip a diagnosis of migraine than with cluster headache.
Focal neurologic symptoms such as positive or neg-
ative visual phenomena or sensory disturbance are
required for a diagnosis of migraine aura. Since essen- Why was the diagnosis of migraine missed
tially all patients with any sort of aura also have visual in this patient?
aura, it is only necessary to establish a history of visual
The criteria for diagnosing migraine are the same for
aura in order to make a diagnosis.
men and women. Unfortunately, they do not fully
reflect differences between the sexes in the clinical pro-
Severe unilateral headaches in a man file and presentation of migraine. Evidence from the
American Migraine Prevalence and Prevention Study
Case shows that, on average, men with migraine have fewer
clinical features of migraine than women with the
A 30-year-old man presented for management of clus-
disorder. That is, men with migraine are less likely
ter headache diagnosed by another physician. He
than women with migraine to report nausea, vomit-
reported an average of three headaches a month that
ing, photo or phonophobia. That does not mean they
tended to occur towards the end of the month: “They
do not have these symptoms, but rather that they have
come in bunches. As soon as one is over I might
fewer of these symptoms than women. Because doctors
get another one a day or so later.” Headaches were
rely on these characteristic historical features to make
located over his forehead bilaterally and he described
the clinical diagnosis of migraine, men are at a disad-
them as throbbing, with occasional mild nausea. He
vantage in receiving a correct diagnosis.
was not sure how long headaches would last with-
In this case, it seems likely that the patient’s physi-
out treatment because for many years he had been
cian correctly recognized that cluster headache is more
using subcutaneous sumatriptan as soon as headaches
common in men and migraine is more common in
women. She failed, though, to realize that neither
headache type occurs exclusively in one sex.
Is this history consistent with a diagnosis of
cluster headache? Discussion
Not all elements of the history in this case fit with clus- In addition to the fact that men with migraine report
ter headache. The bilateral nature of the pain is not typ- fewer migraine-associated features than women, there
ical of cluster headache, which is a strictly unilateral are other ways in which migraine may differ in
headache, usually located behind an eye. The throb- men compared with women. For example, men with
bing nature of the pain is more typical of migraine migraine report that, on average, their attacks are
than of cluster headache, in which the pain is usu- shorter, less severe, and less disabling than migraine
ally described as sharp. The patient does report that attacks in women. Males with migraine also are not
headaches come “in bunches,” but the period of sev- exposed to the potent migraine trigger of monthly
eral days between attacks is not typical of true clus- changes in sex hormones as are females with the dis-
ter headaches, in which short headaches usually occur order, so there is no increase in migraine prevalence
daily or even several times a day. The patient’s response at puberty in males. This is in contrast to the situa-
to sumatriptan is not helpful in clarifying the diag- tion in females, where migraine prevalence increases

Chapter 1: Confusing one benign headache with another

substantially at puberty and remains higher than the ious. She was taking 160 mg of long-acting propranolol
prevalence in men even into old age. This suggests and 200 mg of topiramate daily. At the last visit she
that female hormones have an enduring effect on reported that the medications were ineffective and that
migraine susceptibility. In addition to this lifelong she was “tired and can’t think straight. I am about to
impact on disease risk, women with migraine also lose my job.”
experience periodic increases in migraine attack fre-
quency because of hormonal changes with the men- Has this patient developed chronic
strual cycle.
Although the prevalence of migraine in men is migraine?
lower than in women, migraine is an extremely com- Chronic migraine is unlikely in this patient. Although
mon disorder in both sexes. By age 80, the cumulative daily, her headaches are short, usually lasting three
incidence of migraine in men reaches almost 18% – hours or less. Another clue that this patient may
meaning that almost one in five men will experience not have migraine is that the headaches have not
migraine during his lifetime. As high as this number responded to aggressive treatment with the typi-
is, though, it is certainly lower than the 44% lifetime cally used migraine preventive treatments of propra-
cumulative incidence of migraine in women. Thus it is nolol and topiramate. Of course, not all patients with
understandable that some physicians – and patients – migraine improve with appropriate preventive treat-
view migraine as a “woman’s disease.” Unfortunately, ment, but failure to respond to migraine-specific ther-
the result is that when men seek care for their apy may also suggest an alternative diagnosis. Most
headaches, clinicians may be less likely to consider features of this patient’s headaches are consistent with
migraine as a diagnosis. a diagnosis of cluster headache and not migraine.
Additional questioning about the features and dura-
Diagnosis tion of individual headaches, as well as the pattern of
attacks over time, can distinguish the two disorders
Migraine without aura.
(see Table 1.5).
In this case, the patient gave additional history
Tip which further clarified the diagnosis. For ten years
Migraine is more common in women than in men, before the onset of her daily headache she had just one
but it is highly prevalent in both sexes. Diagnosis of or two bouts of daily or near-daily headache each year.
migraine in men may be challenging because men with Those periods of frequent headache lasted on aver-
migraine have fewer typical migraine symptoms than age two months and seemed to come in the fall and
do women. spring of each year. Individual headaches have always
been located behind the left eye and associated with
Frequent, severe episodic headaches in left-sided nasal congestion and tearing of the left eye.
Although they last only two to three hours, headaches
a woman are extremely severe, “like someone is stabbing me in
the eye with a hot poker.” The patient is restless and
Case paces the floor during attacks. Her story is more con-
A 45-year-old woman had been treated for many years sistent with episodic cluster headache that has now
for a diagnosis of migraine. She treated individual become chronic than with chronic migraine.
headaches with 10 mg of oral rizatriptan. Headaches
typically awakened her from sleep and were extremely Why was cluster headache misdiagnosed in
severe, sometimes associated with nausea. They lasted
two or three hours, so by the time the rizatriptan this patient?
became effective the headache was almost over. For The presentation of cluster headache is highly charac-
the last year her headaches had occurred nightly or teristic but the disorder is uncommon. In contrast to
every other night, although prior to that she would migraine, it is more frequent in men than in women.
have headaches “only two or three months at a time It is caused by dysfunction of central nervous system
and I could deal with that.” With no break in her pain control mechanisms and has distinctive circadian
headaches she reported being sleep-deprived and anx- and circannual features. Most physicians have never

Chapter 1: Confusing one benign headache with another

Table 1.5. Distinguishing migraine from cluster headache

Pain features Migraine Cluster headache
Location Often unilateral over the temple or Strictly unilateral; typically highly localized to behind one eye
forehead area but may be bilateral
Duration of attack 4–72 hours (adults) 15 minutes to 3 hours*
Frequency of attacks Sporadic. Can “cluster” in bunches Attacks can occur once every other day up to eight times a day (for more
but rarely follow the distinctive than half of the time during an active cluster bout – headache frequency
pattern of true cluster headache may increase or taper slowly at the beginning or end of a bout)*
Associated features Nausea, vomiting, photo and Agitation or restlessness OR one of the following seven symptoms or signs
phonophobia must occur on the side of the headache: (1) eye redness or tearing; (2) nasal
congestion or runny nose; (3) edema of the eyelid; (4) sweating of the
forehead and face; (5) flushing of the forehead and face; (6) a feeling of ear
fullness; (7) decreased pupil size or ptosis*
Sex ratio Females ⬎ males Males ⬎ females
Behavior during attack Quiet; prefer to lie quietly in a dark Agitated, restless
Temporal features Attacks typically occur at random Attacks commonly occur at specific times of the day or night. Named for
and are not easily predictable the way they “cluster” together occurring daily or almost daily for 2- to
3-month bouts. In episodic cluster headache these bouts are separated by
periods of remission lasting at least a month; in chronic cluster headache
remissions do not occur or are shorter than a month
* According to ICHD-3 beta, all of these criteria must be met in at least five attacks in order to make a diagnosis of cluster headache.

seen or treated a patient with cluster headache and attack. Aura is very rarely seen in cluster headache but
may be especially likely to miss it in women, perhaps should not rule out the diagnosis.
because women with cluster headache are more likely Preventive treatment of cluster headache differs
than men to have migrainous symptoms such as nau- from that of migraine. Typical migraine preventive
sea. Diagnostic delay is common in any case, however, drugs such as topiramate and propranolol are unlikely
with one study showing that the median time from to be helpful for cluster headache. In this case the
onset to symptoms was three years, with a range of one delay in accurate diagnosis has delayed institution of
week to 48 years. appropriate preventive treatment aimed at reducing or
This patient’s partial response to a triptan medi- eliminating attacks of cluster headache. The mainstays
cation may also have contributed to confusion about of prevention for cluster headache are verapamil or
the diagnosis, since many doctors think of triptans as lithium. There are no US Food and Drug Administra-
“migraine medications.” In fact, triptans are useful for tion (FDA)-approved preventive treatments for clus-
treating individual attacks of cluster headache as well ter headache, but clinical experience shows that one
as migraine. or the other of these drugs, or occasionally the com-
bination, brings the disorder under control for most
Discussion The slow onset of action of oral triptans makes
Cluster headache is correctly diagnosed after the ini- them a poor treatment choice for most patients with
tial evaluation only 21% of the time. The most com- cluster headache. Subcutaneous sumatriptan, which
mon incorrect diagnoses made in these patients are patients can self-administer via an auto-injector, has
migraine (34%) and sinusitis (21%). Migraine and a more rapid onset of action. It is the only triptan for-
cluster can be differentiated on the basis of headache mulation that is FDA approved for treatment of cluster
duration, frequency, seasonality, triggering factors, headache.
and pain behavior during a headache. The presence
of autonomic features is usually part of the presenta-
tion of cluster headache but is not strictly necessary for Diagnosis
diagnosis if the patient is agitated or paces during an Chronic cluster headache.

Chapter 1: Confusing one benign headache with another

Tip ally identify non-aura disorders that might be causing

Cluster headache is more common in men but can also
In this case, the symptoms were associated with
occur in women. It is often missed in both sexes.
migraine and occur in consistent temporal relation
to the headaches. The patient was otherwise healthy,
Headache with disturbing visual and Alice in Wonderland syndrome is the most likely
perceptual alterations
Case How should this patient be treated?
A 43-year-old woman reported she had experienced As is the case for more typical forms of aura, there
bad headaches since childhood. In response to the are no clinically available treatments that will specif-
open-ended question “Does anything else happen with ically treat the aura of Alice in Wonderland syndrome.
your headaches?” she tearfully related symptoms that (Intravenous ketamine reportedly aborts aura in about
“might sound crazy.” At age 11 she awoke one morn- half of sufferers, but is not a practical approach to out-
ing and, while still lying in bed, realized that her hands patient therapy.) Rather, treatment is aimed at reduc-
did not feel “like they belonged to me.” When she held ing the number of migraine episodes using typical
them in front of her they looked long and twig-like, migraine preventive therapy, and also focuses treating
not like normal hands. These perceptions disappeared the pain of any headache that accompanies the aura.
in a few minutes, but she then developed a severe, Triptans and other vasoconstrictive agents are not con-
unilateral headache with vomiting. Over the years she traindicated in this disorder or in migraine with more
has had many similar episodes of visual abnormality, typical forms of aura.
all followed by severe headaches that meet criteria for
migraine. Once while driving she noticed that a fence
and the trees behind it appeared weirdly distorted in Discussion
size and shape. This abnormality was limited to the This unusual form of aura is called “Alice in Wonder-
left side of her visual field but was still present when land syndrome” because of its similarity to the expe-
she covered her left eye. The patient has discussed her riences of Lewis Carroll’s fictional Alice in Wonder-
headaches with other physicians, but not her visual land. It was first described in 1955. The visual abnor-
symptoms, because of her fear that they might be inter- malities in Alice in Wonderland syndrome are more
preted as psychiatric in nature. peculiar than those of typical visual aura. The visual
disturbance may also be associated with alterations in
What conditions may be causing these the perception of time, or feelings of depersonalization
or derealization, which seem to be what the patient
symptoms? in this vignette experienced during her first childhood
Bizarre visual illusions and distortions that affect the episode. Figures 1.2 and 1.3 depict an illustration of the
apparent size, volume, shape, or position in space of metamorphopsic visual distortions of this illness.
objects are described with the term “metamorphop- Alice in Wonderland syndrome is said to be more
sia.” Metamorphopsia is part of Alice in Wonderland common in children than adults. In our experience
syndrome, thought to be a particular sort of migraine this apparent difference in prevalence may stem from
with aura, but can also reflect structural eye disease the reluctance of adults to describe symptoms they fear
(usually retinal) and has been reported in patients will result in stigmatization. Children may be less wor-
with idiopathic intracranial hypertension. Metamor- ried about this. In their 2008 book Headache in Chil-
phopsia is a common symptom in age-related macular dren and Adolescents, Winner et al. report that “The
degeneration or other diseases that affect the macula. children rarely seem frightened by these illusions and
Typically, patients will complain of “waves” or “bend- relate the experience in enthusiastic detail. Witnesses
ing” in objects known to be straight, such as door- of the child’s event will either remark that the child
frames or roof lines. Careful examination of the retina, has an unusual, bemused look on the face or describe
as well as ancillary testing when necessary, includ- the child changing body positions so that they can ‘get
ing Amsler grids or fluorescein angiography, can usu- under a low ceiling.’”

Chapter 1: Confusing one benign headache with another

Figure 1.2 Normal vision.

Figure 1.3 Metamorphopsia in left

hemi-field of vision.

Diagnosis is reassuring for patients who may otherwise fear the

symptoms mean they are “crazy.”
Alice in Wonderland syndrome – commonly thought
to be a form of migraine with aura.
“Cluster migraine”
Treatment for Alice in Wonderland syndrome does not Case
differ from that of typical visual aura with migraine. A 27-year-old woman reported frequent headaches
It is important to recognize it because the diagnosis characterized by retro-orbital pain, always on the

Chapter 1: Confusing one benign headache with another

right, up to four hours in duration but usually shorter headache, and is also compatible with a diagnosis of
because she had learned to treat the headaches imme- migraine.
diately with sumatriptan. The pain could become
severe quite quickly, and when the pain was at its worst
it was a 10/10 and had a stabbing quality. With the most Is “cluster migraine” an appropriate
severe pain she also noticed tearing and redness of the diagnosis in this patient?
right eye. When the pain was milder she preferred to The ICHD recognizes Cluster headache, one of the
be lying down, but at its worst, “it doesn’t matter what TACs, and Migraine as two separate categories of
I do, it just hurts.” There was associated photophobia headaches. “Cluster migraine” is not a recognized term
and some nausea but no aura. She had noticed that she in any edition of the ICHD, including ICHD-3 beta,
had times when her headaches occur very frequently, and this term often causes a great deal of confusion
up to twice a day, and other times when she would among patients and providers.
have several weeks without a headache. In the past she Migraine sufferers sometimes hear the term “clus-
had been told she had “cluster migraines.” One of her ter” and relate this to their experience of having inter-
other physicians thought she had cluster headache and vals of frequent headaches interspersed with times
started verapamil, which had improved the frequency when headaches are less frequent. In our experience,
of headaches only mildly. these changes in headache frequency are commonly
the result of environmental factors such as changes in
lifestyle (a move, a new job), stress, changes in sleep
Are this patient’s autonomic features helpful habits, or dietary changes. Intrinsic factors such as
hormonal changes associated with menstrual periods
in making a diagnosis? may also contribute to cyclical changes in suscepti-
Autonomic features (lacrimation, conjunctival injec- bility to migraine. All of these things may lower the
tion, forehead or facial sweating, miosis, ptosis, eyelid threshold for the development of migraine, resulting
edema, rhinorrhea, or nasal congestion) are a defin- in an increase in headache frequency. Another phe-
ing characteristic of the headaches classified as trigem- nomenon often leading to “runs” of migraine activity
inal autonomic cephalagias (TACs), including cluster is rebound headache, in which withdrawal from symp-
headache. The latest version of the headache classifi- tomatic medications used to treat a previous headache
cation, ICHD-3 beta, however, does not require that leads to another headache.
they be present in order to make a diagnosis of clus- The patient described in this case meets criteria for
ter headache if the patient has the characteristic rest- migraine without aura and not for cluster headache,
lessness or physical agitation that are seen in cluster which was described earlier in this chapter. The term
headache. ICHD-3 beta has also added a sensation of “cluster migraine” is not a recognized diagnosis and
fullness in the ear to the list of autonomic features that should be avoided.
may be seen in cluster headache patients.
Headache-associated autonomic signs or symp-
toms are not unique to cluster headache or other Discussion
TACs, however. They may be seen in association with Autonomic activation during headache is a func-
migraine headaches as well. In one population-based tion of the trigeminal autonomic reflex. This reflex
study, one of four patients with migraine had unilat- occurs when nociceptive input from the trigeminal
eral autonomic symptoms with at least some attacks. system, as occurs during headache, stimulates the
One study also found that migraines occurring with trigeminocervical complex. This in turn stimulates
unilateral autonomic symptoms were more severe the superior salivatory nucleus, which gives rise to
than those occurring without. In a prospective study autonomic fibers providing innervation to the head.
comparing patients recruited from a headache cen- Autonomic symptoms are most commonly seen in
ter with migraine headaches with those with cluster the TACs, but can also occur in migraine. Compared
headache, 56% of patients with migraine had cranial to cranial autonomic symptoms in TACs, the symp-
autonomic symptoms with some of their headaches. toms in migraine are more likely to be bilateral, mild
Thus, the presence of autonomic features in this case or moderate in severity, and to occur inconsistently
does not necessarily indicate a diagnosis of cluster with headache attacks. Lacrimation is the autonomic

Chapter 1: Confusing one benign headache with another

symptom most commonly associated with migraines. Stabbing headaches in a migraineur

In contrast, the autonomic symptoms associated with
cluster headaches are unilateral 80% of the time and Case
almost always ipsilateral to the headache when they are
A 39-year-old woman presented to the office with a
“new headache type.” She had a long-standing his-
Distinguishing between “runs” of migraines and
tory of episodic migraine without aura. Headaches
a bout of true cluster headache can occasionally be
had increased in frequency over the last three years.
challenging. A very careful history should be taken
She attributed this to increased stress from having
with attention to headache duration, location of pain,
a second child four years ago and from increased
other associated features such as phono or photo-
responsibilities at work. Her migraines had gone from
phobia, and frequency. By ICHD criteria, migraine
occurring about once a week to as many as three per
and cluster should not overlap. Accurate diagnosis
can be difficult when treatment has eliminated fea-
In the last four months, she had also developed
tures needed for diagnosis as, for example, might occur
severe stabbing or “shock-like” pains behind the eye,
with treatment that shortens the duration of migraine
in the temple, or in the parietal area on either side.
headaches. In these cases, a description of typical
These lasted only a few seconds at a time but some-
untreated headaches can be very useful.
times occurred in volleys of up to 10 or 20 closely
In contrast to migraine, cluster headache bouts are
spaced jabs of pain in the course of an hour. She had
often distinctly seasonal, possibly because of headache
some days without this new type of pain but more typ-
triggering by changes in day length or seasonally
ically the pain happened between one and ten times a
related changes in circadian rhythms. This seasonal
day. She had seen her dentist last week for a routine
periodicity is uncommon in migraine. Within indi-
checkup, and had been told her symptoms might be
vidual patients, the seasonality of cluster headache
due to trigeminal neuralgia. There were no autonomic
bouts may be stereotyped over many years. Some clus-
features and no provoking factors that she was able to
ter headache patients are able to predict months in
identify. She continued to have her typical migraine
advance when a cluster bout will occur.
without aura headaches and these had not changed in
It can also be helpful to determine if the patient
character or frequency. Her neurologic examination
has headaches outside of the identified “cluster” peri-
was normal.
ods. A patient with episodic cluster headache rarely has
any headaches at all outside of their cluster periods or
bouts, while patients with migraine will usually have
occasional isolated headaches even during their “good What is the differential diagnosis of sharp,
periods.” While this is a general rule, it is not always
true, however, since an occasional patient with chronic stabbing pain in a migraineur?
cluster headache (defined as cluster headaches occur- Several disorders can produce brief, stabbing head
ring for over a year without remission or with remis- pain. The differential diagnosis is worth knowing
sion for less than one month) will experience occa- because complaints of sharp, stabbing head pain are
sional isolated headaches. The true cluster headache remarkably common. Possible causes include (1) a
pattern, with periods of headaches occurring from one trigeminal or other neuralgia; (2) primary stabbing
every other day to eight per day followed by complete headache; or (3) one of two uncommon forms of ultra-
remission periods, is almost never seen in migraine. short unilateral stabbing headache in the TAC family
known as SUNCT (Short-lasting, Unilateral, Neuralgi-
Diagnosis form headache attacks with Conjunctival injection and
Tearing) or SUNA (Short-lasting Unilateral Neuralgi-
Migraine without aura.
form attacks with Autonomic features – with the auto-
nomic features required for diagnosis being identical
Tip to those required for a diagnosis of cluster headache).
Migraine headaches can present with autonomic fea- Both SUNCT and SUNA produce pain that is simi-
tures in up to half of patients. “Cluster migraine” is a lar in location to that of cluster headache but much
confusing term which should be avoided. shorter, in the range of 1–600 seconds.

Chapter 1: Confusing one benign headache with another

Table 1.6. Clinical characteristics of primary stabbing uations requiring a workup for secondary causes. If
this patient had presented acutely with a new stab-
Pain occurs as a single stab or may be a series of stabs bing headache and associated neurologic symptoms,
Individual stabs last only a few seconds and may recur with further evaluation would be warranted.
irregular frequency
Stabs may be from one to many per day
Pain primarily occurs in the orbit, temple, or parietal areas, in the
distribution of the first division of the trigeminal nerve Primary or idiopathic stabbing headache has been
No accompanying symptoms, including no autonomic known by many names over the years, including “jabs
symptoms and jolts” and “ice-pick” headaches. The pathophysi-
ology is not well understood. It occurs in about 2%
of the general population, but is even more preva-
The pain characteristics in this case are similar to lent in patients with other primary headache disorders.
those of trigeminal neuralgia, but trigeminal neural- Up to 40% of migraineurs and 27% of patients with
gia typically does not switch sides or locations. Trig- tension-type headache may have comorbid primary
ger zones, areas of the face or mucosa where stim- stabbing headache, and it has also been described in
ulation reliably reproduces the pain, are frequently patients with hemicrania continua (one of the TACs).
but not always seen in trigeminal neuralgia. This In these cases, the stabs tend to occur in the loca-
patient does not report a history of pain triggering. tion most affected by the underlying primary headache
In the absence of autonomic features, a TAC such as type. The frequency of attacks can vary from a single
SUNCT or SUNA is highly unlikely. Thus, the most stab a few times a year to up to 50 per day. In some
likely etiology for the stabbing pain in this patient cases the stabs start out infrequently and progress to
is primary stabbing headache. The clinical charac- volleys of stabs lasting minutes to hours. Stabbing
teristics of primary stabbing headache are listed in headaches occurring outside the distribution of the
Table 1.6. trigeminal nerve have also been described with some
Primary stabbing headache is more prevalent than regularity.
either trigeminal neuralgia or SUNCT/SUNA and is Classically, primary stabbing headache is con-
very frequently seen in patients with migraine, where sidered an indomethacin-responsive headache and
it is often located in the same area where the migraine indomethacin is first-line treatment when treatment
pain is felt. is indicated. If the stabs are infrequent, reassurance
about their benign nature may be all that is needed.
About 30% of patients may not improve with indo-
Does any further workup need to be done methacin. Case reports have also described improve-
ment with gabapentin, celecoxib, and melatonin in
for this new headache type? those who did not respond to or could not tolerate
Secondary causes of stabbing headaches have been indomethacin. Although not a phenomenon described
described in case reports, and include herpetic in the literature, our clinical experience is that
meningoencephalitis, ischemic stroke, acute thalamic occasional stabbing headaches in migraineurs often
hemorrhage, and meningioma. Posterior fossa lesions improve with appropriate treatment of the migraine
may present with SUNA- or SUNCT-like symptoms. itself. This approach may spare the patient treatment
In this case, the four-month duration of symptoms with indomethacin, which like all anti-inflammatory
without worsening, the nonfocal neurologic exami- drugs can have serious gastric side effects.
nation, and the relative youth and good health of In this case, the patient’s underlying migraine
the patient are all reassuring. In our clinical expe- headaches were increasing in frequency due to envi-
rience, the development of stabbing headache in a ronmental factors. Because of the increased frequency,
migraineur often accompanies worsening of the pri- she was started on amitriptyline for prevention of
mary headache. However, stabbing headaches present- migraine. At her follow-up visit two months later,
ing in older patients, with progressive worsening of the migraines had returned to their previous weekly
symptoms, an abnormal neurologic examination, or pattern and the stabbing headaches had completely
failure to respond to appropriate treatment are all sit- resolved.

Chapter 1: Confusing one benign headache with another

Table 1.7. Clinical characteristics of new daily persistent

Diagnosis headache
Primary stabbing headache in the setting of migraine
Head pain for more than 3 months that is constant
without aura.
Unremitting from onset or within 24 hours of onset
No completely characteristic features, although migrainous
Tip features such as photophobia, phonophobia, or nausea are
Primary stabbing headache is commonly associated frequently described
with migraine and does not necessarily require a Female predominance with F:M ratio 1.4–2.5:1
workup for secondary headache if there are no accom- Onset of pain is clearly remembered by the patient, although
panying neurologic features. fewer than half of patients recall a triggering event

Chronic continuous headache with benefits had expired. She reported being frustrated by
acute onset her symptoms and inability to plan activity.

Case What is the differential diagnosis in

A 29-year-old right-handed female presented for eval-
uation of a “life-altering headache.” There was no fam- this case?
ily history of migraine and she recalled no headache The differential diagnosis for unremitting headache
problems during high school or college. On Septem- with acute onset is lengthy. Most of the potential
ber 8 (about four months prior to consultation), she causes are secondary forms of headache in which the
had awakened with a headache that never resolved. She head pain is due to some underlying process. For this
recalled nothing unusual around that time, specifically reason, an extensive workup for secondary causes of
no serious physical or emotional trauma or illness that headache (as was done in this case) is necessary. In this
had been associated with the onset of headache. She case, the patient had appropriate testing but no sec-
did recall having a mild upper respiratory infection ondary cause of headache had been identified.
several days before the headache began. She described A number of primary headache disorders can pro-
the headache as a pressure sensation behind the left duce the clinical scenario of acute onset, unrelenting
eye, usually 3/10 in severity but reaching 5/10 at its headache, and should be considered in this patient.
worst, always present from the time she woke up in These include hemicrania continua, chronic migraine,
the morning to the time she went to sleep. This pain chronic tension-type headache, and new daily persis-
had been unremitting since onset. She had mild nausea tent headache (usually referred to by its initials of
with the headache but otherwise no associated symp- NDPH).
toms or autonomic features and the headache did not One highly characteristic feature of NDPH is that
wake her from sleep. many patients (about half) recall the precise date when
Evaluation had included a normal magnetic reso- the headache started. This is so characteristic of the ill-
nance angiogram and venogram of the brain; normal ness that the diagnostic criteria for NDPH state that
CT angiogram; normal lumbar puncture with opening the onset must be “distinct and clearly remembered.”
pressure of 150 mm cerebrospinal fluid (CSF); normal In our experience many patients also note the exact
routine laboratory studies; negative Lyme titers; and time of onset with great precision, for example “My
normal antinuclear antibody (ANA) level, thyroid- headache began at 5:02 p.m. on September 8, 2003.”
stimulating hormone (TSH) level, erythrocyte sedi- Although the headache may pursue a “stuttering”
mentation rate (ESR), and C-reactive protein (CRP) course initially, it quickly becomes constant. In fact,
level. She had been diagnosed with migraine but failed diagnostic criteria require that the headache become
to improve with aggressive trials of typical migraine constant within 24 hours of onset (see Table 1.7 for
preventive drugs, except for a brief reduction in pain other clinical characteristics). When the onset of a
severity with a course of oral prednisone. She had ini- refractory headache is so clearly recalled, and in the
tially been unable to work and had moved back in with absence of abnormal tests or examinations, a diagno-
her parents. Recently, however, she had returned to sis of NDPH is far more likely than another primary
work after several months away because her disability headache disorder.

Chapter 1: Confusing one benign headache with another

Table 1.8. Secondary causes to rule out in case of apparent

What advice can be provided to the patient new daily persistent headache
about the natural history of NDPH, and is Vascular:
any further workup required? – Carotid or vertebral artery dissection
– Cerebral venous thrombosis
Many patients are relieved to be given a name for their – Giant cell arteritis
condition. By the time a patient receives this diagno- Nonvascular:
sis, they have usually had an extensive evaluation and – Disorder of CSF pressure, either high or low
some have developed a conviction that doctors have no – Meningitis
– Sphenoid sinusitis
idea what is wrong and have “given up.” Clarification – Intranasal (contact point) headache (pain caused by
that the condition is recognized, classified, and studied contact of intranasal structures such as nasal septum and
can itself be reassuring. Patients are sometimes disap- nasal turbinate)
– Cervical facet disease
pointed by the lack of good quality information about – Intracranial neoplasm or mass lesion
the natural history of the disorder, and by the lack of – Medication overuse headache
identified treatments for it. Providers often share their
The clinical features of NDPH are highly vari- In addition to more commonly recognized causes
able, and it can resemble migraine or tension-type of secondary headache that are described in Table 1.8,
headache. In the absence of validated treatments for studies currently suggest two other possible patho-
NDPH, treatment is usually aimed at the phenotype physiologic mechanisms for NDPH. The first is a local-
of the particular patient’s headaches. Thus, patients ized inflammatory process of the central nervous sys-
whose headaches have features of migraine, such as tem, perhaps precipitated by a viral illness of some
nausea, may be treated as this patient was, with drugs kind. The second theory is that the headache might
typically used for migraine. A few specific agents have result from a connective tissue process that leads to
been reported in the literature as benefitting small joint laxity and cervical spine hypermobility. Some
numbers of patients. These include the antiepilep- authors have suggested a typical physiognomy for
tic drugs topiramate and gabapentin, the tetracycline NDPH patients, classically a tall, thin female with a
derivative doxycycline, corticosteroids, nerve blocks, long neck, joint laxity, and hypermobility in the cer-
and mexiletine. vical spine.
Unfortunately, treatment of this disorder is often NDPH appears to be rare. In epidemiologic stud-
unsuccessful. It is still useful to follow the patient reg- ies, patients with NDPH are usually grouped together
ularly and to consider periodically whether other eval- with other people who have “chronic daily headache,”
uation is warranted. It may be useful, for example, and they account for about 10% or less of such patients
to consider evaluation for an underlying sleep disor- in most studies. Early descriptions of NDPH empha-
der. One other test to consider in patients who have sized that it was often self-limited and could remit
not had it is a lumbar puncture. This can identify dis- after months to years. Most headache specialists, how-
orders of CSF pressure, which in our experience can ever, see patients in whom the disorder persists and is
mimic almost any form of primary headache, includ- resistant to treatment. It is more common in females,
ing NDPH. A good therapeutic relationship with a with a sex ratio of about 2:1 and an earlier age of
patient can also prevent them from trying unproven onset in females. Although initial reports of NDPH
or dangerous therapies out of desperation. described headache with tension-type characteristics,
recent case series suggest that more than half the time
the headache may have associated migrainous features.
Discussion Because of this, the ICHD-3 beta criteria for the dis-
NDPH was first described in 1986 and is currently clas- order mention that the pain “lacks characteristic fea-
sified as one of the “other” primary headache disorders tures, and may be migraine-like or tension-type like, or
(other than migraine, tension-type, and the TACs, that have elements of both.” Similarly, no features are spec-
is). It is a diagnosis used with some trepidation by most ified by the diagnostic criteria for the location of the
clinicians, since it is essentially a clinical description of headache or the quality of the pain.
the patient’s presentation without clear evidence of any As with the patient in this case, the onset of NDPH
pathophysiologic explanation. in many patients has been associated with trigger

Chapter 1: Confusing one benign headache with another

Table 1.9. Suggested workup for apparent new daily Table 1.10. Clinical characteristics of hemicrania continua
persistent headache
r Unilateral, unremitting head pain lasting more than 3 months,
Neuroimaging: brain MRI with and without gadolinium, MRV with mild baseline pain but periodic exacerbations of severe
Intra- and extracranial MRA r During exacerbations
Lumbar puncture with opening pressure Either: agitation or restlessness OR
one of the following seven symptoms or signs must occur on
the side of the headache:
events, particularly a prior infectious, usually viral, ill- (1) eye redness or tearing;
(2) nasal congestion or runny nose;
ness, a stressful experience, or even an extracranial sur- (3) edema of the eyelid;
gical procedure. It is difficult to evaluate whether these (4) sweating of the forehead and face;
things are causally linked to headache or whether such (5) flushing of the forehead and face;
(6) a feeling of ear fullness;
events are recalled more easily by patients or clini- (7) decreased pupil size or ptosis
cians who are eagerly searching for a cause of disabling, r Response to indomethacin is complete at therapeutic doses
refractory headaches. In addition to any testing that is
indicated by a particular patient’s clinical presentation,
Table 1.9 lists investigations that should be strongly doses of ibuprofen gave some relief. She had been told
considered to rule out head trauma or disorders of CSF to discontinue this, however, because of worries that
pressure that can sometimes masquerade as NDPH. her daily use of the drug might be causing medica-
tion overuse headache. Her only other medication was
Diagnosis over-the-counter pseudoephedrine, which she took for
New daily persistent headache. “a problem with my sinus on the right.”

Tip What additional question might clarify the

Acute onset of unremitting headache is characteristic diagnosis in this case?
of NDPH, but many secondary headaches present sim-
It is understandable that a diagnosis of migraine has
ilarly. Although NDPH can have features of migraine,
been made in this case. Migraine is the most common
tension-type headache, or both, it is a distinct disorder.
form of disabling headache in women of childbearing
age, and it is often unilateral. Furthermore, it becomes
Strictly unilateral headache chronic in some patients, who often have mild to mod-
erate constant baseline pain with superimposed exac-
Case erbations of headache, just as this patient describes.
A 46-year-old woman presented with a 20-year his- When pressed, this patient also reported occasional
tory of disabling headache that had not responded well nausea, another feature that may suggest migraine. It is
to treatment. She had been diagnosed with chronic easy to overlook the curious fact that her symptoms all
migraine due to severe headaches located on the right appear to be right-sided. There is, however, an alterna-
side of her head only. When queried specifically about tive diagnosis that is compatible with her continuous,
associated symptoms of nausea, vomiting, photo and unilateral headache: hemicrania continua, a primary
phonophobia, the patient endorsed only occasional headache with autonomic features.
mild nausea. She also reported “shadow headaches,” The clinical criteria of hemicrania continua are
by which she meant that in between episodes of severe listed in Table 1.10. In the new version of the headache
headache she had continuous milder pain, also on the diagnostic classification, it is now considered one of
right side of the head. Her neurologic examination was the trigeminal autonomic cephalgias (formerly it was
normal, as was neuroimaging done in the last year. considered with the “other” primary headaches). Like
She brought past medical records that documented cluster headache and the other TACs, hemicrania con-
prior unsuccessful treatment trials with a large number tinua is a side-locked, strictly unilateral headache with
of preventive drugs for migraine, including botulinum associated autonomic features. It differs from the other
toxin injections. The doses and duration of these tri- TACs, however, in that it is continuous. It is also
als appeared to have been adequate. Triptans were uniquely responsive to a particular nonsteroidal anti-
never effective for headache exacerbations, but large inflammatory medication, indomethacin. Response to

Chapter 1: Confusing one benign headache with another

this drug is required to make a firm diagnosis of continuous headache should prompt consideration of
hemicrania continua. (Unfortunately, this diagnostic hemicrania continua, but this is especially true if the
requirement means that a definite diagnosis cannot be patient is partially responsive to nonsteroidal anti-
made in patients who present with a clinical picture inflammatory treatment.
compatible with hemicrania continua but cannot take The temporal pattern of hemicrania continua is
or do not tolerate indomethacin.) low-grade baseline headache with occasional exacer-
With the possibility of this alternative diagnosis bations. Patients may initially report only the more
in mind, the most useful additional question for the severe headaches and neglect to mention the lower-
patient described in the case becomes “Have you ever grade baseline headache because it is less bothersome
had pain on the left side of the head?” In this case, the to them. This is one reason hemicrania continua may
patient reports that she has never had left-sided pain. be misdiagnosed as migraine. An additional clue to the
Further questioning revealed that this nasal conges- correct diagnosis was the nasal congestion for which
tion was right-sided, and that her husband had told she took daily decongestants, which upon question-
her that her right eye was sometimes “droopy.” Upon ing was ipsilateral to the headache, as was the pre-
further discussion, the patient stated that these symp- viously undisclosed ptosis. Patients with hemicrania
toms had never come up because no one had ever asked continua may also report a sense of something being
before. stuck in the eye, such as an eyelash or a piece of sand,
sometimes called a “foreign body sensation.” This is
What treatment should be tried next? another symptom patients frequently do not report
The patient’s additional history makes a diagnosis
When headache treatment fails, a number of pos-
of hemicrania continua likely, but a definite diag-
sible explanations may be considered: (1) incorrect
nosis requires a complete response to therapeutic
diagnosis; (2) exacerbating factors have been missed;
doses of indomethacin. The reported effective dose
(3) treatment has been inadequate; or (4) unrealis-
of indomethacin in hemicrania continua ranges from
tic expectations or comorbidity exist. In this case,
50 to 300 mg/day. This dosing regimen is also
additional and open-ended questioning about the
commonly used in patients with primary stabbing
headache history provided important clues to correct
headache, another indomethacin-sensitive primary
diagnosis and treatment.
headache syndrome. Common practice is to begin
with indomethacin 25 mg orally three times a day
for three days; if the patient tolerates this but has no
headache improvement the dose is increased to 50 mg
three times a day for three more days, and again if Hemicrania continua.
no response to 75 mg three times a day for a further
three days. The authors of ICHD-3 beta recommend
that doses of up to 225 mg/day should be tried “if Tip
necessary.” Hemicrania continua is easily confused with chronic
There are no good treatment alternatives for migraine. Important characteristics, including accom-
patients who cannot use indomethacin. Patients panying autonomic features and the side-locked char-
may occasionally experience a partial, less complete acter of the pain, may not be volunteered by the
response to other anti-inflammatory agents, but most patient.
patients who cannot tolerate indomethacin also have
difficulty tolerating other anti-inflammatory drugs. Further reading
Migraine vs. tension-type headache
Discussion Calhoun AH, Ford S, Millen C, et al. The prevalence of neck
This patient reported a long history of headaches pain in migraine. Headache. 2010;50(8):1273–7.
diagnosed as migraine that have been refractory to Diener HC, Pfaffenrath V, Pageler L, et al. Headache
appropriate treatment for migraine. This is a common classification by history has only limited predictive value
story in patients who ultimately receive the diagno- for headache episodes treated in controlled trials with
sis of hemicrania continua. Any history of side-locked, OTC analgesics. Cephalalgia. 2009;29(2):188–93.

Chapter 1: Confusing one benign headache with another

Fernandez-de-las-Penas C, Madeleine P, Caminero AB, van Vliet JA, Eekers PJ, Haan J, Ferrari MD, Dutch RUSSH
et al. Generalized neck-shoulder hyperalgesia in chronic Study Group. Features involved in the diagnostic delay
tension-type headache and unilateral migraine assessed of cluster headache. J Neurol Neurosurg Psychiatry.
by pressure pain sensitivity topographical maps of the 2003;74(8):1123–5.
trapezius muscle. Cephalalgia. 2010;30(1):77–86.
Jensen R. Pathophysiological mechanisms of tension-type Alice in Wonderland syndrome
headache: a review of epidemiological and experimental Podoll K, Robinson D. Migraine Art – The Migraine
studies. Cephalalgia. 1999;19(6):602–21. Experience from Within. Berkeley, California, North
Smetana GW. The diagnostic value of historical features in Atlantic Books. 2009; 85–146.
primary headache syndromes: a comprehensive review. Todd J. The syndrome of Alice in Wonderland. Can Med
Arch Intern Med. 2000;160(18):2729–37. Assoc J. 1955;73(9):701–4.
Tepper SJ, Dahlof CG, Dowson A, et al. Prevalence and Winner P, Lewis D, Rothner AD, eds. Headache in Children
diagnosis of migraine in patients consulting their and Adolescents. Lewiston, NY, BC Dekker, 2008.
physician with a complaint of headache: data from the
Landmark Study. Headache. 2004;44(9):856–64. “Cluster migraine”
Wober C, Holzhammer J, Zeitlhofer J, Wessely P, Barbanti P, Fabbrini G, Pesare M, Vanacore N, Cerbo R.
Wober-Bingol C. Trigger factors of migraine and Unilateral cranial autonomic symptoms in migraine.
tension-type headache: experience and knowledge of the Cephalalgia. 2002;22(4):256–9.
patients. J Headache Pain. 2006;7(4):188–95.
Lai TH, Fuh JL, Wang SJ. Cranial autonomic symptoms in
migraine: characteristics and comparison with cluster
Diagnosing migraine with aura headache. J Neurol Neurosurg Psychiatry. 2009;80(10):
Eriksen MK, Thomsen LL, Olesen J. The Visual Aura Rating 1116–19.
Scale (VARS) for migraine aura diagnosis. Cephalalgia. Obermann M, Yoon MS, Dommes P, et al. Prevalence of
2005;10:801–10. trigeminal autonomic symptoms in migraine: a
population-based study. Cephalalgia. 2007;27(6):504–9.
Migraine in men
Stewart WF, Linet MS, Celentano DD, Van Natta M, Ziegler NDPH
D. Age and sex-specific incidence rates of migraine with Goadsby PJ. New daily persistent headache: a syndrome not
and without aura. Am J Epidemiol. 1991;134(10):1111– a discrete disorder. Headache. 2011;51(4):650–3.
Robbins MS, Evans RW. The heterogeneity of new daily
Stewart WF, Wood C, Reed ML, Roy J, Lipton RB, AMPP persistent headache. Headache. 2012;52(10):1579–89.
Advisory Group. Cumulative lifetime migraine
incidence in women and men. Cephalalgia. 2008;28(11): Rozen TD. New daily persistent headache: clinical
1170–8. perspective. Headache. 2011;51(4):641–9.
Young WB. Expert Commentary on New Daily Persistent
Cluster headache in women Headache. Headache. 2011;51(4):654–6.
Rozen TD, Fishman RS. Female cluster headache in the
United States of America: what are the gender Hemicrania continua
differences? Results from the United States Cluster Lipton RB, Silberstein SD, Saper JR, Bigal ME, Goadsby PJ.
Headache Survey. J Neurol Sci. 2012;317(1–2):17–28. Why headache treatment fails. Neurology. 2003;60(7):
Rozen TD, Fishman RS. Cluster headache in the United 1064–70.
States of America: demographics, clinical characteristics, Rossi P, Tassorelli C, Allena M, et al. Focus on therapy:
triggers, suicidality, and personal burden. Headache. hemicrania continua and new daily persistent headache.
2012;52(1):99–113. J Headache Pain. 2010;11(3):259–65.

Mistaking primary headache for another

2 condition

In Chapter 1, we discussed pitfalls that can lead to “Sinus headache”

confusing one type of primary headache disorder with
another. Primary, nondangerous headaches may also
be confused with or incorrectly attributed to multi- Case
ple other medical conditions. It is of course possible A 55-year-old man reported a history of headaches
for patients to have two separate forms of headaches since age 15. These were initially intermittent but
(e.g. allergy-related headache and migraine), but this had gradually worsened over the last several years
situation is probably not as common as many doc- and become constant. The episodic headaches he
tors assume. Some medical conditions, for example had when he was younger had been unilateral;
hypertension, are commonly but mistakenly assumed severe; lasted up to 24 hours; and were associated
to be frequent causes of headache. Other conditions, with photophobia, phonophobia, and nausea. He had
such as sinus disease or temporomandibular joint been diagnosed with “cluster and vascular tension
dysfunction, may exacerbate an underlying primary headache.”
headache disorder but are less commonly its sole He had been referred to an otolaryngologist
cause. And finally, for some disorders, such as Arnold because of frequent episodes of bilateral facial pain
Chiari malformations or pituitary lesions, it may and rhinorrhea. An MRI scan showed mild mucosal
be impossible to determine with certainty whether thickening (Figure 2.1) in the maxillary sinuses bilat-
the primary headache is related to the structural erally, and he was told that his headaches were prob-
lesion. ably due to sinus disease. After several rounds of
Most of the medical conditions incorrectly antibiotic treatment failed to produce improvement in
assumed to produce headache are common, and headaches, he had undergone a total of seven surgi-
under the right circumstances all of them can be the cal sinus procedures over the course of 20 years in an
true cause of headaches. Their prevalence, however, unsuccessful attempt to treat headaches.
means that often their association with headaches is At the time of consultation he reported bilat-
due to chance. This matters because assumptions of eral pain over the maxillary sinuses and forehead
causality may lead to overly aggressive treatment of area, as well as the posterior head and neck. He
the disorder assumed to be causing the headaches. rated the pain as 5 or 6 on a 0–10 scale at base-
Such treatment will not only be ineffective for the line, with daily exacerbations up to 8 or 9 lasting five
headaches but may also lead to other side effects or to six hours that were associated with photophobia,
problems. A good example is sinus surgery that is phonophobia, osmophobia, and nausea. The headache
done in the mistaken belief that migraine headaches increased when he bent over and he noted a sensa-
are due to sinus disease. We see many patients whose tion of pressure and fullness in the maxillary regions
migraines continue to be a problem but who also suffer bilaterally.
from persistent post-surgical pain or dysesthesias. The patient’s exacerbations of pain responded
In this chapter, we highlight the pitfalls that can well to symptomatic treatment with sumatriptan,
lead to an incorrect diagnosis of headache due to but trials of typical preventive medications for
another condition when the actual problem is a pri- migraine did not reduce the frequency or intensity of
mary headache such as migraine. headache.

Chapter 2: Conditions that mimic primary headache

Table 2.1. Clinical features of chronic vs. recurrent

Recurrent acute
Chronic rhinosinusitis rhinosinusitis
More than 3 months of a Four or more episodes
combination of purulent nasal per year of acute sinusitis
drainage, nasal obstruction, with clearing of
facial pressure/pain, or reduced symptoms and findings
sense of smell in between
Inflammation documented by
direct observation or imaging

sinusitis was correct, but it seems reasonable to assume

that the patient had at least recurrent acute bacterial
rhinosinusitis if not chronic rhinosinusitis in addition
to the primary headache syndrome.

Won’t imaging clarify the diagnosis of

In this situation, diagnostic confusion is not usually
resolved by diagnostic testing. It is important to note
Figure 2.1 An MRI showing mucosal thickening in both maxillary
that there is considerable expert consensus that abnor-
sinuses. Source: MRI malities on CT scan are not sufficient to make a diag-
112010 rgbca.png. nosis of rhinosinusitis. Some imaging abnormalities
commonly assumed to indicate rhinosinusitis are not
specific; for example, mucosal thickening may be seen
How is chronic sinus disease diagnosed? in up to 40% of the asymptomatic population. Most
The term rhinosinusitis is preferred to sinusitis because expert guidelines suggest that a patient must meet clin-
the middle turbinate is often involved in sinus dis- ical criteria for a diagnosis of sinusitis first, after which
ease. A clinical practice guideline from the American imaging is considered to be confirmatory rather than
Academy of Otolaryngology – Head and Neck Surgery suggestive. Thus, abnormalities on sinus imaging or
(AAOHNS) defines acute bacterial rhinosinusitis endoscopy that are consistent with sinusitis are not
(ABRS) as up to four weeks of purulent nasal drainage, by themselves sufficient to assume that they are caus-
accompanied by nasal obstruction, facial pain/ ing headache. CT is, however, particularly helpful in
pressure/fullness, or both. Recurrent acute rhino- determining any anatomic abnormalities which may
sinusitis is diagnosed when four or more episodes of be exacerbating recurrent rhinosinusitis or chronic
ABRS separated by symptom-free periods occur in rhinosinusitis.
one year. Chronic rhinosinusitis is defined as 12 weeks
of two or more of the following symptoms: muco- What sinus-related headaches do headache
purulent drainage, nasal obstruction/congestion,
facial pain/pressure/fullness, or decreased sense of specialists recognize?
smell, as well as documented inflammation as evi- The International Classification of Headache Disor-
denced by either purulent mucus or visible mucosal ders (ICHD)-3 beta recognizes two types of headache
edema, nasal polyps, or radiographic imaging show- related to sinus disease. The first is headache attributed
ing paranasal sinus inflammation. Table 2.1 lists to acute rhinosinusitis. The diagnostic criteria for
criteria for distinguishing chronic rhinosinusitis from this form of headache require clinical, endoscopic,
recurrent bouts of acute disease. or imaging evidence consistent with rhinosinusitis. In
In this case we are not provided with enough infor- addition, however, some evidence of causality must be
mation to be certain whether the patient’s diagnosis of present in order to attribute any associated headache

Chapter 2: Conditions that mimic primary headache

to the sinus abnormalities. Evidence that sinus abnor- Table 2.2. Clinical characteristics of headaches attributed to
rhinosinusitis and disorders of the nasal cavity
malities are causing headache can come in the form
of a temporal relationship with the development of Sinus Headache due to
or any change in headache, or by exacerbation of the (rhinosinusitis) disorders of the
headache with pressure over the paranasal sinuses. If headache nasal cavity
the physical sinus abnormalities are unilateral, an ipsi- Pain location Frontal headache If imaging or physical
lateral headache also stands as evidence of a causal radiating to face, ears, abnormalities are
or teeth unilateral, headache
relationship, according to these new criteria. Inter- should also be
estingly, the requirement that headaches be frontally unilateral. Otherwise,
located or accompanied by pain in the face, ears, or no specific location is
teeth has been removed from the new version of the
Imaging Direct or imaging Direct or imaging
classification. evidence of changes evidence of a
ICHD-3 beta also recognizes headache attributed consistent with hypertrophic or
to chronic or recurring rhinosinusitis, commenting chronic or acute inflammatory process
sinusitis in the nasal cavity
that while the association of headache with chronic
sinusitis has been controversial, new evidence appears Aggravating/ Headache develops Pain often resolves
alleviating or changes in with local application
to support a connection. Imaging or endoscopic evi- features conjunction with the of anesthetic to the
dence of current or past infection or inflammation in clinical course of contact point region
the sinuses is required, along with at least two other sinus inflammation
things that support a causal connection. These can
be either the development of or change in headache
character that parallels the degree of sinusitis, the
onset of headache in conjunction with the disorder, can likewise have serious complications. We see many
or exacerbation of the headache with pressure over patients in our practice who have developed chronic
the paranasal sinuses. As with headache attributed to pain problems or dysesthesias after having many sinus
acute sinusitis, a unilateral headache is also grounds to operations.
assume a causal connection when associated with ipsi- Many symptoms assumed to indicate a sinus ori-
lateral imaging abnormalities. gin of pain are not necessarily specific to rhino-
The ICHD-3 beta appendix also includes proposed sinusitis and can be seen in migraine. These include
criteria for a diagnosis of headache attributed to dis- a pain location in the face or area over the sinuses,
order of the nasal mucosa, turbinates, or septum, for changes in pain that correlate with the weather or sea-
which evidence is limited. This headache is hypothe- sons, and pain associated with rhinorrhea, sinus pres-
sized to occur due to irritation or inflammation related sure, or nasal congestion. Fluctuations in barometric
to a hypertrophic or other process in the nasal cav- pressure associated with changes in weather are a fre-
ity. It resolves fully after application of topical anes- quently reported migraine trigger, for example, but are
thetic or changes in severity in relation to gravitational often interpreted as a trigger for sinus problems. Many
or positional changes. Table 2.2 summarizes the clin- patients who treat presumed “sinus headaches” with
ical characteristics of headaches due to sinus or nasal decongestants report that this treatment is modestly
disease. effective and assume that this proves the headaches
are due to sinus problems. Decongestants, however,
are vasoconstrictors and furthermore are often sold
Discussion in combination with simple analgesics such as aspirin
As with the patient in the vignette, migraine is or acetaminophen. This combination of medications
commonly mistaken for a sinus problem. Patients is, not surprisingly, sometimes helpful for migraine
may receive repeated courses of powerful antibiotics headaches and thus response to this treatment does not
to treat the presumed sinus infection. Overuse of confirm the presence of sinus headaches. The major-
antibiotics is a serious public health problem, and ity of patients using these medications for presumed
can cause harm to individual patients as well. Sinus “sinus headache” will have an even better response
surgery performed in the mistaken belief that a to migraine-specific drugs. In one study, for exam-
patient with migraine has headaches due to sinusitis ple, the majority of self-diagnosed patients with “sinus

Chapter 2: Conditions that mimic primary headache

headaches” experienced greater than 50% of reduction cal, intermittent migraine headaches, which occurred
in pain with triptan use. on average once a week and responded well to suma-
In the Sinus, Allergy, and Migraine study (SAM), triptan. She was also taking six to eight tablets daily of
100 consecutive participants who responded to a ques- a butalbital-containing medication to control her daily
tionnaire and believed that they had sinus headache headache pain.
were enrolled. After a detailed history and physical The patient suspected her new headache was a
exam, only 3% were classified as having headache due symptom of chronic Lyme disease. She lived in an
to rhinosinusitis. The remainder met diagnostic crite- endemic area and had noticed a tick bite four years
ria for migraine with or without aura (52%), probable ago. She was immediately and empirically treated with
migraine (23%), various forms of chronic migraine or doxycycline. A friend, however, told her that Lyme dis-
other primary headaches (12%). In another study of ease often is not cured by short courses of antibiotic
almost 3000 patients with self- or physician-diagnosed treatment and advised that she see a Lyme specialist.
“sinus headaches,” 80% of the patients presenting with That physician had been treating her with regular high
“sinus headaches” met diagnostic criteria for migraine doses of vitamins and intravenous ceftriaxone through
with or without aura. It is thus abundantly clear that a percutaneous indwelling catheter. On review of sys-
migraine is often mistaken for sinusitis. tems, the patient had a large number of symptoms that
Rhinosinusitis and migraine may also coexist, she attributed to chronic Lyme disease.
however, complicating the diagnosis. The majority of Her neurologic examination was normal. A brain
patients who attribute their headaches to sinus dis- MRI and lumbar puncture were also normal. Cere-
ease do have prior diagnoses of allergic rhinitis, acute brospinal fluid (CSF) studies showed a normal protein
rhinosinusitis, and (to a lesser extent) chronic rhino- and no antibodies to Borrelia burgdorferi.
sinusitis, although the accuracy of those diagnoses is
not clear. In the SAM study, a third of patients expe-
rienced chronic cranial autonomic symptoms whose
How common is headache in patients with
exacerbations triggered migraine headaches, indicat- Lyme disease and is it likely that this
ing to the authors that seasonal allergic rhinitis could patient’s headaches were due to persistent
at times trigger migraine.
neurologic infection?
Headache is common in patients with Lyme disease
Diagnosis but by no means universal. About half of patients
Migraine without aura and chronic rhinosinusitis. with new positive Lyme serology will report headache
and headache is more common in patients with other
Tip symptoms of central nervous system and systemic
A diagnosis of headache attributed to rhinosinusitis involvement. In patients who only had positive CSF
must be made clinically and then confirmed by imag- serology and headache, the headache features could
ing or endoscopy. Treatments such as antibiotics or resemble either migraine or tension-type headache.
surgery that are given for mistaken diagnoses of sinusi- Rarely, headache will be the only presenting feature
tis expose patients to harm and prevent the use of more of Lyme disease, which is usually not diagnosed until
effective migraine-specific treatment. lumbar puncture shows characteristic CSF abnormali-
ties of high protein, lymphocytic pleocytosis, and Bor-
relia burgdorferi-specific intrathecal antibodies.
Chronic headache after Lyme disease It is, however, unlikely that this patient has persis-
tent neurologic infection or headaches due to Lyme
Case disease. She did not report the characteristic ery-
A 28-year-old woman with a prior history of episodic thema migrans rash of Lyme disease and was treated
migraine presented for evaluation of a new and differ- promptly with appropriate antibiotic therapy for her
ent type of headache, which she described as a con- tick bite. She has never had peripheral nervous sys-
stant, generalized pressure with a severity of 4/10 on a tem complications or CSF findings consistent with
scale of 0–10. It was not associated with nausea but she Lyme neuroborreliosis. Rather, her long-standing his-
was sensitive to light and noise. She still had her typi- tory of episodic migraine and daily use of combination

Chapter 2: Conditions that mimic primary headache

Figure 2.2 A deer tick of the sort that can transmit Lyme disease.

analgesics make medication overuse headache and

chronic migraine far more likely diagnoses than Lyme

What are the symptoms of chronic Lyme Figure 2.3 The characteristic erythema migrans rash of Lyme
disease. Source:
neuroborreliosis and how is it treated? migrans - erythematous rash in Lyme disease - PHIL 9875.jpg.
Chronic Lyme neuroborreliosis usually results from
untreated or unsuccessfully treated Lyme disease
which can become disseminated and affect the brain
The diagnosis of early, uncomplicated disease is usu-
and nervous system. Neurologic complications usually
ally based on clinical features such as a history of tick
appear a month or later after the tick bite. They include
exposure and the characteristic erythema migrans rash
Bell’s palsy, painful meningoradiculitis, or meningi-
(Figures 2.2 and 2.3). Fever, headache, joint pains,
tis. The latter is a rare complication of Lyme dis-
fatigue, and other constitutional symptoms may occur.
ease, possibly more common in children, and often
Serologic tests for Lyme disease are often negative in
associated with papilledema. Late complications of
this early stage of the disease, and are not required for
Lyme neuroborreliosis can include cerebral vasculitis,
diagnosis: the erythema migrans rash itself is diagnos-
progressive encephalitis, or encephalomyelitis. Lyme
tic. Most clinical guidelines agree that early infection
encephalitis characteristically produces evidence of
responds well to short courses of oral antibiotics such
nonspecific parenchymal involvement on neuroimag-
as doxycycline.
ing. In the absence of parenchymal involvement, oral
Subjective symptoms that persist six months or
antibiotic treatment is usually considered satisfac-
more after treatment of Lyme disease are sometimes
tory for this stage of the disease, although some
termed “post-Lyme disease syndrome.” Headache is
patients may continue to have troublesome symptoms
often a prominent symptom in such patients. How-
and require treatment with intravenous ceftriaxone or
ever, it seems unlikely that the patient in this case ever
penicillin G.
had Lyme disease. Instead, she has a long history of
episodic migraine, and her chronic headaches may be
Discussion due to medication overuse or represent the develop-
The spirochete Borrelia burgdorferi, which is trans- ment of chronic migraine. Most authorities agree that
mitted to humans through the bite of deer ticks, is there is little good evidence that lengthy treatment
endemic in the United States. Infection risk is high- with antibiotics improves outcomes in patients with
est in wooded areas of New England, the northern post-Lyme disease syndrome. In fact, many harmful
Pacific coast, and the area around the Great Lakes. consequences of such prolonged treatment have been

Chapter 2: Conditions that mimic primary headache

reported, including biliary damage as a result of pro-

longed use of ceftriaxone.

Medication overuse headache; probable chronic mi-
graine without aura.

Persistent Lyme disease is an unlikely cause of
headache in patients who do not have objective find-
ings of central nervous system involvement.

Headache in the setting of

temporomandibular abnormalities
Case Figure 2.4 The temporomandibular joint. Source: http://en.
A 42-year-old woman was seen in a headache clinic for
evaluation of chronic headache. She reported having
headaches since high school and described herself as
someone who “always has headaches.” The headaches How likely is this daily headache to be due
were initially intermittent but had become daily over
the last three years. They were worse in the morn- to pathology of the TMD?
ings. She reported generalized head pain but the pain It can be difficult to make a diagnosis of headache
was most severe around the temples and jaw bilater- attributed to TMD because many people grind or
ally. The pain was described as dull and achy and rated clench their teeth and occasional jaw pain is common.
4 on a scale of 0–10. When severe, the pain would ICHD-3 beta requires that the diagnosis can only be
become throbbing and was associated with photopho- made when there is clinical or radiologic evidence of
bia, phonophobia, osmophobia, nausea, and some- pathology of the jaw or TMJ or associated structures.
times vomiting. Stress, anxiety, and worsening jaw ten- Furthermore, some evidence of causation is required
sion seemed to make the headaches worse, and they in the form of development or change of the headache
also seemed worse around her menstrual periods. in conjunction with any pathologic changes, or exacer-
Recently the patient had become aware that she bation of pain by active or passive jaw movements or
clenched her teeth constantly. After evaluation by provocative maneuvers. If the joint pathology is uni-
a dentist she was diagnosed with displacement of lateral, the headache pain must be ipsilateral to the site
the temporomandibular (TMJ) joint (Figure 2.4), a of the pathology.
form of temporomandibular disorder (TMD). She also A number of historical or examination features can
received diagnoses of bruxism and myofascial pain be useful when trying to assess the possibility of TMD
in the muscles of mastication. She was provided with pathology. These include the presence of localized jaw
a night guard to prevent nighttime bruxism, but her pain or physical findings such as impaired jaw range
headaches persisted. She suspected that TMJ dysfunc- of motion, joint sounds with movement (e.g. clicking),
tion was the cause of the chronic headache and had and muscle and joint tenderness to palpation.
purchased several custom-made oral appliances from In this case we are not provided with the details
her dentist, aimed at reducing bruxism and improving of the initial dental examination. On examination
her pain. Because the cost of these appliances was not in the headache clinic, however, the patient did
covered by insurance, however, the patient had paid for not give a clear history of pain development that
them out of pocket. Her headaches continued, though, correlated closely with demonstrated TMJ pathol-
and surgery on the TMJ had been advised. However, ogy. Furthermore, there was no impairment of jaw
before proceeding she wanted a second opinion. opening and minimal pain on palpation over the

Chapter 2: Conditions that mimic primary headache

Table 2.3. Clinical features of headache attributed to Table 2.4. Selected causes of postpartum headache
temporomandibular disorder
Primary causes:
Imaging or clinical evidence of pathology in the TMJ or nearby Migraine
structures or masticatory muscles Other primary headaches
Evidence of a causal link between headache and this pathology, Secondary causes:
as demonstrated by onset or worsening of pain in association Vascular:
with the pathology, or worsening of pain with active or passive Ischemic stroke
movements of the jaw or with provocative maneuvers including Hemorrhagic stroke, including subarachnoid hemorrhage
pressure on masticatory muscles or the TMJ Venous sinus thrombosis
Reversible cerebral vasoconstriction (RCVS)
Arterial dissection
temporomandibular joint. Many details of the patient’s Vasculitis
Hypertension/Posterior reversible encephalopathy
history seemed more compatible with a diagnosis of syndrome (PRES)
chronic migraine, such as the associated headache fea- Nonvascular:
tures of nausea, vomiting, and phono and photopho- Preeclampsia/eclampsia
bia. She was advised to postpone treatment of the Idiopathic intracranial hypertension
Post-dural puncture headache
TMJ and was treated with more aggressive preventive
therapy for migraine. Her headaches became episodic
and were more easily controlled with migraine-specific episodic migraine with aura that had improved during
therapy. the pregnancy.
After delivery she had the acute onset of a severe
Discussion bifrontal headache, associated with nausea and blurred
Jaw pain is common and many patients seen in spe- vision, which she identified as similar to her prior
cialty headache practice have already been evaluated migraines but much more severe. The headache and
and treated by a dentist for presumed TMJ pathology. nausea seemed to worsen when she was sitting up. Her
It is difficult to estimate the true prevalence of TMDs. neurologic examination showed mild diffuse hyper-
One large review suggested that they occur in roughly reflexia but was otherwise normal.
10% of the population, mostly in middle-aged adults. Because of the acute onset of headache an
Like migraine, they are more common in females than evaluation, including noncontrast head CT and
males. They may therefore coexist with migraine in MRI/MRA/MRV with contrast, was done and was
many patients, but this does not mean they are a causal negative. Urinalysis showed no proteinuria and her
or contributing factor to headache problems. ICHD-3 blood pressure was 132/85 mm Hg. A parenteral anal-
beta criteria for a diagnosis of headache attributed to gesic medication was administered. By the time the
TMD are summarized in Table 2.3. diagnostic testing was completed her headache had
started to improve and she declined a recommended
lumbar puncture.
Chronic migraine.
What is the differential diagnosis of
Tip postpartum headache?
Familiarity with the presentation and findings in TMD The differential diagnosis for postpartum headache
as distinct from migraine can help prevent confusion is lengthy. Selected causes are listed in Table 2.4.
when migraine masquerades as TMD. They include both benign primary headaches such
as migraine, and dangerous secondary causes of
headache such as cerebral venous thrombosis. A
Postpartum headache detailed headache history and thorough neurologic
examination can be helpful in differentiating between
Case primary and secondary causes of headache, but a
A 24-year-old woman developed acute headache six high suspicion of secondary causes is appropriate.
hours after a normal spontaneous vaginal delivery, for Postpartum headaches that are similar to a patient’s
which she had epidural anesthesia. She had a history of pre-existing headaches, particularly if the headaches

Chapter 2: Conditions that mimic primary headache

persisted during pregnancy, are probably more likely with a prior history of migraine. Other factors asso-
to be benign. Likewise, a nonfocal neurologic exami- ciated with an increased risk of postpartum headache
nation is reassuring. In this case, the finding of hyper- are known dural puncture, previous headache history,
reflexia is of unclear significance. Furthermore, her and multiparity.
report of an acute headache more severe than her usual Secondary causes are responsible for 25–50% of
headaches is concerning. postpartum headaches; these are discussed elsewhere.
Among the primary headache disorders that produce
Should further diagnostic testing be done? postpartum headache, recurrence of migraine is prob-
ably the most frequent. In one study of 1000 women
If so, what? with postpartum headache, migraine or another pri-
Headache characteristics that should prompt workup mary headache disorder was identified as the etiology
for secondary headache in the postpartum period are in 75% of cases. Migraine may reappear within a week
similar to the “red flags” that warrant workup in other to one month after delivery and has been attributed to
settings. These include a new headache or change causes such as declining estrogen levels, sleep disrup-
in an established headache pattern, changes in men- tion, and stress.
tal status, new focal neurologic deficits, uncontrol- Even when a diagnosis of migraine is made in a
lable vomiting, or intractable headache. A history of patient with postpartum headaches we believe it is
primary headache disorder such as migraine is not prudent to avoid vasoconstrictive medications such as
always reassuring in the postpartum setting because triptans because the postpartum period in migraineurs
migraine is associated with an increased likelihood may be a situation of heightened susceptibility to the
of several dangerous causes of headache, including vasoconstrictive effects of triptans.
peripartum stroke, preeclampsia, and angiopathy or
reversible cerebral vasoconstrictive syndrome.
Appropriate diagnostic investigation depends on Diagnosis
the suspected cause of secondary headache. Imaging Migraine without aura.
is frequently needed to assess for the possibility of
intracranial vascular problems. A head CT and lum- Tip
bar puncture can be performed to rule out intracere- A high index of suspicion for dangerous causes of post-
bral or subarachnoid hemorrhage, and MRA/MRV are partum headache is appropriate, but the most common
appropriate to evaluate for vasculitis, vasoconstrictive cause of postpartum headache in a known migraineur
syndromes, or cerebral venous thrombosis. Angiog- is migraine.
raphy may be needed when cerebral vasoconstriction
or aneurysm are strongly suspected. An MRI of the
brain with contrast may show meningeal enhance- A middle-aged man with low-lying
ment in cases of post-dural puncture headache, and a cerebellar tonsils
lumbar puncture may show low opening pressure. If
preeclampsia is suspected, monitoring for proteinuria, Case
hyperreflexia, and hypertension is warranted. A 49-year-old male was evaluated for frequent
In this case, the initial workup for causes of sec- headache. He reported a family history of migraine.
ondary headache was negative. Because the patient’s He had no history of childhood headache or of peri-
headaches started to resolve after the MRI/MRA/MRV odic symptoms related to migraine such as motion
and this testing was negative, a lumbar puncture was sickness. Headaches had begun in his mid 30s. They
deferred. She was monitored in the Mother and Baby were infrequent and the pain was generalized and
Unit for an additional three days without recurrence of moderate in intensity. The headaches could last
headache, and felt well at the time of discharge. for days and generally did not respond to simple
analgesics. They were attributed to sinusitis.
Discussion When the patient was in his late 30s the headaches
About one-third of women experience headache in the became more severe and were associated with migrain-
postpartum period, and both primary and secondary ous features such as nausea. He sought treatment
postpartum headaches are more frequent in women and was diagnosed with migraine. He was prescribed

Chapter 2: Conditions that mimic primary headache

Table 2.5. Clinical and MRI characteristics of headache

associated with Chiari malformation, type I
Headache that is occipital or suboccipital or precipitated by
Valsalva maneuvers or lasts less than 5 minutes OR headache that
has developed or changed in conjunction with a CM1 OR
headache associated with posterior fossa dysfunction, indicated
by lower cranial nerve or cervical spinal cord dysfunction or
otoneurologic or brainstem symptoms such as dizziness, vertigo,
Demonstration of a CM1 by MRI:
ࣙ5 mm caudal descent of the cerebellar tonsils OR
ࣙ3 mm caudal descent of the cerebellar tonsils plus posterior
fossa crowding manifested by:
compression of the posterolateral cerebellar CSF spaces
reduced height of the supraocciput
increased slope of the tentorium
kinking of the medulla oblongata

Table 2.6. Clinical characteristics of cough headache

Headache precipitated only by coughing, straining, or Valsalva
Sudden in onset
Location not specified
Lasts 1 second to 60 minutes
Figure 2.5 A sagittal T1 MRI image showing moderate tonsillar

tion prescribed by another physician. He was work-

oral zolmitriptan 5 mg. This effectively aborted indi- ing but had been unable to exercise because of the
vidual attacks of headaches but headache frequency headache. He had gained 30 pounds in the preced-
increased and by his mid 40s the patient had two to ing year. His sleep was fragmented and he reported
three episodes of severe headaches weekly. In addition, daytime somnolence. His neurologic examination was
he noted gradual onset of a constant low-grade head normal. A lumbar puncture showed a normal opening
and posterior occipital pressure, which he described pressure and no CSF abnormalities.
as “like a tourniquet” or tight band around his head
and neck. The pain was not postural. There were no How can headache attributed to a Chiari I
abnormalities on neurologic examination. An MRI
of the head showed mild cerebellar tonsillar descent malformation be diagnosed?
(Figure 2.5) into the foramen magnum. Headache related to a CM1 is characteristically located
The patient was seen by a neurosurgeon, who in the occipital region of the head and is intermittent
believed that his symptoms were related to a Chiari 1 rather than constant. It is often provoked by maneu-
malformation (CM1) and recommended surgery. The vers or activities that temporarily raise CSF pressure,
patient sought a second opinion from another neuro- such as coughing or sneezing. It generally does not
surgeon, who strongly advised against surgery and have features more characteristic of migraine or other
referred him for headache consultation. primary headaches, such as associated nausea or pho-
At the time of his visit to the headache center tophobia. Table 2.5 lists the clinical and imaging fea-
the patient was experiencing three headache exacer- tures of this disorder.
bations weekly, each lasting one to two days, super- CM1 is easily confused with cough headache,
imposed on the generalized tight sensation that was which is described in Table 2.6. Cough headache is a
always present. The severe headaches were present primary headache disorder in which headache is pre-
when he awakened. They were described as throb- cipitated by cough or other Valsalva-type maneuvers,
bing, generalized, and associated with photophobia but is not associated with identifiable structural abnor-
and nausea. The patient was taking frequent over-the- malities. It is a diagnosis of exclusion, and should be
counter medications to treat his headaches and was made only after imaging has been done to assess for
occasionally using a butalbital combination medica- possible abnormalities of the posterior fossa.

Chapter 2: Conditions that mimic primary headache

Roughly 40% of patients who present with appar- this was an incidental finding and that he had a pri-
ent “cough headache” in fact turn out to have CM1. mary headache disorder for which medical manage-
Other reported secondary causes of headache precipi- ment was indicated.
tated by cough or Valsalva include carotid or vertebro- The patient improved with a brief course of
basilar diseases and cerebral aneurysms. Thus, neuro- oral steroids and discontinuation of daily analgesics.
imaging plays an important role in differentiating Amitriptyline and topiramate were added as pre-
secondary from primary causes of cough headache. ventive agents and a triptan drug was restarted for
CM1 can also present as a more protean chronic headache exacerbations. He was evaluated and found
headache, with or without exertional features, usu- to have sleep apnea, for which treatment was started.
ally located in the posterior occipital or suboccipi- At a follow-up visit, he considered his headaches well
tal regions. If associated with other neurologic signs controlled on this regimen and had been able to
and symptoms, the diagnosis is usually not particularly resume daily exercise.
challenging. Often, however, headache is either the ini-
tial or only manifestation of the condition.
A diagnosis of symptomatic CM1 does not seem Discussion
very likely in the patient described in this case, despite A Chiari type 1 malformation is a cerebellar anomaly
the documented cerebellar tonsillar descent. He does in which there is descent of the cerebellar tonsils to
not describe headache in association with cough and at least 3–5 mm below the foramen magnum. This
his clinical presentation is more characteristic of abnormality often occurs alone but can also be associ-
migraine. Cerebellar tonsillar descent or “brain sag” ated with a spinal cord syrinx, hydromyelia, or hydro-
can occur in patients with spontaneous intracranial cephalus. Although generally considered to be a con-
hypotension, so a lumbar puncture is recommended genital lesion, symptoms often do not begin until
in patients who have any cerebellar tonsillar descent adulthood. This delay may be due to altered CSF
to assess the possibility of a CSF leak. dynamics with aging or to changes in connective tis-
sue that occur over time. Many people with docu-
What treatment should be recommended to mented cerebellar tonsillar descent – up to a third – are
this patient? Descent of the cerebellar tonsils is a frequent inci-
The patient likely had a past history of headaches com- dental finding on MRI performed for other reasons.
patible with a diagnosis of episodic migraine with- It is fairly common, with an incidence estimated to
out aura. Unusual presenting features were the late be around 0.5% of the population and a female to
age of onset and the initial nondescript features of male ratio of 3:2. There is no clear correlation between
the headache. Later, though, a more typical pattern of the degree of cerebellar tonsillar descent and the clin-
migraine emerged and his headache pattern was more ical severity of headache. Thus, radiologic findings
easily recognizable as a transformation from episodic alone should almost never prompt surgical interven-
to chronic migraine. Although the posterior occipital tion. Rather, the clinical context of the findings must be
neck pressure he reported might suggest a secondary carefully considered. Headache related to CM1 is often
headache cause such as CM1, there were few other located in the occipital or suboccipital regions and is
features in the history or examination to suggest this often worsened by cough Valsalva, postural change, or
diagnosis. Additionally, when carefully measured, the bending over.
degree of tonsillar descent on the patient’s MRI was less MRI correlates of CM1 have been characterized
than 3 mm, meaning that he did not meet radiologic based on large case series of affected patients. Herni-
criteria for a CM1. ation of the cerebellar tonsils 3–5 mm below a line
At times the historical features of a headache pat- drawn between the tip of the clivus and the rim of the
tern do not allow a clear diagnosis to be made and fur- foramen magnum is the most consistently used mea-
ther testing such as diagnostic neuroimaging may be surement to determine the diagnosis. Lesser amounts
warranted. If misinterpreted, however, imaging may of descent are considered to represent merely ton-
confuse the situation further. This patient was advised sillar ectopia, and this is generally thought to be
that although it was possible the MRI findings were asymptomatic. Other MRI features may also be
related to his headache, it was much more likely that present, including elongation of the tonsils, caudal

Chapter 2: Conditions that mimic primary headache

Table 2.7. Classification of Chiari malformations cheek or teeth. The pain was described as sharp, stab-
Malformation Description of congenital bing, and excruciating: “definitely a 10/10, worse than
type malformation childbirth.” Individual episodes of pain lasted up to 30
Type I Elongation of the tonsils and the medial parts minutes and usually occurred at night.
of the inferior lobes of the cerebellum into The patient had seen a dentist and undergone
cone-shaped projections, which accompany removal of her wisdom teeth with no benefit. She
the medulla oblongata into the spinal canal
was then diagnosed with TMJ disorder and provided
Type II Displacement of the parts of the inferior with a night guard. Her headache pattern seemed
vermis, pons, and medulla oblongata
together with elongation of the fourth to improve after that, but the next year headaches
ventricle (most cases are associated with recurred. Later she returned to the dentist. Several
spina bifida) fillings were replaced and a suspected diseased tooth
Type III Herniation of the entire cerebellum into the was removed. With each procedure she would seem to
cervical canal. This is associated with severe
neurologic deficits
improve transiently before experiencing a return of her
prior headache pattern. During a typical headache the
Type IV Cerebellar hypoplasia
patient paced the room, noting that it was better for
her to walk than to remain in bed with a headache.
Her primary care physician had suggested that she
displacement of the brainstem and fourth ventricle, might have cluster headache and referred the patient
crowding of the posterior fossa with obliteration of for evaluation. At the time of the neurologic evaluation
the subarachnoid space, and bulbo-cervical kinking, she reported her headaches were unresponsive to treat-
an increased angle formed by the brainstem and cervi- ment with opioid analgesics and her most recent bouts
cal spine at their margin. Figure 2.5 illustrates some of of headache had been separated by only six months.
these findings. The features of other degrees of Chiari She was frustrated and commented: “I’ve been search-
malformation are described in Table 2.7. ing for an answer for 20 years. It’s only on the left
Chronic migraine; tonsillar ectopia. What is the most likely diagnosis in
this case?
Tip This patient does meet diagnostic criteria for a diagno-
The clinical context of headache occurring in the sis of episodic cluster headache. Despite this, an accu-
setting of low-lying cerebellar tonsils must be care- rate diagnosis was not made until 20 years after her
fully considered. Caution is warranted in making a headaches began. Cluster headache is characterized by
diagnosis of headache attributed to CM1, in order a severe unilateral orbito-temporal pain lasting from
to avoid exposing patients to possible unnecessary 15 minutes to 3 hours associated with ipsilateral auto-
surgery which can cause serious complications. nomic features such as tearing, nasal congestion, eyelid
edema, ptosis, or conjunctival edema.
Cluster headache presenting as Although cluster headache is more common in
men, it does occur in females. Recent epidemiologic
a dental problem data suggest that the sex ratio is less lopsided than
originally suspected. Currently the male:female ratio
Case is estimated to be about 2:1.
A 49-year-old female was evaluated for a 20-year his- Bouts of episodic cluster headache typically last
tory of episodic left-sided headache. Headaches typi- 2–12 weeks and are separated by pain-free periods of
cally occurred daily for periods of one to two months at least 2 weeks but typically longer. Cluster headache
during the spring every year, although she could go for is termed chronic when a remission is absent for a year
several years without having any pain episodes. During or more or is present but of short duration, less than 14
her one- to two-month bouts of headache, she experi- days in length.
enced headaches located behind the left eye. They were In this vignette, though the pattern meets crite-
associated with tearing and the pain radiated to the left ria for a diagnosis of episodic cluster headache, the

Chapter 2: Conditions that mimic primary headache

patient’s bouts are becoming more frequent. It is pos- A revision of the ICHD criteria for cluster to include
sible that this is the beginning of a conversion from this localization has been suggested as a possible step
episodic to chronic cluster headache. It is not uncom- to mitigate the confusion.
mon for physicians or patients to attribute the uncom- Worsening or triggering of a cluster headache pat-
mon and unilateral pattern of cluster headache to local tern after dental procedures has also been reported,
disease in the sinuses or teeth and direct inquiry to explained in part by deafferentation of pain fibers in
those disorders. the region of the dental procedure leading to trigger-
ing of brainstem and spinal cord pain networks. The
possibility that a dental procedure might precipitate
Was the dental work this patient had a cluster attack further underlines the need to elimi-
indicated? nate diagnostic confusion and avoid unnecessary pro-
It is unlikely that the dental work this patient had was cedures in these patients.
in fact necessary. A dentist who is unfamiliar with the Given the pain severity and associated disability of
diagnosis of cluster headache can easily attribute clus- cluster headache, it is important for dental personnel
ter headache pain to a local dental process. The uni- to recognize the features of this headache pattern; they
lateral nature of the pain, as well as the fact that pain may well be the first to see and diagnose the patient.
is often felt in the jaw area, means that patients fre- Those in dentistry should redouble their efforts to ren-
quently seek dental evaluation. Studies indicate that der an accurate diagnosis and to avoid unnecessary
almost half of cluster patients will have visited a and inappropriate dental treatments. Timely manage-
dentist for evaluation before the diagnosis of clus- ment in order to minimize patient disability is the
ter headache is eventually made. Unfortunately, many goal.
undergo repeated, invasive dental procedures in an
attempt to treat the pain. As with the patient described Diagnosis
in this vignette, it can be years before an accurate diag- Episodic cluster headache.
nosis of cluster headache is made. This is unfortunate
because medical therapy for cluster headache is often
remarkably effective. Tip
Cluster headache is a unilateral pain disorder that is
frequently mistaken for a dental problem. Familiar-
Discussion ity with the characteristic pattern of cluster headache
For a variety of reasons, episodic cluster can be a chal- is important to avoid costly and unnecessary dental
lenging diagnosis and a delay in diagnosis from onset procedures.
of symptoms of up to five years is described, though the
delay may be longer in females with cluster. Episodic An older woman with focal neurologic
cluster will at times present to the dentist and the diag-
nosis can be missed. In one study just fewer than half deficits but no headache
of patients eventually diagnosed with cluster presented
for dental evaluation as part of the initial evaluation of Case
their pain and many of these patients received unnec- A 71-year-old woman was seen in the emergency
essary or inappropriate dental management for their department. She reported that five hours ago, while
symptoms. reading a magazine, she suddenly developed a visual
The current International Classification of disturbance. She found the type on the magazine page
Headache Disorders (ICHD) criteria for cluster blurry and could not make out the individual letters.
include an orbital, supraorbital, and/or temporal She was able to see the face of her husband. After a
localization for the pain. However, ipsilateral radia- few minutes she noticed a thin bright white line in the
tion of the pain to the face and teeth is not unusual in upper part of her visual field in the right eye. These
cluster. It is well documented that cluster may in part symptoms continued for 20 minutes, when she devel-
present with a midface localization or radiation of oped numbness of her left cheek that gradually spread
pain, and this is thought perhaps the main explanation to the side of her nose and corner of her mouth. Her
for diagnostic confusion among dental practitioners. tongue was not numb. Five minutes later her husband

Chapter 2: Conditions that mimic primary headache

noticed that she was unable to express herself and that Table 2.8. Clinical characteristics of “late-life migraine
accompaniments” From C. Miller Fisher
her speech consisted of nonsense words. The visual
changes lasted 40 minutes, the speech problems for 10 r Most common aura is visual, followed by paresthesias,
minutes and the numbness for an hour. She had no then aphasia/dysarthria, then (rarely) paralysis
r Symptoms usually build, one upon another
associated headache. r There may be a spread or “march” of symptoms
The patient had experienced a similar episode one r Progression from one symptom to the next is common
r ࣙ2 similar episodes required for confirmation
month earlier. She had been seen in the emergency r Duration of each individual symptom is usually 15–25
department and had a workup including brain imag- minutes
ing, Holter monitor, and carotid studies, all of which r Symptoms may or may not be followed by headache
r Generally indicates a benign course
were negative. She was diagnosed with a transient r Exclude other conditions. Requires normal imaging
ischemic attack and instructed to take 81 mg of aspirin
daily. She had osteoarthritis and hypertension that was
well controlled on lisinopril and was otherwise well.
She was not a smoker and her cholesterol levels were What is the epidemiology of late-life
normal. migraine accompaniments, and should they
In the past she had experienced periodic throb-
bing, unilateral headaches preceded by neurologic be treated?
symptoms similar to those she had now. Her pre- An analysis of data from the Framingham Study
vious headaches had been diagnosed as migraine showed that migrainous visual accompaniments in late
but she had not had any severe headaches for the life are not rare. In over 2000 queried subjects, 1.23%
last 20 years. After hearing about the patient’s most reported migrainous visual symptoms. The major-
recent event, the neurology resident on call diag- ity of subjects reported stereotyped episodes and in
nosed a second transient ischemic attack and was con- over three-quarters of these subjects the episodes had
cerned that this had occurred despite aspirin ther- begun after the age of 50. Well over half of peo-
apy. She wondered whether the patient should be ple who experienced such episodes reported that the
placed on additional antiplatelet therapy, a statin, or visual events occurred without headache, and the typ-
anticoagulated. ical duration of an episode was 15–60 minutes. In con-
trast to a group of subjects who had been diagnosed
with transient ischemic attacks, patients with late-life
This patient had a negative workup a month migraine accompaniments were far less likely to expe-
ago. Should she be reinvestigated now or rience stroke subsequently. The authors of this study
concluded that these events were not rare, were benign,
could this be a benign process? and that “invasive diagnostic procedures or therapeu-
The patient probably does not need additional test- tic measures are generally not indicated.”
ing or investigation for this second episode of tran- Although antiplatelet therapy such as low-dose
sient neurologic deficits, but it is easy to understand aspirin is sometimes used in patients with transient
the neurology resident’s anxiety. The patient is in an ischemic attacks, there is no evidence to support its
age group where transient ischemic attacks and strokes use in patients with late-life migraine accompaniments
are common. She also gives a clear history of neu- or other forms of aura. In the past some have advo-
rologic deficits, although the reversible nature of her cated this approach on the assumption that decreased
symptoms is more suggestive of a transient ischemic platelet aggregability might reduce the risk of subse-
attack than a stroke. However, her symptoms have quent ischemic events in such patients.
been fully evaluated with no suggestion of a throm- This practice is called into question, however, by a
botic or embolic cause. In fact, this patient’s symptoms recent analysis of data from the Women’s Health Study.
are likely to be what the famous neurologist C. Miller In subgroup analyses of this randomized trial, women
Fisher termed “late-life migraine accompaniments.” who had migraine with aura who had received 100 mg
Fisher’s “rules” for the diagnosis of late-life of aspirin every other day actually had a higher risk
migraine accompaniments (Table 2.8) are still useful of subsequent myocardial infarction than women who
clinically, and help to put the case of the patient in the had not received aspirin. Although these data come
vignette into clearer context. from subgroup analyses, and the risk was limited to

Chapter 2: Conditions that mimic primary headache

women who had smoked or had hypertension, they do and a scotoma. After about an hour the scotoma grad-
sound a cautionary note. In the absence of evidence ually faded and he developed a right-sided headache
that aspirin is helpful, and some suggestion of pos- that responded to treatment with zolmitriptan. The
sible harm, we do not currently recommend treating scintillations, however, have continued. In the past the
patients with aura with aspirin therapy. visual symptoms associated with his headaches had
never lasted longer than 40 minutes. Initial neurologic
Discussion examination, including fundoscopy, was normal. The
patient was sent for urgent ophthalmologic consulta-
Attacks of aura without migraine become more com-
tion, which showed no apparent ocular disease but
mon as patients with migraine age. Most patients with
did reveal an upper and lower quadrant right visual
late-life migraine accompaniments have a prior history
field deficit. A CT of the head with and without con-
of migraine and aura. In our practice we are reluctant
trast showed no abnormalities. The patient was oth-
to make a diagnosis of aura without migraine in the
erwise well and hematologic and chemistry tests were
absence of a previous history consistent with migraine.
normal. About six hours after onset the visual symp-
Headache expert Dr. William Young, writing about
toms resolved spontaneously. An MRI scan of the head
Dr. Fisher’s masterful description of this clinically
done a month after the emergency department visit
common phenomenon, has commented that “Once
also proved to be normal.
you learn of this symptom complex, you see it over and
The patient then disappeared from care, reappear-
over again. It is easy to diagnose . . . even obvious . . . No
ing four years later. He reported that he traveled fre-
one since has changed [Fisher’s] clinical characteri-
quently for his work as a consultant. He had continued
zation, much less improved upon it. His writing is
to have episodes of prolonged aura interspersed with
strange by today’s standards, with his introduction,
episodes of typical visual aura lasting no longer than
results section, and conclusion all mixed together, and
40 minutes. On several occasions he had sought eval-
much of the article is composed of illustrative cases.
uation in emergency departments in different cities,
But once you slog through, you really will have the fla-
where he had undergone several repeat CT scans, elec-
vor of the condition, and when you are confronted with
troencephalogram (EEG), and lumbar puncture with
it, you are reassured, you reassure your patient, and
no abnormalities found.
common sense rules your management.”

Diagnosis What is the likely diagnosis in this case?

Typical aura without headache (late-life migraine This patient experiences aura symptoms that are com-
accompaniments or “acephalgic migraine”). pletely characteristic of his typical visual aura with the
exception of their duration. Such an occurrence nat-
Tip urally raises the suspicion of cerebral infarction, but
testing in this case does not show evidence of ischemia.
Aggressive treatment or diagnostic investigation is not
Unfortunately, the current ICHD-3 beta does not pro-
warranted in patients with a clinical scenario consis-
vide a diagnostic label for his episodes of prolonged
tent with late-life migraine accompaniments. While
aura. The criteria for typical visual aura specify that
antiplatelet therapy such as aspirin may be helpful for
aura symptoms should not exceed 60 minutes. The
patients with transient ischemic attacks, there is no
criteria also include the diagnosis of “persistent aura
evidence to support its use in patients with late-life
without infarction” for cases where aura symptoms last
migraine accompaniments.
longer than a week in the face of normal test results.
Patients whose auras fall between 60 minutes and
Prolonged aura a week, or who have repetitive attacks of aura with
very short intervals of normality reside in a diag-
Case nostic “no man’s land.” Many terms have been used
A 36-year-old man with a 20-year history of migraine to describe this clinical situation, including “compli-
with typical visual aura presented in the emergency cated migraine,” “sustained visual aura,” “persistent
department. He reported that four hours previously migraine aura,” and “migraine aura status.” Because
he had one of his typical auras, including scintillations most patients with prolonged aura have other attacks

Chapter 2: Conditions that mimic primary headache

that fulfill criteria for some other form of aura, the

authors of ICHD-3 beta recommend headaches in
those cases should be coded to that other diagnosis.
Otherwise, since they fulfill all but the duration crite-
rion for a diagnosis of typical aura, they should be clas-
sified as probable migraine with aura, specifying the
atypical feature (prolonged aura or acute-onset aura)
in parentheses.

What is the differential diagnosis of

prolonged aura?
The differential diagnosis of prolonged aura includes
vertebrobasilar transient ischemic attack, carotid or Figure 2.6 This drawing illustrates visual abnormalities of the kind
vertebral dissection, cerebral vasculitis, reversible described by a patient experiencing repeated attacks of visual aura.
After Haas. Prolonged migraine aura status. Ann Neurol. 1982;11:
cerebral vasoconstrictive syndrome, occipital lobe 197–9. From Used with permission.
epilepsy, or even cerebral venous thrombosis. In the
case of this patient, those possibilities were elimi-
on which to base treatment recommendations. Case
nated by normal testing results. The passage of time
reports provide empirical support for trials of lamot-
without the development of additional problems also
rigine, acetazolamide, or valproic acid.
makes a dangerous explanation for his prolonged
auras unlikely. Unfortunately, this patient’s peripatetic
lifestyle meant that he was subjected to repeat diagnos- Diagnosis
tic testing to rule out dangerous diagnoses that might Migraine with aura.
have accounted for his presentation.
Discussion Aura symptoms that last longer than an hour but less
Persistent aura symptoms are not especially common than a week are difficult to classify. In patients who also
but numerous case reports make it clear that they do have attacks of aura of typical duration, it is reasonable
occur. Persistent visual aura is often distressing for to assign a diagnosis of migraine with aura.
patients, who frequently report that they are either
unable or afraid to drive. Figure 2.6 shows a draw- A middle-aged woman with vertigo
ing that illustrates visual abnormalities similar to those
depicted by a 70-year-old man who suffered from and a history of migraine
repetitive attacks of aura for five weeks; looking at
this it is not hard to understand how disabling per- Case
sistent attacks of aura can be. In commenting on the A 54-year-old woman was referred to the headache
patient’s drawing, noted aura experts report that “Each clinic by her primary care physician for possible
repetition began with slowly undulating thick gray migrainous vertigo. She had a history of migraine
lines, which changed in a few minutes into a pinwheel without aura in her 20s, which had largely abated after
of bright whirling color in his left visual field. Sev- the birth of her last child in her late 30s. About three
eral minutes later this image slowed down and dis- years ago she started having episodes of vertigo, lasting
appeared. After more than a week of suffering these about two hours and initially occurring twice a month.
hallucinations, he also developed brief attacks of ‘elec- When she had the vertigo, turning her head quickly
trical’ paresthesias in his left hand. These were less fre- exacerbated it. It did not seem to be provoked by any
quent than the visual phenomena and alternated with environmental factors such as dehydration. There was
them irregularly. Throughout his ordeal, he had a dull no headache associated with the vertigo, though she
headache over his right eye.” Because this is a rela- did note photo and phonophobia during some of the
tively rare condition, there are no large clinical trials episodes.

Chapter 2: Conditions that mimic primary headache

Table 2.9. Clinical characteristics of migraine with brainstem toms in addition to those believed to originate in the
Aura must be followed within an hour by headache
Aura must occur with at least two brainstem symptoms,
which include decreased level of consciousness, tinnitus, What treatments are available to
vertigo, dysarthria, diplopia, hypacusis, or ataxia
this patient?
Each symptom can last 5–60 minutes, at least one symptom
is unilateral, and at least one spreads gradually and/or at Until recently, research regarding treatment for
least two symptoms occur in succession vestibular migraine has been limited by the absence
of clear, agreed-upon diagnostic criteria. There have
been no randomized controlled trials of treatments
She had seen an otolaryngologist. Her examina- for vertigo linked to migraine and high quality evi-
tion, done at a time when she was not symptomatic, dence to guide treatment is therefore lacking. One
was normal. Specifically, a Dix–Hallpike maneuver did small and underpowered trial suggested benefit from
not provoke nystagmus. An MRI of the brain was nor- zolmitriptan for acute treatment. Larger chart review
mal. She was diagnosed with benign positional parox- or survey studies suggest that medications prescribed
ysmal vertigo and started on meclizine, which was for migraine were generally effective in treatment of
ineffective and made her tired. Her primary care physi- the vertigo in patients with both vertigo and migraine.
cian thought the vertigo might be related to her history Small interventional studies, typically open label, have
of migraine and referred her to the headache clinic. evaluated beta-blockers, tricyclic antidepressants, cal-
cium channel blockers, and antiepileptic drugs. One
Could this patient’s vertigo be migrainous interventional study included a dietary modification
that appeared to be beneficial for a subset of patients,
in origin? although in our view such open-label results should
The first two versions of the ICHD did not include be interpreted with caution.
criteria for migrainous vertigo or migraine-associated Given the lack of a solid evidence base to guide
vertigo. ICHD-3 beta, however, lists criteria for treatment decisions, we typically start by treating
vestibular migraine in the appendix, which is a place vestibular migraine as we do any other form of
for candidate criteria for disorders that need further migraine, with an emphasis on prevention if attacks are
validation and testing. These criteria describe migrain- frequent. Specific preventive medications to be con-
ous vertigo as the occurrence of moderate to severe sidered include propranolol, amitriptyline, and top-
vestibular symptoms lasting 5 minutes to 72 hours, iramate. Lifestyle modifications including adequate
where at least half of the episodes are associated with and regular sleep, maintaining good hydration, and
headache, visual aura, or photophobia and phonopho- a healthy diet are also emphasized. For treatment of
bia. These symptoms must occur in a patient who acute episodes of headache or vertigo, a trial of triptans
meets diagnostic criteria for migraine. may be fruitful, even in patients without migrainous
It should also be noted that vertigo is listed as one headache accompanying the episodes of vertigo. Seda-
of the brainstem symptoms characterizing “migraine tive medications such as meclizine or benzodiazepines,
with brainstem aura,” formerly known as basilar often used in the treatment of vertigo of peripheral
or basilar-type migraine. These criteria are listed in origin, are another option. Lastly, physical therapy or
Table 2.9. The classification notes that while there vestibular therapy might be considered.
is some overlap between the two syndromes, most
patients with vestibular migraine experience vertigo
for longer than an hour. They also rarely experience Discussion
vertigo in the same relationship to headache that Vertigo is the sensation of rotational motion when
is seen with typical migraine with aura, where the the head and body are actually at rest. Patients often
aura immediately precedes and is usually followed by describe the feeling that they or the room are spinning;
headache. Instead, headache may not always occur this sensation of movement distinguishes “vertigo”
in relation to vertiginous symptoms, and may coex- from “dizziness.” Vertigo may occur spontaneously or
ist with rather than follow them. In addition, migraine may be triggered, often by a change in head position
with brainstem aura requires typical visual aura symp- but also at times by visual triggers. Any of these types

Chapter 2: Conditions that mimic primary headache

of vertigo may be accompanied by migrainous symp- Tip

toms, whether headache or sensitivity to other sen-
A diagnosis of vestibular migraine may be considered
sory modalities, to suggest a diagnosis of vestibular
for patients with a history of migraine when the ver-
tigo is associated with other migrainous symptoms and
There is a wide variation in the duration of vertigi-
other etiologies have been ruled out.
nous symptoms, from minutes to days. The pathophys-
iology of vestibular migraine is not well understood,
but an overlap between vestibular input and circuits Further reading
involved in migraine seems likely. Both clinical and Sinus disease
research interest in the topic has increase significantly Eross E, Dodick D, Eross M. The Sinus, Allergy and
over the last several years. The ICHD-3 beta notes Migraine Study (SAMS). Headache. 2007;47(2):213–24.
that the temporal course of vestibular migraine rarely
Schreiber CP, Hutchinson S, Webster CJ, et al. Prevalence
fits the pattern for typical aura (duration 5–60 min- of migraine in patients with a history of self-reported or
utes, immediately preceding headache) and therefore physician-diagnosed “sinus” headache. Arch Intern Med.
“episodes of vestibular migraine cannot be regarded as 2004;164(16):1769–72.
migraine auras.”
Although the ICHD-3 beta has included these Lyme disease
diagnostic criteria in order to stimulate further Halperin JJ, Shapiro ED, Logigian E, et al. Practice
research aimed at validating the disorder, it is worth parameter: treatment of nervous system Lyme disease
noting that other criteria for the diagnosis exist. (an evidence-based review): report of the Quality
Otolaryngologists may use the Neuhauser criteria. Standards Subcommittee of the American Academy of
A diagnosis of definite migrainous vertigo by these Neurology. Neurology. 2007;69(1):91–102.
criteria requires: (1) recurrent vestibular vertigo; (2) Nord JA, Karter D. Lyme disease complicated by
migraine according to the International Headache pseudotumor cerebri. Clin Infect Dis. 2003;37(2):e25–6.
Society (IHS); (3) migrainous symptoms during at
least two vertiginous attacks (migrainous headache, TMD
photophobia, phonophobia, or aura symptoms); and LeResche L. Epidemiology of temporomandibular
(4) vertigo not attributed to another disorder. Prob- disorders: implications for the investigation of etiologic
factors. Crit Rev Oral Biol Med. 1997;8(3):291–305.
able migrainous vertigo requires items 1 and 4 and
another migrainous feature (migraine per IHS crite- Schiffman E. Diagnostic criteria for headache attributed to
ria, migrainous symptoms during vertigo, migraine temporomandibular disorders. Cephalalgia. 2012;32(9):
triggers for vertigo, response to anti-migraine
Postpartum headache
Chronification of vertiginous symptoms is not as
well described in the literature, but anecdotally we Cardona L, Klein AM. Early postpartum headache: case
discussions. Semin Neurol. 2011;31(4):385–91.
have occasionally seen patients evolve from episodic to
chronic vertigo, seemingly in a process like that which Klein AM, Loder E. Postpartum headache. Int J Obstet
Anesth. 2010;19(4):422–30.
occurs with headache. The symptoms may become less
severe but more frequent or continuous over time,
Chiari malformation
a phenomenon commonly seen with other migraine-
associated symptoms such as photophobia. When the Grazzi L, Andrasik F. Headaches and Arnold-Chiari
syndrome: when to suspect and how to investigate. Curr
vertigo is not accompanied by headache or is not tem-
Pain Headache Rep. 2012;16(4):350–3.
porally associated with headache it is often unclear
Massimi L, Peppucci E, Peraio S, Di Rocco C. History of
whether the vertigo is a migrainous symptom. In those
Chiari type I malformation. Neurol Sci. 2011; 32(Suppl
cases we often empirically treat patients as if they have 3):S263–5.
chronic migraine.
Mea E, Chiapparini L, Leone M, et al. Chronic daily
headache in the adults: differential diagnosis between
symptomatic Chiari I malformation and spontaneous
Diagnosis intracranial hypotension. Neurol Sci. 2011;32(Suppl
Vestibular migraine. 3):S291–4.

Chapter 2: Conditions that mimic primary headache

Pascual J, González-Mandly A, Martı́n R, Oterino A. Kurth T, Diener H-C, Buring JE. Migraine and
Headaches precipitated by cough, prolonged exercise or cardiovascular disease in women and the role of aspirin:
sexual activity: a prospective etiological and clinical subgroup analyses in the Women’s Health Study.
study. J Headache Pain. 2008;9(5):259–66. Cephalalgia. 2011;31(10):1106–15.
Taylor FR, Larkins MV. Headache and Chiari I Wijman CA, Wolf PA, Kase CS, Kelly-Hayes M, Beiser AS.
malformation: clinical presentation, diagnosis, and Migrainous visual accompaniments are not rare in late
controversies in management. Curr Pain Headache Rep. life: the Framingham Study. Stroke. 1998;29(8):1539–
2002;6(4):331–7. 43.
Young WB. A knockout punch: C. Miller Fisher’s migraine
Cluster presenting as a dental problem accompaniments. Headache. 2000;48:726–7.
Bahra A, Goadsby PJ. Diagnostic delays and
mis-management in cluster headache. Acta Neurol Repetitive and prolonged migraine aura
Scand. 2004;109(3):175–9. Chen WT, Fuh JL, Lu SR, Wang SJ. Persistent migrainous
Balasubramaniam R, Klasser GD. Trigeminal autonomic visual phenomena might be responsive to lamotrigine.
cephalalgias. Part 1: cluster headache. Oral Surg Oral Headache. 2001;41:823–5.
Med Oral Pathol Oral Radiol Endod. 2007;104(3): Haan J, Sluis P, Sluis LH, Ferrari MD. Acetazolamide
345–58. treatment for migraine aura status. Neurology.
Bittar G, Graff-Radford SB. A retrospective study of 2000;55:1588–9.
patients with cluster headaches. Oral Surg Oral Med Oral e25968/e26078/e26305/index en.html.
Pathol. 1992;73(5):519–25. Haas DC. Prolonged migraine aura status. Ann Neurol.
Gross SG. Dental presentations of cluster headaches. Curr 1982;11:197–9.
Pain Headache Rep. 2006;10(2):126–9.
Rozen TD, Fishman RS. Female cluster headache in the Vestibular migraine
United States of America: what are the gender Bisdorff AR. Management of vestibular migraine. Ther Adv
differences? Results from the United States Cluster Neurol Disord. 2011;4(3):183–91.
Headache Survey. J Neurol Sci. 2012;317(1–2):17–28. Furman JM, Marcus DA, Balaban CD. Vestibular migraine:
Shoji Y. Cluster headache following dental treatment: a case clinical aspects and pathophysiology. Lancet Neurol.
report. J Oral Sci. 2011;53(1):125–7. 2013;12(7):706–15.
Lempert T, Neuhauser H, Daroff R. Vertigo as a symptom
Late-life migraine accompaniments of migraine. Ann NY Acad Sci. 2009;1164:242–51.
Fisher CM. Late-life migraine accompaniments as a cause Neuhauser H, Leopold M, von Brevern M, Arnold G,
of unexplained transient ischemic attacks. Can J Neurol Lempert D. The interrelations of migraine, vertigo, and
Sci. 1980;7:9–17. migrainous vertigo. Neurology. 2001;56:436–41 .

Missing dangerous causes of headache

In most patients with dangerous underlying causes of New headache in an elderly woman
headache there are signs and symptoms that clearly
point to the causal disorder. Typical features that sug- Case
gest a dangerous or secondary cause of headache
include fever, altered consciousness, or abnormal neu- A 75-year-old woman with no personal or family his-
rologic findings. One example of a rarely missed diag- tory of prior headache was referred to the headache
nosis is temporal arteritis presenting as a classic case center for consultation. She reported headaches over
of new-onset headache in an elderly person, accompa- the last two months. The headaches began gradually
nied by anemia, jaw claudication, and visible tortuosity but increased rapidly to the point where they were
of the temporal artery. daily and constant. They were located over the top of
When such symptoms or signs are subtle or absent, her head and radiated to her forehead and face. The
however, dangerous causes of headache can be missed pain was variable: sometimes dull and sometimes a
or mistakenly attributed to other disorders. Headache pressure sensation “like my head is going to fly off.”
is also a common symptom of many illnesses. Patients The pain had been especially intense over the last two
may have more than one condition known to be asso- weeks. It ranged from 4 to 6 on a 0–10 pain scale. When
ciated with headache, making it difficult to assess asked, the patient said the headaches were associated
causation. with photo and phonophobia but she did not have nau-
Finally, a number of disorders can produce sea, vomiting, or worsening of headaches with physical
headaches that superficially resemble nondangerous activity or change of position.
causes of headache such as migraine or tension-type The patient denied jaw claudication, visual dis-
headache. Both of these headache types are com- turbances, or associated autonomic symptoms. She
mon, so many patients presenting for evaluation of reported that her sinuses felt “congested” and the
headache will have a prior history consistent with headache temporarily improved when she sneezed.
one of these benign headache syndromes. Migraine, She was treating her headaches with over-the-counter
for example, is a highly prevalent cause of severe acetaminophen plus phenylephrine. She had a 15
headache, with a cumulative lifetime incidence of 44% pound weight loss over the last year and reported
in women and 18% in men. that she has been told she is anemic. She was recently
A prior history of migraine or tension-type seen in the emergency department because of a severe
headache in a patient with a new or different form headache, where a neurologic examination was nor-
of headache is not necessarily reassuring, however. In mal and a CT scan of the head did not reveal any
fact, patients with a well-established history of benign explanation for her headaches. At that visit, her systolic
headaches may be at especially high risk of having a blood pressure ranged from 160 to 180 mm Hg. A diag-
dangerous cause of headache mistakenly attributed to nosis of headache due to poorly controlled hyperten-
their prior nondangerous headache disorder. sion was made and her blood pressure medication was
Cases in this chapter illustrate situations in which adjusted.
underlying medical or structural abnormalities A sinus CT was obtained on an outpatient basis
responsible for headache were not recognized, and when headaches did not improve despite better con-
headache was mistakenly attributed to a nondanger- trol of her blood pressure. The report she brought with
ous headache or other disorder. her read as follows:

Chapter 3: Missing dangerous causes of headache

IMPRESSION: toms, a cognitive error known as premature closure.

1. Multifocal sinus disease, involving the bilateral In one series of diagnostic errors, premature closure –
anterior/posterior ethmoid air cells, sphenoid failure to consider other diagnostic possibilities once a
sinus, and left frontal sinus. This could account for diagnosis had been identified that “fit the facts” – was
the patient’s headache. the most common cause of missed diagnoses.
2. No acute intracranial abnormality. In this case the initial diagnoses only partially “fit
the facts.” Although the patient’s blood pressure was
She was started last week on amoxicillin/clavulanic indeed high in the emergency department, there were
acid twice daily but her headaches continued. no other signs or symptoms consistent with a diag-
nosis of hypertensive encephalopathy. This makes it
unlikely that her elevated blood pressure had occurred
What are the leading diagnostic possibilities abruptly, which is the only setting in which hyperten-
and how would you proceed in evaluating sion is well validated as a cause of headache. Chronic
them? hypertension is not commonly considered a cause of
headache. Additionally, this patient’s headaches did
New-onset headache in an elderly person is a “red flag”
not improve even with better control of her blood
situation that should prompt a careful search for an
underlying causative condition. While hypertension
The CT scan shows extensive sinus disease, which
or sinusitis may seem like plausible explanations for
might account for the patient’s headaches – indeed, the
her problem, they do not entirely explain other worri-
radiology report suggests so – but antibiotic treatment
some features such as weight loss or anemia. Giant cell
did not produce improvement in her symptoms. Some
arteritis is an additional possibility that must be seri-
sinus infections respond poorly to antibiotics, how-
ously considered in any patient over the age of 50 who
ever, so sinusitis may in fact be contributing to this
presents with a new headache.
patient’s symptoms, but it is far from being the only
After consultation at the headache center, this
possible explanation. Diagnostic delay is not uncom-
patient had blood work and additional imaging per-
mon in giant cell arteritis. Predictably, delayed diag-
formed. Results included a hematocrit of 30, an ery-
nosis is more common in patients like this one who do
throcyte sedimentation rate (ESR) of 94 (normal 1–
not present with “typical” features such as jaw claudi-
20), and a C-reactive protein (CRP) of 34.6 (normal
cation or visual disturbance.
0–5). A magnetic resonance angiogram showed nor-
Visual loss is the most feared complication of giant
mal cerebral vessels and extensive mucosal thicken-
cell arteritis. Visual symptoms in combination with
ing of the sphenoid sinus and ethmoid air cells with
new-onset headaches would prompt most clinicians
moderate mucosal thickening of the frontal sinus. The
to consider a diagnosis of giant cell arteritis. How-
patient was continued on amoxicillin and clavulanic
ever, visual problems are found in only a quarter to a
acid twice daily and 60 mg of prednisone was added
third of patients at presentation. This patient did have
each morning; a temporal artery biopsy was also per-
headache, which is the most common initial symptom,
formed. The patient’s headaches resolved within two
found in almost 90% of patients. Unfortunately, this
days and the temporal artery biopsy was positive for
symptom alone was not enough to prompt consider-
giant cell arteritis.
ation of giant cell arteritis in the presence of attractive
competing explanations.
Why was the correct diagnosis delayed?
Somewhat surprisingly, this elderly patient had been
seen by her primary care physician and in the emer- Discussion
gency department with complaints of new headache The gold standard for diagnosis of giant cell arteritis
but had not had either an ESR or CRP checked. is temporal artery biopsy. Most patients will also have
Instead, her headaches were attributed to elevated elevated inflammatory markers. CRP may be more
blood pressure and later a possible sinus infection. sensitive than the ESR, but either can be normal in
In hindsight, it seems clear that after these initial the occasional patient with documented disease. Thus,
diagnoses were considered, the clinicians involved did most clinicians order both tests to improve diagnostic
not reflect on alternative causes for the patient’s symp- sensitivity.

Chapter 3: Missing dangerous causes of headache

The clinical presentation of giant cell arteritis is was treated with intravenous promethazine, with some
highly variable, and this can make diagnosis difficult. symptomatic improvement.
As was done in this case, steroid treatment should
be initiated immediately if clinical suspicion of giant Does this sound like migraine?
cell arteritis is high. Most experts agree that a few
While some features of the patient’s headaches are con-
days of steroid treatment will not reduce the likelihood
sistent with a diagnosis of migraine, other things, such
of a positive biopsy result. A normal temporal artery
as the fever and chills, are not. The headaches are of
biopsy does not rule out a diagnosis of giant cell arteri-
new onset and his pain is side-locked. All of these
tis; false-negative biopsies can occur if unaffected por-
things suggest that further evaluation is warranted. In
tions of the artery (so-called “skip lesions”) are sam-
this case, a CT scan of the brain was ordered by the
pled in the biopsy. In this patient’s case, a high clinical
emergency department and performed as soon as the
index of suspicion for giant cell arteritis was present
patient’s pain had improved with treatment. The films
and the clinicians involved agreed that they would
showed an air/fluid level in the left sphenoid sinus.
proceed with steroid treatment even if the biopsy
Laboratory tests done in the emergency department
were normal. She may well have coexistent sinus dis-
showed an elevated CRP of 12 and an ESR of 32.
ease, too, and her clinicians planned to consult an
In view of his new-onset headaches and his-
tory of fever and elevated inflammatory markers, the
patient also has a lumbar puncture and is sent for
Diagnosis blood cultures, and oral antibiotics are started. The
Giant cell arteritis. cerebrospinal fluid (CSF) is clear, the opening pres-
sure normal, and there is no pathogen growth at 48
Tip hours. Blood cultures, however, grow Haemophilus
Many elderly patients have a large number of medical
conditions that might plausibly cause or contribute to
new-onset headaches. It is important to consider fully What is the next step in management?
all possible explanations for headache rather than pre- The positive blood cultures confirmed septicemia. The
maturely settling on the first diagnosis that appears to source of the infection was most likely the left sphe-
fit the facts of the case. noid sinus, given the patient’s prominent eye, head,
and face pain and sinus abnormalities on CT scan.
A young man with head pain, nausea, A careful search for sources of head and neck infec-
tion should start with a detailed ear, nose, and throat
and fever examination. In this case, an otolaryngologist noted
pus in the left superior turbinate and confirmed the
Case clinical impression of acute sphenoid sinusitis. The
A healthy 22-year-old man was seen in the emer- patient was treated with high-dose antibiotics and sur-
gency department for new-onset headaches. He was gical debridement of the sphenoid sinus and recovered
well until two weeks ago, when he began to experience uneventfully.
intermittent severe left eye pain and head and face pain Acute sphenoid sinusitis is not a common con-
associated with fever, chills, nausea, and vomiting. The dition. It can be difficult to diagnose since typical
pain worsened with coughing, sneezing, or physical signs and symptoms of sinusitis often are not present.
exertion. The headache was located over the vertex For example, nasal congestion, discharge, or postnasal
and occiput and the face pain in the V1 distribution. drips are not common in sphenoid sinusitis, which
There was a history of an upper respiratory infection instead tends to present with headache and fever. The
a week before onset of the headaches, but at the time pain may be felt over the vertex, occiput, or in the peri-
of presentation the patient had no nasal or respira- orbital region, and nausea is common. Pain is often
tory symptoms. Physical and neurologic examination worsened by coughing or physical exertion. It is easy
revealed no abnormalities other than a slightly ele- to see how such a presentation could be mistaken for
vated temperature of 101 degrees Fahrenheit. A diag- migraine, as it was in this case. Fortunately, however,
nosis of new-onset migraine was made and the patient diagnostic workup for secondary causes of headache

Chapter 3: Missing dangerous causes of headache

Table 3.1. Structures adjacent to the sphenoid sinus that may second division of the trigeminal nerve, so mid-facial
be affected by sphenoid disease
pain is more common with maxillary sinusitis. The
Cranial nerve 2 frontal sinuses are innervated by the supraorbital
Cranial nerve 3 and supratrochlear nerves of the first division of the
Cranial nerve 4 trigeminal nerve.
First division of cranial nerve 5
Second division of cranial nerve 5 Discussion
Cranial nerve 6 Some forms of sinusitis can be serious. This is espe-
cially true of sphenoid sinusitis. Untreated, infection
Dura mater
can spread to involve the cavernous sinus or cause
meningitis. Headache is the most common present-
Cavernous sinus ing symptom, and several case series suggest that the
Internal carotid artery headache often awakens patients from sleep. Sphenoid
Sphenopalatine ganglion sinusitis is also on the list of disorders that can produce
Sphenopalatine artery sudden onset of extremely severe headache – so-called
“thunderclap” headache – perhaps as a result of exten-
Pterygoid canal and nerve
sion of infection through the wall of the sinus.
The location of the headache is variable: it is often
was pursued even after the preliminary diagnosis of over the vertex in addition to causing pain in the face or
migraine had been made. over the sinus region. Sensory changes in V1 or V2 or
Acute sphenoid sinusitis can be very serious, pus in the middle or superior turbinates on examina-
so aggressive treatment is warranted. The sphenoid tion can help make the diagnosis, but these findings are
sinuses lie within the sphenoid bone in the mid- not always present. X-rays or CT may show opacifica-
dle cranial fossa, adjacent to many critical structures tion of the sinus or thickening of the mucosa, but there
(Table 3.1). These include the pituitary and the cav- is no clear correlation between the extent of opacifica-
ernous sinus. If infection spreads to the cavernous tion and the severity of illness. CT is probably some-
sinus or subarachnoid space, life-threatening com- what better than MRI for the diagnosis of sinusitis, but
plications can develop. The most common causative either is probably sufficient for initial evaluation. In a
organisms are bacterial, especially staphylococcal or review of 300 CT and MRI scans, the sphenoid sinus
streptococcal infections, but fungal infections or was visualized in all of the studies.
Gram-negative infections can also occur.
What is the nociceptive innervation of Acute sphenoid sinusitis.

the paranasal sinuses? Tip

Sensation from the nasal mucosa and sinuses is trans-
Acute sphenoid sinusitis can be missed when it mim-
mitted through the first and second divisions of the
ics migraine by producing peri-orbital pain and nau-
trigeminal nerve with a very small contribution from
sea. It should be considered in any case of new-onset
the greater superficial petrosal branch of the facial
headache, particularly in the presence of fever or signs
nerve. In particular, the sphenoid and ethmoid sinuses
of inflammation with pain over the vertex of the head.
are served by the posterior ethmoidal branch of the
nasociliary nerve, which arises from the first divi-
sion of the trigeminal nerve. This may account for Weekend headache
the prominent occurrence of orbital and peri-orbital
pain in cases of sphenoid sinusitis, since pain from Case
the sinuses is usually referred to the corresponding A 23-year-old woman had been seen multiple times
dermatome. over the last six months for complaints of headache.
In contrast, the maxillary sinus is innervated by She had no prior history of headache although an
the superior alveolar and infraorbital branches of the aunt had migraine. She described the headaches as

Chapter 3: Missing dangerous causes of headache

constant. When headaches began they mostly oc- Upon further questioning the patient confirmed
curred on weekends or holidays but quickly became that she had no prior history of headache. She worked
daily. Typically they were worst in the mornings and as a technical editor and was able to do much of
continued to be more severe on weekends. The pain her work at home. As her headaches have worsened
was described as throbbing, and was rated on average she had been spending more and more time at home
6 on a scale of 0–10. and remarked that “I can’t remember the last time I
Headaches were associated with dizziness, nau- was in my office . . . the worse I get the more time I
sea, occasional vomiting, and, on two occasions, spend at home and the more time I spend at home the
loss of consciousness. With both episodes of loss of worse I seem to get.” When asked about relatives with
consciousness she was taken to the emergency depart- headache problems, the patient again mentioned her
ment. Testing included an MRI scan of the head, elec- aunt. She added that she shared a house with her sis-
trocardiogram, Holter monitor, and electroencephalo- ter, who worked as a flight attendant. Her sister had
gram, all of which were normal with the exception of recently also developed low-grade headache and nau-
occasional episodes of supraventricular tachycardia. sea, but her headaches did not seem to be as severe.
On both occasions she slowly improved during
a short hospitalization for observation and was dis- Does the additional history change your
charged with a diagnosis of chronic migraine. Treat-
ment aimed at migraine had not been helpful, however. mind about possible diagnoses?
The patient reported modest, temporary relief from The patient’s offhand comment about getting worse as
anti-inflammatory medications but none at all from she spent more time at home was a clue to her diag-
triptans or a variety of preventive medications such as nosis, as was the history of a similar illness in some-
propranolol and topiramate. one who lived with her. Headaches were initially most
noticeable on weekends, when she was most often at
home. As her illness progressed, she began to spend
Based on this patient’s clinical presentation, more time at home, at which point the headaches wors-
ened and became daily.
what are some diagnostic considerations? All of these facts are consistent with a toxic envi-
The patient’s age and sex, the throbbing character of ronmental exposure in the home as a possible cause
the headache, its duration and frequency, as well as of headache. When her physicians learned that the
associated features of nausea and vomiting all fit with patient’s headaches began last October with the onset
a diagnosis of chronic migraine. It is also the case that of cold weather they became suspicious about possible
many people with migraine report they are more likely carbon monoxide exposure. On testing, the patient’s
to have headaches on weekends, holidays, vacations, or carboxyhemoglobin level was 28% and her sister’s was
at other times of “letdown” from stress. 18%. Inspection of their home showed a faulty furnace.
Other features of this patient’s headache, however, Carbon monoxide poisoning can be very diffi-
did not fit with migraine: migraine is not typically cult to diagnose when it results from chronic expo-
associated with loss of consciousness, and it is rare sure to carbon monoxide from defective space heaters,
for it to evolve so quickly to the chronic form of the kitchen stoves, or furnaces. The symptoms are non-
disorder. Could this be new daily persistent headache specific and easily attributed to other disorders such
which, as its name implies, presents as a constant as migraine. In this case the patient’s age, sex, and
headache? Perhaps, but the episodes of loss of con- headache features, including the weekend occur-
sciousness still cannot be explained by that diagnosis. rence of the headaches, all fit with a diagnosis of
In this case, the rapid evolution of the headache migraine.
after onset, the unexplained episodes of loss of con- Because migraine is so common in young women,
sciousness, and the lack of response to extensive tri- it is a diagnosis that most physicians can easily call to
als of treatment aimed at migraine suggested that mind. This is an example of a cognitive bias known
alternative diagnoses should be considered. In cases as “availability bias,” in which the differential diag-
of difficult-to-diagnose headache, most experts agree nosis is strongly influenced by what is easily recalled.
that the crucial next step is “history, history, and more Here it was easy to assume that the weekend headaches
history.” reported by the patient were the typical “letdown”

Chapter 3: Missing dangerous causes of headache

headaches of migraine, and not so easy to remember Tip

that at a more fundamental level they were related to
Weekend headaches are a hallmark of migraine, but
absence from work – and exposure to the home envi-
not all weekend headaches are migraine. Occult car-
ronment – factors that might have reminded her doc-
bon monoxide poisoning can produce headache that is
tors about environmental exposures that can produce
difficult to distinguish from migraine. In patients with
refractory headaches occurring in winter and with
similarly affected cohabitants, a high index of suspi-
Discussion cion for carbon monoxide poisoning is appropriate.
Carbon monoxide is odorless, so victims are not aware
of exposure. It binds more tightly to hemoglobin than A young woman with seizure
oxygen, with the result that oxygen delivery to the
brain and other tissues is impaired. Diagnosis is based
and headache
on finding a venous carboxyhemoglobin level above
Headache is the most common symptom of car- A 19-year-old woman with epilepsy was referred to
bon monoxide poisoning and is thought to result the headache clinic for management of very severe
from cerebral vasodilation in response to hypoxemia. headaches occurring after seizures. She was diagnosed
Nausea and a general sense of malaise or other flu- with complex partial seizures two years ago and was
like symptoms are also common. The higher the car- undergoing treatment trials with antiepileptic drugs.
boxyhemoglobin level and the more sustained the Her seizures were less frequent with topiramate but
exposure, the worse the long-term outcome. Chronic still occurred about once every other month.
encephalopathy can result with prolonged or very high Following her seizures she experienced unilateral,
level exposure. throbbing, severe headache with photophobia, nausea,
Carbon monoxide poisoning is treated by removal and vomiting. She rated the headaches as “10+” on
from the exposure and with oxygen therapy; 100% a 0–10 pain scale and reported they could last up to
oxygen is commonly administered for several hours. a whole day. The pain was not relieved by ibuprofen.
If carboxyhemoglobin levels are above 25%, guide- Because she and her epileptologist were primarily con-
lines recommend the use of hyperbaric oxygen at cerned about controlling seizures, the headaches had
2–3 atmospheres, if available. Hyperbaric oxygen not been a focus of attention until recently.
rapidly increases the amount of oxygen that is dis- Upon further questioning she admitted having
solved in plasma and the half-life of carboxyhe- occasional milder but similar headaches outside of
moglobin is shortened from 4–5 hours in room air to the seizure episodes, often around her menstrual
about 20–25 minutes. There can be complications to period. Her mother had “sick headaches” when she was
hyperbaric treatment, including tension pneumotho- younger.
rax or cerebral edema. It is contraindicated in patients
with chronic obstructive pulmonary disease. What is the relationship between seizures
One study found that roughly 15% of patients with
headache who presented during the winter months to and headaches?
an urban emergency department had elevated carbon The International Classification of Headache Disor-
monoxide levels. Another study found that smoking a ders (ICHD) recognizes two forms of headache asso-
high number of cigarettes daily, use of a stove to heat ciated with epileptic seizures. The first, hemicrania
living spaces, and similar illness in a family member epileptica, is diagnosed when headache is a primary
were predictive of carbon monoxide poisoning, under- manifestation of the seizure itself. The diagnostic cri-
scoring the value of a good environmental exposure teria include headache lasting seconds to minutes,
history. with features of migraine, developing synchronously
with a partial epileptic seizure and resolving imme-
diately after the seizure. The pain must be ipsilateral
Diagnosis to the ictal discharge. This is a relatively uncommon
Carbon monoxide intoxication. phenomenon.

Chapter 3: Missing dangerous causes of headache

Much more common is post-ictal headache, which episodes with antiepileptic therapies. If headaches also
is diagnosed when a headache occurs after a seizure. occur outside of seizure activity, an antiepileptic drug
The criteria include a headache with features of with evidence of efficacy for headache may be a good
tension-type or migraine headache, developing within option. Antiepileptic drugs that may be effective for
three hours after a partial or generalized epileptic both seizures and migraine include topiramate and
seizure and resolving within 72 hours after the seizure. valproate. Evidence is less certain for gabapentin.
Prevalence estimates for post-ictal headache range There is little evidence to guide the choice of
from 25% to 50%. symptomatic treatments for post-ictal headache. Case
Migralepsy, a disorder in which migraine is the reports and clinic experience both support the prac-
putative trigger for an epileptic seizure, has been tice of using migraine-specific therapies to treat
very rarely described and is a controversial diagno- patients whose post-ictal headaches have migrainous
sis. Migraine and epilepsy share many features. Both features. Anecdotally, some patients who have both
are chronic neurologic disorders of long duration with migraine and post-ictal headaches note that their post-
episodic manifestations and exacerbations. Both dis- ictal headaches are more severe than their regular
orders are characterized by paroxysmal “all or none” migraines.
events. If triptans or nonsteroidal anti-inflammatory
While the exact pathophysiology of migraine with- drugs are not effective for post-ictal headache, this
out aura is unclear, it is generally accepted that may be the rare situation in which the use of butalbital-
migraine aura is associated with excessive cortical containing drugs or opioid drugs is defensible. As
excitation, as is seizure activity. In migraine, cortical with treatment of regular migraines, antiemetic treat-
hyperexcitability is theorized to lower the threshold ment should be considered if nausea is a prominent
for cortical spreading depression, in which an orderly accompaniment of the headache.
wave of neuronal depolarization travels across the sur- The patient in this case was treated with subcu-
face of the cortex. This is followed by a period of taneous sumatriptan and oral prochlorperazine. This
depressed cortical activity as the neurons struggle to regimen was effective for most of her interictal and
repolarize. Cortical spreading depression is generally post-ictal headaches. At a follow-up visit she reported
accepted as the cause of migraine aura. having a few very severe headaches for which this regi-
A number of similar molecular mechanisms are men was not effective and she was given a prescription
theorized to play a role in at least some cases of both for a butalbital-containing medication to use in those
disorders, such as abnormalities in ion channel func- circumstances.
tion leading to increased glutamatergic tone. The pos-
sibility that migraine and epilepsy might share some
underlying mechanisms is supported by epidemiologic Discussion
evidence that shows the two disorders are comorbid. Although epilepsy and headache may be comorbid
The prevalence of epilepsy is higher in migraineurs (1– conditions, they can usually be distinguished on clini-
71%) than in the general population (0.5–1%). Esti- cal grounds alone. Because of this, electroencephalog-
mates of the prevalence of migraine in the epilepsy raphy (EEG) is not useful in the initial evaluation of
population range from 8.4% to 24%. patients with typical headaches. This is reflected in a
In this case, the patient has an established diagnosis practice parameter of the Academy of Neurology. EEG
of partial complex seizures, and the temporal relation- may be warranted in patients who have atypical visual
ship between the seizure and the resulting headaches aura where conditions such as occipital lobe epilepsy
is clear. A diagnosis of post-ictal headache is there- are being considered.
fore appropriate. Based on her history of other similar Headache as the sole manifestation of seizure is
headaches occurring in the absence of seizure, it seems very rare, while headache occurring after a seizure
likely that she has a diagnosis of migraine as well. with other manifestations is more typical. Post-ictal
headache is relatively common, affecting up to half of
patients with epilepsy. It may be more common in chil-
What treatment options are available? dren. Despite being common, post-ictal headache is
In cases where headaches occur only after seizures, often overlooked. This may be because the headaches
the most effective treatment is prevention of seizure are not very severe, or may be because the seizure is

Chapter 3: Missing dangerous causes of headache

more concerning and “overshadows” the subsequent Intermittent acute angle-closure glaucoma is
headache. Because these headaches may be bother- another possibility, but it does not typically last for
some and refractory to minor analgesic treatments, two weeks: the duration of a typical episode of angle-
however, they often warrant medical attention. closure glaucoma is only 30 to 60 minutes. It is usually
associated with other symptoms such as halos around
lights. This patient has reported only blurry vision.
Diagnosis Additionally, this patient does not report exposure
Post-ictal headache; episodic migraine. to provocative agents that are associated with the
development of acute angle-closure glaucoma, such as
Tip a dark or dim environment, prolonged near work, or
Post-ictal headache is common and can be severe. sneezing.
Failure to ask patients with epilepsy about post- Proptosis is not a feature of either cluster headache
ictal headache, and/or failure to recognize migrainous or glaucoma. A mass lesion of the orbit or retro-orbital
symptoms, may delay treatment. area is a possibility but it is unlikely that would produce
intermittent proptosis and redness. Cavernous sinus
thrombosis could produce these symptoms, but it does
Intermittent red eye and headache not usually remit and recur.

A 44-year-old woman reported intermittent redness What is the next step in evaluation of
and bulging of her right eye in association with this patient?
headache. For five years, the episodes had occurred The intermittent bulging and congestion of the eye
at least once a year and lasted for up to two weeks. suggest the possibility of a vascular lesion and the
During the episodes she had severe, 10/10 ipsilateral need for further imaging of the cerebral vessels. On
headache without migrainous features. Toward the end examination during an episode, this patient had right
of the episodes her vision became blurry. She had been proptosis, ptosis, and slightly decreased right visual
treated with antibiotic and steroid eye drops without acuity as well as dilated episcleral vessels. The optic
benefit. Recently her symptoms had lasted longer than discs were normal but the right retina showed some
usual and while she was visiting a family member in venous engorgement. A carotid cavernous fistula was
the hospital a casual physician observer suggested she suspected after the patient mentioned that she could
should seek emergency evaluation. “hear my heartbeat” when the eye was bulging.
A magnetic resonance angiogram did not show
What are the possible causes of intermittent a fistula but the right cavernous sinus was enlarged.
Conventional cerebral angiography confirmed a right
red eye and headache? carotid cavernous fistula.
Redness of the eye can occur because of autonomic
dysfunction, venous congestion, or an inflammatory
condition. In this case, an inflammatory condition Discussion
such as orbital cellulitis seems unlikely given the inter- Carotid cavernous fistulas can cause eye pain, visual
mittent and unilateral nature of the symptoms and lack loss, and proptosis, probably due to marked orbital
of association with other symptoms such as fever. Cel- and cerebral venous congestion. In this case, the diag-
lulitis is also unlikely to resolve spontanteously. nosis of a carotid cavernous fistula was suggested by
Cluster headache can present with a unilateral the recurrent and remitting nature of the episodes in
red eye and ipsilateral pain, but by definition the combination with suspicious findings on examination,
individual attacks cannot last for longer than three particularly the ocular bruit that was audible to the
hours. Some patients with chronic cluster headache patient.
do develop a low-grade constant background discom- In carotid cavernous fistulas there is abnormal
fort in the orbital and temporal area, but the severe communication between the carotid artery and the
nature of the pain in this case is not consistent with that cavernous sinus, usually through dural branches of the
possibility. internal or external carotid artery. These fistulas can

Chapter 3: Missing dangerous causes of headache

result from trauma, but the cause of non-traumatic fis- and management for the presumed benign condition
tulas is not known. is appropriate.
Elevated pressure in the cavernous sinus can pro- In this case, though, a diagnosis of sinus headache
duce substantial venous congestion in the orbit, with was made in the absence of any obvious symptoms of
resulting visual loss. While small fistulas may seal sinusitis. The patient is at an age where migraine com-
over, larger and more symptomatic lesions are gener- monly begins, but she does not have typical features of
ally treated with endovascular techniques to close the migraine such as nausea or sensitivity to light or noise.
connection. Despite initial improvement, her headache problem is
progressing. Neuroimaging thus seems appropriate to
Diagnosis evaluate the possibility of underlying explanations for
Carotid cavernous fistula.
A brain CT showed a lesion and early papilledema
was noted on funduscopic examination by the neurol-
Tip ogist. Brain MRI showed a 4 cm enhancing lesion in
A red, bulging eye accompanied by headache is not the left posterior fossa with significant mass effect on
consistent with a primary headache disorder and the cerebellum and the adjacent brainstem, and partial
requires further workup. compression of the fourth ventricle. The patient was
admitted for further evaluation and ultimately diag-
A young woman with morning nosed with a medulloblastoma.

This seemed like a benign headache.
A 24-year-old female reported new-onset bilateral,
What was missed?
frontal, and retro-orbital headache onset every morn- This was a new-onset headache that did not clearly
ing that resolved later in the day. There were no meet criteria for any primary headache. Although the
associated symptoms. There was no prior history of initial visit note indicated that there were no neu-
headache. Her neurologic examination was normal. rologic abnormalities, no funduscopic examination
She was diagnosed with sinus headache and advised to was documented in the record. Thus, it is not at all
try over-the-counter decongestants. clear that the neurologic examination was normal.
The headaches improved but then returned about Had a careful funduscopic examination been done,
three months later and were more intense. She papilledema might have been detected and the diag-
returned to her doctor, who wrote in the chart that nosis of medulloblastoma made sooner. It also seems
there were “no neurologic abnormalities.” A computed likely that the initial improvement in the headache led
tomographic scan of the head was ordered and she was to a false feeling of reassurance.
referred for neurologic evaluation.

Is it appropriate to order imaging in Discussion

Headache is present at some point in up to 50% of brain
this case? tumor patients, but it is rarely the presenting symptom.
Headache is common and usually benign, especially Nonetheless, doctors worry about cases like this one, in
when there are no associated abnormalities and the which a patient presenting with headache and no other
neurologic examination is normal. If such patients features turns out to have a deadly brain tumor.
meet criteria for one of the primary headache dis- The response to cases like this should not be to
orders such as migraine or tension-type headache, image indiscriminately all patients with new-onset
further evaluation is unlikely to be fruitful. In these headache. Even in patients with new-onset headache
circumstances the chance of finding a clinically mean- the prior probability of a dangerous cause of headache
ingful abnormality on neuroimaging has been esti- is low when there are no associated symptoms – this
mated at about 0.1%. Instead of imaging, reassurance particular case notwithstanding – and most imaging

Chapter 3: Missing dangerous causes of headache

findings will be “incidentalomas” that do not change of mild benefit. He was significantly limited in his daily
clinical care. activities.
Rather, this case underscores the importance of a
thorough neurologic examination, including a fundus- What are possible diagnoses for this
copic examination, in every patient with new-onset
headaches. It also illustrates the value of follow-up vis- headache presentation and what testing
its and clinical monitoring. Although the underlying would you consider?
tumor causing this patient’s headaches was missed at This patient has a history of a previous primary
the first visit, appropriate testing was done when she headache disorder, episodic migraine, and transforma-
returned to report that she was worse. tion from the episodic to the chronic form of the disor-
Careful monitoring of the clinical course of der might be one explanation for his daily headaches.
headaches is a suitable way to gather more informa- Both snoring and sleep-disordered breathing are risk
tion about a headache problem where the diagnosis is factors for progression of migraine.
not clear. This “watchful waiting” strategy avoids the Headache that is present every morning and more
harms and costs that come from over-testing while severe than at other times of day is commonly reported
minimizing the likelihood that an important problem in many severe chronic headache disorders. It does not
will be overlooked. clearly point to a particular diagnosis. However, noc-
turnal hypoxia could also produce morning headache
Diagnosis and can be seen in association with sleep apnea syn-
drome. This seems to fit with the report of apneic
Medulloblastoma. episodes during sleep.
An overnight polysomnogram in this patient
Tip showed moderate obstructive sleep apnea. He was
started on continuous positive airway pressure therapy
Omitting the funduscopic examination in the evalua-
(CPAP) and within one week his daily headaches had
tion of headache can lead to errors in diagnosis. A care-
resolved. He continued to have occasional intermittent
ful examination of the optic fundi is an essential part
migraine headaches.
of the physical examination of every patient presenting
with headaches.
Obstructive sleep apnea is often associated with
More morning headaches headache and is a risk factor for transformation of
migraine from episodic to chronic. All patients with
Case chronic headache or headache upon awakening should
A 58-year-old obese man with a history of episodic be questioned about symptoms of obstructive sleep
migraine since his teens presented for evaluation of apnea, such as frequent nighttime awakenings, wit-
subacute onset chronic headache about one year ago. nessed apneas by bed partners, unrefreshing sleep,
The headache was holocephalic, moderate to severe in daytime somnolence, and irritability (Table 3.2). Cer-
intensity, pounding, associated with mild photophobia tain risk factors for obstructive sleep apnea are also
and phonophobia, and worse with activity. There was easily assessed in a routine office visit, including ele-
no aura. He woke up with a headache every morning vated body mass index (BMI), retrognathia, and pres-
and sometimes awakened from sleep with a headache ence of a large neck circumference. Some patients do
as well. He described his sleep as broken, and said not have full apneas, but may have nocturnal hypoxic
he had been “a snorer” his whole life. His girlfriend episodes that could contribute to headache as well.
noticed periodic apneas over the last two years. After Obstructive sleep apnea and nocturnal hypoxic
nights when he didn’t sleep well, his headaches were episodes are evaluated with a polysomnogram, which
usually worse. A brain MRI was normal. Treatment includes continuous oxygen saturation monitoring.
with multiple abortive and preventive migraine medi- Referral to a sleep specialist or for a polysomno-
cations for a diagnosis of presumed migraine was only gram should be considered in any patient who reports

Chapter 3: Missing dangerous causes of headache

Table 3.2. Features associated with obstructive sleep apnea family history of migraine or other chronic headaches
Morning headaches the patient was thought to have migraine. A brain
MRI scan showed a small pituitary adenoma with-
out anatomic compression or extension. The patient
Daytime somnolence
reported no endocrine symptoms. The patient was
Irritability advised that the finding was very unlikely to be the
Unrefreshing sleep cause of his headache and additional trials of treatment
Obesity for migraine were attempted.
Large neck circumference Three years later the patient developed severe,
Poorly controlled hypertension
chronic headache and was unable to work. A repeat
MRI showed an increase in the size of the adenoma
although there was still no evidence of extension or
symptoms of obstructive sleep apnea, or has several compression. Careful questioning elicited symptoms
risk factors for obstructive sleep apnea in addition to consistent with early acromegaly: an increase in shoe
refractory headache. Preliminary studies suggest that and glove size and hypertrophy of the gums. After
most patients with headaches and obstructive sleep neurosurgical evaluation he was offered surgical
apnea have improvement of the headaches when the removal of the adenoma through a trans-sphenoidal
disorder is effectively treated. approach. One month after surgery he was almost
This patient had a known primary headache dis- headache-free and planning to return to work.
order, which might have predisposed him to hav-
ing headaches in response to another stimulus, in
this case obstructive sleep apnea. However, treatment Should the patient initially have been
with migraine therapy did not completely resolve the
headache. When he received treatment for obstruc- advised that the MRI finding was unrelated
tive sleep apnea he reverted to his previous pattern to the headache pattern?
of episodic migraine. This illustrates the importance Probably not, since headache is surprisingly common
of looking for other potential causes or aggravat- among patients with pituitary tumors. In fact, it is the
ing factors of headache in patients with known pri- presenting feature in 40–70% of cases. It does not seem
mary headache disorders whose headaches change in to be related to the size of the tumor. In one small
character. cohort of patients who had pituitary tumors and trou-
blesome headaches, almost half had improvement in
Diagnosis headaches following surgery, suggesting a causal rela-
Obstructive sleep apnea causing progression from tionship. However, 15% of patients reported worsening
episodic to chronic migraine. of headaches following surgery. Headache alone is gen-
erally not considered a sufficient indication for surgery
because of this uncertainty.
Obstructive sleep apnea can be a cause of difficult-to-
treat chronic headaches and headaches often respond
well to treatment of apnea.
What are the characteristics of headaches in
people with pituitary tumors?
A middle-aged man with side-locked One study evaluated the phenotypic characteristics of
84 patients with pituitary adenomas. Headache was
headache side-locked in 88% of cases, and over half described the
pain as throbbing. In fact, the majority had headaches
Case with many of the features of migraine. The presence of
A 35-year-old man had a five-year history of episodic migrainous features was associated with a family his-
left-sided headache that was side-locked (i.e. did not tory of migraine. Perhaps people with a genetic pre-
shift sides). The headache had been severe at times disposition to migraine are also more likely to develop
and was difficult to control. Although there was no headache in the presence of a pituitary tumor.

Chapter 3: Missing dangerous causes of headache

Thus, in contrast to the general rule that headaches Fatal migraine

due to brain tumors have features of tension-type
headache, headache associated with pituitary lesions Case
more often fits criteria for episodic or chronic
A 53-year-old man with a long history of migraine
with visual and dysphasic aura died following a two-
Other headache patterns described with pitu-
month attack of constant, severe left-sided headache.
itary lesions include cluster headache, SUNCT (Short-
The headache was associated with a persistent visual
lasting, Unilateral Neuralgiform headache attacks with
defect typical of his previous visual aura, as well as
Conjunctival injection and Tearing), or primary stab-
hemiplegia, and was assumed to be simply a particu-
bing headache. No specific tumor type is associated
larly long episode of his usual migraine.
with a specific headache type. Interestingly, only a
few pituitary headaches meet the official criteria for
intracranial neoplasm headache or pituitary hyper- Can migraine be lethal?
secretion headache, and thus revisions to the classifi- We are cheating a bit here, because this case is not
cation have been proposed. Almost half of patients will a patient from our practice. Rather, it was reported
report severe disability due to headache. almost 130 years ago by Féré, a resident of the famous
French neurologist Jean-Martin Charcot. Bousser and
Discussion Welch point to this as “the first case of lethal migrain-
ous stroke” but note that “in the absence of autopsy,
Pituitary headache may be more severe than typi-
the precise cause of death is unknown.” The term “fatal
cal migraine and when unilateral may often be side-
migraine” occasionally appears in the neurologic lit-
locked. Less commonly, pituitary headache may mimic
erature, usually referring to cases in which an appar-
cluster or other more rare forms of primary headache.
ent attack of migraine aura persists and culminates in
Pituitary headache may also be chronic and disabling.
death from stroke. Obviously, it is always difficult in
Growth hormone (GH)- or prolactin-secreting pitu-
such cases to know whether the headache in question
itary tumors are more likely to be associated with
was an early symptom of the stroke, or whether the
headache than nonsecreting tumors. The lack of asso-
stroke was caused by the headache.
ciation between headache and tumor size, volume, or
compression suggests that traction is not the most
likely mechanism of headache. The pathophysiology of How might migraine aura lead to stroke?
these headaches is more likely to reflect the hormonal The way in which an attack of migraine aura might
activity of the tumor than its physical characteristics. cause stroke is not definitely known. A leading pos-
Significant headache relief has been reported in sibility is that an infarct might be caused by cere-
some instances after treatment with somatostatin bral hypoperfusion due to migraine aura. Decreases
analogs, for example octreotide, in GH-producing in cerebral blood flow are known to occur in the set-
tumors, or dopamine agonists, for example cabergo- ting of aura, but typically are not sufficient to pro-
line or bromocriptine, in prolactinomas. duce ischemic symptoms. Migraine with aura is a com-
mon, long duration condition, and sufferers may have
Diagnosis hundreds of auras over a lifetime. Yet true migrain-
Headache associated with growth hormone-secreting ous infarction is rare. In unusual circumstances,
pituitary adenoma. however – such as volume depletion or hypercoagula-
bility – perhaps the hypoperfusion that occurs during
aura is amplified and sufficient to produce a stroke.
Tip Other theories are that migraine with aura is asso-
Pituitary tumors are a common incidental finding on ciated with other stroke-causing conditions such as
neuroimaging for headache, and it is difficult to deter- patent foramen ovale or cerebral artery dissection. It is
mine whether they are the cause of headache. When also possible that vasoconstrictive treatments used for
unilateral the headache is often side-locked, a histor- individual attacks of migraine may occasionally pro-
ical feature that may indicate the headache is more duce stroke. Still another explanation is that migraine
likely to be due to a non-migraine process. is associated with the presence of vascular risk factors

Chapter 3: Missing dangerous causes of headache

such as endothelial dysfunction that raise stroke risk. and in the absence of autopsy results, this diagnosis of
Finally, the increased susceptibility to cortical spread- “fatal migraine” cannot be confirmed.
ing depression that is a hallmark of the migraine-prone
brain may also act more generally to boost susceptibil-
ity to brain ischemia. Tip
True migrainous infarction is rare, and careful evalu-
ation is needed to distinguish it from other causes of
Discussion stroke or persistent aura without stroke.
Current diagnostic criteria in ICHD-3 beta allow a
diagnosis of “migrainous infarction” only in patients
known to have migraine with aura. The stroke must Severe headache in a new mother
develop during a typical aura, and imaging must show
a defect in the relevant brain location. Other expla- Case
nations for stroke must be excluded with a careful A 28-year-old woman developed acute headache six
workup. Several cases have been reported in which hours after a forceps-assisted vaginal delivery for
patients met criteria for migrainous infarction but later which she had epidural anesthesia. She had a his-
proved to have unusual explanations for stroke, includ- tory of episodic migraine with aura that she ordinar-
ing cardiac myxoma and cerebral aneurysms. ily treated with sumatriptan. Her headaches had not
Unsurprisingly, a large proportion of true migrain- improved during pregnancy but she treated them only
ous infarcts occur in the occipital lobe, which fits with with acetaminophen and bed rest.
the fact that most auras are visual and arise from that Eight hours after delivery the patient reported a
part of the brain. In addition to cases of true migrain- severe bifrontal headache with nausea and blurred
ous infarction, there are numerous case reports of long vision. This was similar to her prior migraines but
lasting aura symptoms that resolve without subsequent much more severe. Her neurologic exam showed mild
stroke. A technique known as magnetic resonance dif- diffuse hyperreflexia but was otherwise normal. Uri-
fusion with apparent diffusion coefficient maps has nalysis showed no proteinuria and her blood pressure
been used to distinguish between long-lasting aura and was 132/82 mm Hg. The patient reported that she had
true migrainous infarction. been expecting a bad headache after delivery and asked
For historical interest, let us close with yet another to be treated with sumatriptan.
case from the past: this patient is a 47-year-old man
with a long history of migraine with typical visual aura.
His visual auras always consist of a left scintillating sco- Would you give this patient sumatriptan?
toma, which develops gradually, fades away within 20 The patient is no longer pregnant so fears about unin-
minutes, and is followed by a headache that meets cri- tended effects on the baby are not a reason to avoid
teria for migraine. One day he had a typical attack of the use of sumatriptan. Sumatriptan is also consid-
visual aura but the visual abnormality did not resolve ered compatible with breast-feeding by the American
and he was left with a permanent upper left quadrantic Academy of Pediatrics, so plans to nurse the baby like-
defect. Your thoughts? wise are not a contraindication to its use.
The case is that of a real person, a pathologist by Headache occurs in about a third of women dur-
the name of Frank Mallory (attributed to Polyak and ing the postpartum period and is more common in
reported by Bousser and Welch). He had migraine those with a prior history of migraine. In fact, in one
without aura and at age 47 experienced a persistent study of 1000 women with postpartum headache the
upper left quadrantic defect that arose during a typi- cause was migraine or another primary headache over
cal visual aura. When he died 30 years later an autopsy three-quarters of the time. This patient had a his-
revealed a calcarine infarct. tory of migraine and furthermore her headaches had
not improved substantially during pregnancy. This is
consistent with epidemiologic evidence that women
Diagnosis who have migraine with aura are less likely to expe-
The first vignette is consistent with a diagnosis of rience pregnancy-related improvement of headaches
migrainous infarction and death, but 130 years later than women who do not have aura.

Chapter 3: Missing dangerous causes of headache

Figure 3.1 MRV images of cerebral venous thrombosis showing reduced flow in the right transverse sinus.

It is easy to understand how all of these things The differential diagnosis of postpartum headache
suggested that the patient’s headache was a migraine is long, however, and several other disorders could
and that sumatriptan would be an appropriate way reasonably explain the findings in this case. Causes
to treat it. The patient was given sumatriptan but her of secondary postpartum headache can be divided
headache did not improve. In fact, she became increas- into vascular causes (stroke, venous sinus throm-
ingly groggy and confused and several hours after bosis, arterial dissection, vasculitis, reversible cere-
a second dose of sumatriptan she had a generalized bral vasoconstrictive syndrome), and nonvascular
seizure. causes (preeclampsia/eclampsia, idiopathic intracra-
nial hypertension, post-dural puncture headache).
Notably, seizures occur in almost 40% of patients with
Could this be something other than cerebral venous thrombosis, making that a top diag-
nostic possibility along with eclampsia.
migraine? The patient in this case was treated emergently
It no longer seems particularly likely that the patient’s with intravenous magnesium and phenytoin, and had
headache was due to migraine. Although a diagnosis an MRI, MRA, and MRV looking for vascular abnor-
of migraine “fits the facts,” this seems another case in malities. MRV confirmed transverse sinus thrombosis
which the cognitive error of premature closure might (Figure 3.1). The patient was begun on low-molecular-
have occurred. Other less common causes of headache weight heparin and oral warfarin.
in the postpartum period also might “fit the facts.”
For example, preeclampsia can occur in the post-
partum period, and may be more common in patients Discussion
with a history of migraine, especially migraine with The biggest risk of cerebral venous thrombosis occurs
aura. The patient’s blood pressure was higher than during the third trimester of pregnancy and the month
might be expected in the postpartum setting, and she following delivery. Headache is the most common
had hyperreflexia on examination. Now she has had a symptom, occurring in about 90% of cases. In one
seizure. Could this be eclampsia? series, roughly a third of patients had headaches

Chapter 3: Missing dangerous causes of headache

with migrainous features, which underscores the chal- Graber ML, Franklin N, Gordon R. Diagnostic error in
lenge of distinguishing the early headaches of cerebral internal medicine. Arch Intern Med. 2005;165(13):
venous thrombosis from those of migraine. 1493–9.
Headache in cerebral venous thrombosis is pre- Poole TR, Graham EM, Lucas SB. Giant cell arteritis with a
sumably due to increased intracranial pressure from normal ESR and CRP. Eye 2003;17(1):92–3.
abrupt cerebral venous congestion. Papilledema is also
a common finding. About a third of patients with Sphenoid sinusitis
cerebral venous thrombosis experience an intracranial Digre KB, Maxner CE, Crawford S, Yuh WTC. Significance
hemorrhage. Headaches can persist for some time after of CT and MR findings in sphenoid sinus disease. AJNR
cerebral venous thrombosis is treated. Am J Neuroradiol. 1989;10:603–6.
Guidelines from the American Heart Association Lew D, Southwick FS, Montgomery WW, Weber AL, Baker
and American Stroke Association encourage testing AS. Sphenoid sinusitis. A review of 30 cases. N Engl J
for prothrombotic conditions such as factor V Lei- Med. 1983;309(19):1149–54.
den, protein S, or protein C deficiency in women who Proetz AW. The sphenoid sinus. BMJ. 1948;2:243–5.
have had cerebral venous thrombosis, and recommend
the use of low-molecular-weight heparin throughout Carbon monoxide poisoning
future pregnancies. Balzan MV, Agius G, Galea Debono A. Carbon monoxide
Acute-onset severe headaches, particularly those poisoning: easy to treat but difficult to recognise.
associated with focal neurologic signs, are suggestive Postgrad Med J. 1996;72:470–3.
of vascular etiologies. In this clinical scenario, even in Heckerling PS. Occult carbon monoxide poisoning: a cause
a patient with a history of benign headaches, extensive of winter headache. Am J Emerg Med. 1987;5(3):201–4.
workup is often warranted. In this case, it is unlikely Heckerling PS, Leikin JB, Maturen A, Perkins JT. Predictors
that the patient was directly harmed by sumatrip- of occult carbon monoxide poisoning in patients with
tan, but that would not be true if her headache had headache and dizziness. Ann Intern Med. 1987;107(2):
been due to some other common causes of postpartum 174–6.
headache, such as reversible cerebral vasoconstrictive
syndrome. The use of vasoconstrictive drugs such as Post-ictal headache
the triptans or ergots could produce stroke in those Förderreuther S, Henkel A, Noachtar S, Straube A.
cases. Headache associated with epileptic seizures:
epidemiology and clinical characteristics. Headache.
Diagnosis Ito M, Nakamura F, Honma H, et al. A comparison of
post-ictal headache between patients with occipital lobe
Cerebral venous thrombosis. epilepsy and temporal lobe epilepsy. Seizure. 1999;8(6):
Jacob J, Goadsby PJ, Duncan JS. Use of sumatriptan in
Tip post-ictal migraine headache. Neurology. 1996;47(4):
Although postpartum migraine is common, up to a 1104.
quarter of postpartum headaches are due to other Rogawski MA. Migraine and epilepsy – shared mechanisms
causes, including serious vascular problems that could within the family of episodic disorders. In: Noebels JL,
be worsened by vasoconstrictive medications. For this Avoli M, Rogawski MA, Olsen RW, Delgado-Escueta
reason, it is generally prudent to avoid triptans or AV, editors. Jasper’s Basic Mechanisms of the Epilepsies
[Internet], 4th edition. Bethesda, MD, National Center
ergots in the treatment of early postpartum headache.
for Biotechnology Information (US). 2012.
Scher AI, Bigal ME, Lipton RB. Comorbidity of migraine.
Further reading Curr Opin Neurol. 2005;18(3):305–10.
Temporal arteritis
Ezeonyeji AN, Borg FA, Dasgupta B. Delays in recognition Carotid cavernous fistula
and management of giant cell arteritis: results from a Miller NR. Dural carotid-cavernous fistulas: epidemiology,
retrospective audit. Clin Rheumatol. 2011;30(2):259– clinical presentation, and management. Neurosurg Clin
62. N Am. 2012;23(1):179–92.

Chapter 3: Missing dangerous causes of headache

Medulloblastoma Levy MJ, Matharu MS, Meeran K, Powell M, Goadsby PJ.

Ang C, Hauerstock D, Guiot MC, et al. Characteristics and The clinical characteristics of headache in patients with
outcomes of medulloblastoma in adults. Pediatr Blood pituitary tumours. Brain. 2005;128(Pt 8):1921–30.
Cancer. 2008;51(5):603–7. Melmed S, Casanueva F, Cavagnini F, et al. Consensus
Chan AW, Tarbell NJ, Black PM, et al. Adult statement: medical management of acromegaly. Eur J
medulloblastoma: prognostic factors and patterns of Endocrinol. 2005;153(6):737–40.
relapse. Neurosurgery. 2000;47(3):623–31; discussion
631–2. Stroke in the setting of migraine
Leary SE, Olson JM. The molecular classification of Bousser MG, Welch KMA. Relation between migraine and
medulloblastoma: driving the next generation clinical stroke. Lancet Neurol. 2005;4:533–42.
trials. Curr Opin Pediatr. 2012;24(1):33–9. Féré C. Note sur un cas de migraine ophtalmique à accès
Malheiros SM, Franco CM, Stávale JN, et al. répétéssuivis de mort. Rev Med (Paris). 1883;3:194–201.
Medulloblastoma in adults: a series from Brazil. Kurth T, Chabriat H, Bousser MG. Migraine and stroke: a
J Neurooncol. 2002;60(3):247–53. complex association with clinical implications. Lancet
Neurol. 2012;11:92–100.
Obstructive sleep apnea
Polyak S. The Vertebrate Visual System. Chicago, IL,
Bigal ME, Lipton RB. Modifiable risk factors for migraine University of Chicago Press, 1957.
progression. Headache. 2006;46(9):1334–43.
Bigal ME, Lipton RB. What predicts the change from Cerebral venous thrombosis
episodic to chronic migraine? Curr Opin Neurol.
2009;22(3):269–76. Breteau G, Mounier-Vehier F, Godefroy O, et al.
Cerebral venous thrombosis 3-year clinical outcome in
Neau JP, Paquereau J, Bailbe M, et al. Relationship between 55 consecutive patients. J Neurol. 2003;250:29–35.
sleep apnoea syndrome, snoring and headaches.
Cephalalgia. 2002;22(5):333–9. Saposnik G, Barinagarrementeria F, Brown RD Jr., et al.; on
behalf of the American Heart Association Stroke
Council and the Council on Epidemiology and
Pituitary headache Prevention. Diagnosis and management of cerebral
Giustina A, Gola M, Colao A, et al. The management of the venous thrombosis: a statement for healthcare
patient with acromegaly and headache: a still open professionals from the American Heart
clinical challenge. J Endocrinol Invest. 2008;31(10): Association/American Stroke Association. Stroke.
919–24. 2011;42(4):1158–92.

Pitfalls in diagnostic testing: imaging and

4 lumbar puncture

Many causes of headache can be identified based on Refractory migraine in the emergency
the history and physical examination. Patients who
clearly meet criteria for a primary headache disor- department
der generally do not need additional diagnostic test-
ing. In some cases, however, there is clinical suspi- Case
cion of a possible alternative diagnosis and testing is A primary care physician received a nighttime call
needed to “rule out” a sinister cause of pain. In the from an emergency department physician about a
majority of these situations, clinical suspicion cen- patient of hers, a 20-year-old woman, who had inter-
ters on possible structural or mechanical explanations mittent severe headaches. These occurred on average
for headaches. For this reason, a good case can be twice a month and met criteria for migraine. They typi-
made that imaging tests and lumbar puncture (LP) cally responded well to oral sumatriptan. On occasion,
are the most important diagnostic tools in headache though, the patient awakened with a well-established
medicine, with a correspondingly large number of headache and prominent vomiting and was unable to
mistakes that can be made and adverse events that can keep her oral medications down. She then sought treat-
occur. ment in the emergency department, as she had this
We all accept that failure to order a needed test, time.
misinterpreting test results, or lack of follow-up of Her records showed a total of six emergency
worrisome test results can lead to avoidable harm. We department visits over the last three years for simi-
are less accustomed to thinking about the possibility of lar headaches. The emergency department physician
harm from ordering unnecessary tests, or other test- reported that the patient’s neurologic examination was
related harms that are more difficult to quantify. These normal, and the results of past testing, including three
include the iatrogenic anxiety that can be produced CT scans of the head, one MRI study, and an LP, had
by false-positive test results and the long-term health been normal.
risks of testing, such as the risk of cancer from test-
related radiation.
This chapter reviews cases in which imaging tests
or LP were overused, underused, or misused, as well as What additional testing is needed?
situations in which complications occurred as a result The answer is “none.” Severe, poorly controlled pain
of testing. A common thread in all of these cases is is a medical emergency and an appropriate reason to
that the decision to order a test and the interpreta- seek emergency care. Nothing about this patient’s situ-
tion of test results must be considered in the con- ation, however, suggests a dangerous underlying cause
text of the individual patient and the specific clini- of headache or points to the need for additional test-
cal situation. Test results that do not fit the clinical ing. This patient meets criteria for migraine. She has
picture should prompt careful reevaluation of the had previous emergency department visits for similar
situation. Recognition of adverse events from test- headaches. Vomiting is a common accompaniment of
ing and the appropriate management of test-related her headaches, and her consequent inability to keep
complications are also important aspects of headache down oral medication is the main reason she is seeking
management. care now.

Chapter 4: Imaging and lumbar puncture

It is unlikely that another CT scan of the head Discussion

or other testing will identify anything that will alter
It is not uncommon to encounter patients with
her management. Furthermore, the radiation risks of
headache who have had multiple imaging procedures,
diagnostic imaging are not negligible. They are highly
including many that involve exposure to ionizing
dependent on age, with younger people being at higher
radiation. Repeated, probably unnecessary, scans are
risk than adults. The authors of a recent review of
ordered when physicians and patients are frustrated
the risks of diagnostic CT scans concluded that “In
about a lack of response to treatment and fear some-
summary, there is direct evidence from epidemiologic
thing might have been missed. In other cases, the
studies that the organ doses corresponding to a com-
results of previous scans may not be easily available,
mon CT study . . . result in an increased risk of cancer.
with the result that physicians are unaware of previous
The evidence is reasonably convincing for adults and
testing. The risk of unneeded testing may be especially
very convincing for children.”
high in the emergency department since physicians are
A more reasonable plan would be to forgo addi-
often unfamiliar with the patient and fear missing seri-
tional diagnostic testing and treat the patient with
ous causes of headache. In ordering diagnostic tests,
subcutaneous sumatriptan. If this is effective and well
though, the possible adverse effects of testing must be
tolerated, she could be given a prescription for a suma-
weighed against the likely benefits to the patient. In
triptan auto-injector so that she can give herself subcu-
this case, the balance of benefit to harm for any addi-
taneous injections of the medication when oral treat-
tional testing is not favorable.
ment is not possible. As we mention elsewhere in
this book, patients who have a history of prominent
vomiting in association with migraine often benefit Diagnosis
from having such a non-oral “rescue” treatment option Migraine.
Can the radiation risks of a single head CT be The potential adverse health effects of radiation expo-
quantified? sure should be taken into consideration when ordering
diagnostic testing for headache.
Yes. Radiation doses are measured in units of ioniz-
ing energy absorbed per unit of body mass. The radi-
ation dose varies depending upon the type of imag- A young woman with stable headaches
ing procedure, the number of scans obtained during
the procedure, the equipment used, and other factors. Case
Radiation doses from CT scans are much greater than A 22-year-old woman sought treatment for twice-
doses from conventional X-rays. The estimated radi- monthly headaches that began four years ago. Her
ation dose from a typical CT scan of the head, for most severe headaches were throbbing, located behind
example, is 340 millisieverts. The radiation dose from the right or left temple, and associated with nausea
a single CT scan of the head may seem small, but and photophobia. They typically responded to treat-
there is good research evidence that even exposures ment with an over-the-counter combination medica-
in this range are associated with a statistically signif- tion that contained acetaminophen, aspirin, and caf-
icant increase in the lifetime risk of cancer. These risk feine. Occasionally treatment was ineffective and she
estimates have not been extrapolated from studies of missed a day of work. Her neurologic examination,
higher-dose exposures, but are instead based on long- including funduscopic examination, was normal and
term follow-up of groups with similar exposures. The she had no other medical problems and was taking no
estimated lifetime attributable risk of subsequent brain medications except for headache.
cancer as a result of a single CT scan of the head at age A friend of hers had similar headaches that
20, for example, has been estimated at roughly 0.01%. respond well to almotriptan. The patient requested a
At age ten, the risk is 0.02%. For an individual patient prescription for that medication. She also requested
the risks are low, but on a population level it is esti- an MRI scan of her head, saying “I don’t think my
mated that diagnostic radiation exposure may account headaches are caused by something serious, but you
for 1.5–2% of all cancers. never know. I would feel better if I had a normal scan.”

Chapter 4: Imaging and lumbar puncture

The physician diagnosed migraine, ordered an MRI – that is, test results showing abnormalities that are
scan, and prescribed almotriptan. unlikely to be clinically significant.
A month later the patient returned for a follow- Although this study suggested short-lived reduc-
up visit to review her MRI results. She reported that tions in anxiety in patients with chronic daily headache
almotriptan had worked well for the only headache she who underwent testing, similar studies in other con-
had since her last visit. The MRI shows no mass lesion ditions have not clearly shown that testing reassures
or other abnormalities but the radiologist’s report patients. In a systematic review of five controlled tri-
described a small arachnoid cyst in the posterior fossa. als in a variety of conditions, four of five studies did
Despite an explanation that this finding was unlikely to not find that patients were reassured. The authors con-
be related to her headaches, the patient was not reas- cluded that “the results point in the direction of diag-
sured. She later called to request referral to a neuro- nostic tests making hardly any contribution to the level
surgeon and to obtain a copy of her scan for a second of reassurance.” They suggested that “a clear explana-
opinion. tion and watchful waiting” might make testing unnec-
essary and commented that “if diagnostic tests are
used, it is important to provide adequate pre-test infor-
Were there medical reasons to scan mation about normal test results.”
this patient?
The patient in this vignette met diagnostic criteria When diagnostic imaging is indicated,
for migraine. Her condition had been stable for sev- should patients be warned of the possibility
eral years. A practice parameter from the American
Academy of Neurology discourages the use of diag-
of incidental findings?
nostic imaging or other testing in these circumstances. Discussing possible testing outcomes with the patient
This is because the prevalence of meaningful abnor- before test results come back may be advisable. This
malities in this clinical situation is low, on the order is because patients may have well-established beliefs
of approximately 0.2%, with an upper 95% confidence about the seriousness of their symptoms that make it
limit of approximately 0.6%. It is difficult to argue that difficult for them to believe and be reassured by normal
a brain scan was medically necessary in this patient. or clinically irrelevant test results. The authors of one
Most clinicians have been faced with patients like this study suggested that “providing an explanation about
one who do not have a medical reason for testing but the meaning of normal test results before testing may
who request it nonetheless in order to allay anxiety. weaken patients’ preconceived ideas about their ill-
ness and provide a context to help patients make sense
of the test result.” The authors tested this hypothesis
Does testing reassure patients? in a randomized trial among patients with chest pain
referred for a stress test. Patients who participated in a
Only one study has examined this question in patients discussion of what it would mean if they had normal
with headache. In that study the investigators mea- test results were more likely to report feeling reassured
sured baseline depression and health anxiety levels in
at one and three months, and fewer were taking cardiac
patients with chronic daily headache and then ran- drugs at one month.
domized them to receive or not receive an MRI scan
Based on this evidence, in our practice we make a
of the head. The results showed that patients who
point of discussing possible test results with patients
received a scan were less worried about their health at before we send them for testing. This includes a discus-
three months compared to those who had not received
sion of the possibility that small abnormalities may be
a scan. By a year, however, there were no differences
identified that do not have any diagnostic or treatment
in anxiety between the two groups. The group that implications.
had received a scan, though, had lower overall medi-
cal costs than the group that had not been scanned.
All of the patients in that study had normal test Discussion
results, however, so the findings do not apply to sit- Many doctors feel that patients will not be reas-
uations like this one, in which patients undergo test- sured about the benign nature of their condition with-
ing and are found to have so-called “incidentalomas” out proof from testing that nothing is wrong. It is

Chapter 4: Imaging and lumbar puncture

not uncommon for physicians to acquiesce to patient activities as a result of the headaches and was able to
requests for tests that may not be necessary. Despite exercise if the headaches were low grade.
their good intentions testing can cause unintended A complete review of systems was negative and
adverse effects, including a paradoxical increase in the patient had no prior history of medical problems.
anxiety when results deviate from normality. Order- His general and neurologic examinations were nor-
ing a test may reinforce the patient’s belief that some- mal. The physician did not believe that further diag-
thing is wrong. Some tests involve exposure to radi- nostic testing was indicated. The patient was advised
ation, which may increase the risk of later cancers. to begin a low dose of amitriptyline as a preventive
Test results of uncertain significance may set in motion medication and continue using a triptan to treat severe
yet more diagnostic testing or lead to treatment that headaches.
causes harm, a situation known as the “medical cas-
cade.” Unnecessary testing contributes importantly to
healthcare costs. It is also possible that patients will
Were there medical reasons to scan
continue to feel anxious and worried even after they this patient?
receive normal test results. This young man’s headaches met diagnostic criteria
for chronic migraine. He had experienced a change in
Diagnosis headache pattern – something that might be consid-
ered a “red flag” historical feature – but the transfor-
mation from episodic to chronic headache occurred
a year ago and was gradual. Furthermore, his neuro-
Tip logic examination was normal. Additional reassuring
There is limited evidence to support the view that diag- historical features are the strong family and personal
nostic testing reassures patients who are anxious about history of migraine and the carsickness, which is often
the cause of their headaches. It may be difficult to reas- viewed as a childhood precursor of migraine.
sure patients whose test results reveal minor abnor- Patients like this are familiar to anyone who works
malities of no clinical relevance. in a specialty headache clinic; in fact, this clinical pat-
tern of early childhood onset of migraine headaches
A young man with constant with the gradual development of daily or near-daily
headache is an extremely common clinical scenario.
headaches There is little in this patient’s clinical presentation
to suggest that initial imaging should have been
Case recommended.
A 19-year-old man sought medical treatment for At the patient’s first follow-up visit, he reported
constant, low-grade headaches with superimposed substantial improvement in his headaches with the
episodes of more severe headache several times a week. amitriptyline, and almotriptan worked well for those
He reported a history of carsickness and intermit- that did occur. The patient’s parents, however, were
tent headaches dating back to childhood. Many mem- very concerned about his headaches and requested
bers of his family, including his mother and a sis- neuroimaging. The physician ordered an MRI scan
ter, had severe episodic headaches and had been diag- of the head, which showed a right parietal mass.
nosed with migraine. When he was a child the patient’s Additional workup identified this as a low-grade
headaches occurred on average twice a month, but astrocytoma.
about a year ago they had gradually worsened and
become constant. Severe exacerbations of headache
were described as global; throbbing; and associated
How do you reconcile this finding with the
with photo and phonophobia, nausea, and vomit- fact that neuroimaging is not recommended
ing. These episodes of more severe headaches seemed in cases of stable headache that meet
to be brought on by emotional stress and had been
responsive to both triptans and nonsteroidal anti- criteria for migraine?
inflammatory medications such as aspirin. The patient This is certainly a startling finding. A single case
was not experiencing difficulties in school or daily like this can have a disproportionate effect on a

Chapter 4: Imaging and lumbar puncture

physician’s beliefs about diagnostic testing. Had this be more common than they are. Such cases tend to
finding been undetected at the initial headache eval- stick in the memory and are easily recalled. This is
uation and come to light later on, it is likely that the sometimes called “availability bias” because these diag-
family would have considered that an important diag- noses are easily “available.” It is not difficult to imag-
nosis had been delayed or missed. Low-grade astrocy- ine that this particular case, in which a serious prob-
tomas are slow-growing glial tumors that occur most lem was found in a patient who had little indication for
frequently in younger people. Since they can undergo scanning, might reinforce a doctor’s belief that “every-
malignant transformation, treatment of some type is one should be scanned because otherwise you might
usually recommended and prolonged survival is pos- miss something.”
sible. It seems likely that the fortuitous discovery of this “Defensive medicine” refers to the practice of
patient’s tumor improved his long-term outcome. ordering tests that are not medically indicated in order
to protect the physician from an accusation of medi-
cal malpractice if rare but serious disorders are missed;
Should an event like this change your in other words, situations such as the case described
practice? Since it is never possible to be in this vignette. It would be difficult, perhaps impos-
certain a patient doesn’t have something sible, to convince a jury that the headaches experi-
enced by a patient such as this one were really due to
seriously wrong, shouldn’t we image migraine, especially in hindsight and with an abnor-
everyone with headache? mal test result. Should the possibility of rare, serious
There is no right or wrong answer to these ques- disorders lead doctors to order extensive diagnostic
tions; rather, they are something that individual doc- testing “just in case?”
tors must decide for themselves. We believe that physi- Several features of this case suggest that the
cians should think very carefully before adopting an tumor was unrelated to the patient’s headache prob-
indiscriminate approach to testing in order to reduce lem and represented an incidental imaging find-
liability risks. Medical ethics experts have pointed out ing. This patient’s history is consistent with long-
that defensive testing “essentially transfers test utility standing migraine that began at a time when he
from the patient (for care) to the physician (for self- could not have had a brain tumor. It is also unlikely
protection).” This is in conflict with the physician’s that a tumor like this would present as episodic
professional obligation to put the patient’s interests migraine followed by transformation to a chronic
ahead of his or her own. Since diagnostic testing can pattern. Finally, his headaches improved with treat-
expose patients to harm, ordering tests that are not ment aimed at migraine. Taken as a whole, it seems
indicated also is in conflict with the physician’s ethical far more likely that the tumor, at the time it was
obligation to “first do no harm.” detected, was asymptomatic. In fact, further his-
Finally, it is not clear that such an approach really tory in this patient is that although he had tran-
does reduce the risk of a malpractice lawsuit, since sient improvement in headaches once the tumor was
numerous studies show that the likelihood patients removed, headaches returned rapidly after surgery and
will file a malpractice suit has little to do with actual resumed a chronic pattern. He reestablished care in
malpractice or the extent of harm suffered. Extensive the headache center for management of his chronic
reliance on testing is probably more common among migraine.
less capable physicians. One study of test-ordering Thus, we do not think that this case stands as evi-
behavior among neurologists found that older doc- dence that all patients should be scanned in order to
tors, those who felt more confident in their diagnoses, avoid missing rare events. Rather, it is a reminder that
and those who were board certified ordered fewer “uncommon events occur uncommonly” but can pro-
tests. duce an outsized impact that may tend to sway us from
our principles.

Physicians can be strongly influenced by unusual or Diagnosis
especially dramatic cases, and falsely perceive them to Chronic migraine; incidental low-grade astrocytoma.

Chapter 4: Imaging and lumbar puncture

Table 4.1. Selected causes of thunderclap headache

r Primary
It is not practical, possible, or desirable to image all r
Primary thunderclap headache
patients with headache. We recommend that doc- r
Exertional headache
tors stick to their principles and image only when r
Cough headache
indicated. Headache associated with sexual activity
r Secondary
Reversible cerebral vasoconstrictive syndrome (RCVS)
Subarachnoid hemorrhage
Severe, sudden headache while r
Intracerebral hemorrhage
mowing the lawn r

Cerebral venous thrombosis
Arterial dissection
Case r

Pituitary apoplexy
Colloid cyst
A 44-year-old woman presented to the emergency
department one hour after the sudden onset of an
extremely severe, generalized headache. The headache
occurred while she was mowing the lawn on a warm is to “think blood vessels.” Nonvascular causes of
summer day. There was associated nausea and mild thunderclap headache do exist, though. These include
photophobia but no phonophobia, and the pain was large colloid cysts which can intermittently block the
so severe she didn’t want to move. She denied any third ventricle (producing transient hydrocephalus),
aura prior to the headache. She had a remote history and pituitary apoplexy.
of migraine in her 20s but no significant headaches If secondary causes of thunderclap headache are
recently. She reported that this headache was very dif- ruled out, several primary headache disorders that
ferent from her previous migraines. can sometimes present in this way should be con-
sidered. These include headaches provoked by activ-
ity (such as exertional headache, cough headache, and
What is the clinical term used to describe headache associated with sexual activity) as well as pri-
mary thunderclap headache. More controversial is so-
the sudden onset of extremely severe called “crash migraine.” This is an informal diagnos-
headache, and what is the differential tic term used by some headache experts to describe
diagnosis? migraine attacks that build up quickly to peak inten-
sity, rather than exhibiting the more gradual increase
The sudden onset of extremely severe headache is
in pain intensity that is more common in migraine.
referred to as “thunderclap headache.” The classic clin-
For the patient described in this vignette, all of
ical scenario is a headache that reaches maximal inten-
these diagnoses must be considered. In particular,
sity within seconds or minutes of onset. Thunderclap
her previous history of migraine does not protect her
headache is a clinical presentation rather than a diag-
against the many secondary causes of thunderclap
nosis, and the list of entities that can cause it is exten-
headache. She requires urgent evaluation for danger-
sive (see Table 4.1). The experience is so dramatic that
ous causes of thunderclap headache, and the investiga-
it is often likened to being hit by lightning (perhaps a
tion should be as thorough as it would be in a patient
more illustrative image than a peal of thunder!)
with no prior history of headaches.
Both primary and secondary disorders can cause
thunderclap headache. The best known and most-
feared cause of thunderclap headache is subarachnoid What workup should she have?
hemorrhage, and initial workup appropriately focuses All patients presenting to the emergency department
on this possibility. Other causes include intracerebral with thunderclap headache, including this patient,
hemorrhage, reversible cerebral vasoconstrictive syn- should have a noncontrast-enhanced CT of the head to
drome (RCVS), arterial dissection, stroke, and cerebral evaluate for subarachnoid hemorrhage. It is well estab-
venous thrombosis. This long list of possible vascular lished that CT may be negative in 5% of cases of sub-
sources means that a good rule of thumb in formulat- arachnoid hemorrhage, however, so a follow-up LP in
ing a differential diagnosis for thunderclap headache this situation is mandatory. False-negative findings on

Chapter 4: Imaging and lumbar puncture

obtaining a head CT in this setting, and most also

would perform an LP if the CT is negative. In contrast,
fewer physicians seem to be aware of RCVS or other
potentially dangerous causes of thunderclap headache.
Nonetheless, we believe that the standard of care in
such cases is evolving rapidly and that it is prudent
to perform additional testing when the CT and LP are
normal in patients with thunderclap headache.
This level of diagnostic testing is justified because
many of the conditions which cause secondary thun-
derclap headache are life-threatening. If RCVS is
suspected it is useful to obtain a detailed history
of recent substance use. Vasoconstrictive substances,
both prescribed and illicit, are often associated with
RCVS. Patients should be carefully questioned about
the use of sympathomimetic agents such as decon-
gestants or cocaine. Selective serotonin reuptake
inhibitors and selective serotonin–norepinephrine
Figure 4.1 One image from a CT angiogram showing multifocal reuptake inhibitors, as well as marijuana, are also sus-
segmental areas of vascular narrowing that can be associated with pected of being occasional causes of RCVS.
reversible cerebral vasoconstrictive syndrome (RCVS).
Treatment of thunderclap headache depends on
the underlying cause. If RCVS is found or strongly sus-
pected despite negative imaging, most clinicians would
LP can occur, however, since xanthochromia may not
consider the use of a calcium channel antagonist such
be seen until several hours after the initial develop-
as verapamil or nimodipine. If secondary causes of
ment of headache. This means that visual inspection
thunderclap headache have been convincingly ruled
may not detect xanthochromia when LP has been per-
out, treatment for a primary cause is usually in order.
formed very early after onset of headache. Spectropho-
This may include the use of calcium channel antag-
tometry is helpful in these cases.
onists or such things as indomethacin if exertional
Even when subarachnoid hemorrhage has been
headache is suspected. Reassurance is also important;
ruled out by appropriate investigation, however, fur-
most cases of drug-induced or primary thunderclap
ther diagnostic testing should be strongly consid-
headache resolve spontaneously, usually within a few
ered. Reversible cerebral vasoconstrictive syndrome is
months. Avoidance of vasoconstrictive medications
increasingly recognized as a frequent cause of thun-
during this period of time is important, however. Evi-
derclap headache. In RCVS there is segmental con-
dence is lacking, however, about whether avoidance
striction of cerebral arteries, which produces a classic
of vasoconstrictive substances should be continued
beaded appearance of the vessels. Typically this con-
beyond this period of time.
striction begins in the distal vessels and spreads proxi-
mally. RCVS (as well as arterial dissection) can be rec-
ognized with MRI and MRA. Depending on the his-
tory, a magnetic resonance venogram may also be war- Diagnosis
ranted. In the patient described in the vignette, the Reversible vasoconstrictive syndrome (RCVS) pre-
initial CT and LP were normal but a CT angiogram senting as thunderclap headache.
(Figure 4.1) demonstrated the characteristic beading
appearance of RCVS.
Discussion Thunderclap headache is a clinical presentation rather
Thunderclap headache is an important presentation than a diagnosis. Workup of thunderclap headache in
of headache in emergency settings. In our experi- the emergency department begins with a noncontrast
ence, most physicians are aware of the importance of head CT and LP; if these are negative an MR or CT

Chapter 4: Imaging and lumbar puncture

angiogram should be strongly considered to assess for no evidence of such progression in men, and migraine
RCVS. was not associated with progression of other brain
lesions. There also was no association between the
presence of white matter lesions and cognitive pro-
Incidental white matter lesions in cessing or speed. This finding is consistent with evi-
a migraineur dence from a French study that showed no association
between migraine and the risk of dementia or cogni-
Case tive impairment.
A 34-year-old woman reported intermittent headaches
since age 14. About three-quarters of headaches were
preceded by visual aura lasting roughly 45 min- Is any further workup or treatment
utes. On four occasions these visual symptoms had required?
lasted six to eight hours. Upon hearing this history,
If the white matter lesions are small, few in number,
the patient’s primary care physician was concerned
and the patient has no symptoms other than migraine,
about the possibility of a stroke or other intracra-
as with the patient in this vignette, no further workup
nial lesion and ordered an MRI scan of the head. The
is recommended. If the lesions are extensive, large,
MRI report described “several small areas of abnor-
or distributed in locations that are characteristic of
mal increased signal seen in the white matter of the
specific demyelinating disease, additional investiga-
hemispheres bilaterally on the T2 and FLAIR (fluid
tion may be warranted. White matter lesions seen on
attenuated inversion recovery) sequences. This can
MRI have a broad differential diagnosis. Etiologies to
be seen in migraine but other demyelinating pro-
consider include multiple sclerosis, cerebral autosomal
cess cannot be entirely excluded. Clinical correlation
dominant arteriopathy with subcortical infarcts and
is advised.” Despite an explanation that isolated T2
leukoencephalopathy (CADASIL), mitochondrial dis-
hyperintensities are commonly found in people with
ease, vasculitis, or infectious etiologies such as Lyme
migraine and are not worrisome, the patient did not
disease. Many of these disorders have characteristic
seem reassured. She later called to request testing
appearances on imaging. Most patients with these dis-
for Lyme disease and referral to a multiple sclerosis
orders also have accompanying neurologic signs and
symptoms in addition to headache which suggest a
diagnosis other than migraine.
What is known about white matter lesions There is no specific treatment for white matter
lesions. Obviously, treatment of the headache disor-
in people with migraine? der that the patient initially presented with is usually
White matter lesions on MRI brain scans are more advisable. It is unknown whether preventive treatment
commonly found in migraineurs than the general pop- of migraine might decrease the risk of developing new
ulation. A Dutch population-based study showed that white matter lesions. The use of triptans or ergots is not
patients with migraine were more likely than controls contraindicated in patients with white matter lesions,
to have deep white matter lesions. The increased risk of since the use of these medications has not been asso-
white matter lesions in migraineurs was also found in ciated with their development. However, some stud-
a meta-analysis of case–control studies. A number of ies showed that patients with white matter lesions had
studies have also suggested an increased risk of cere- more cardiovascular risk factors than those without,
bellar infarcts in migraineurs. for example smoking or hypertension. It remains pos-
In the Dutch study, the risk of white matter lesions sible that there are subgroups of migraineurs who are
was greater for patients with migraine with aura (com- at elevated risk of not only white matter lesions but also
pared to no aura) and in those with higher attack fre- ischemic brain lesions. Thus, it is prudent to search for
quency (at least 12 attacks per year). After nine years and address modifiable cardiovascular risk factors in
of follow-up, the original lesions persisted but had not patients with migraine.
increased in size. However, women with migraine had There is no evidence to support the use of
an increased incidence of new white matter lesions, but antiplatelet or antithrombotic therapy such as low-
the overall volume of lesions remained low. There was dose aspirin in migraine patients with white matter

Chapter 4: Imaging and lumbar puncture

lesions. In fact, a recent analysis of data from the Tip

Women’s Health Study showed that women who had
Nonspecific white matter FLAIR hyperintensities are
migraine with aura randomized to daily aspirin had an
often seen on MRI in patients with migraine and if the
increased risk of myocardial infarction compared with
pattern is not suggestive of another disorder, no fur-
those who did not receive aspirin. In view of this find-
ther workup or treatment is required. Patients should
ing, we do not recommend aspirin therapy in patients
be reassured that these lesions are common and are not
with migraine who have white matter lesions and no
associated with any adverse outcomes.
other indication for aspirin therapy.

Familial occurrence of white

Discussion matter lesions
Patients with migraine who have scattered small
white matter lesions on MRI can be reassured that the Case
lesions do not indicate a dangerous underlying con- A 44-year-old woman was referred to the headache
dition. It can be difficult to reassure patients, however, clinic for evaluation of an abnormal MRI brain scan.
if this finding is unexpected, and particularly when The patient reported a three-month history of near-
the radiologist’s report mentions possible diagnoses daily headaches in the context of substantial work and
such as multiple sclerosis. This is a situation where family stress. Headaches were located behind both eyes
discussing possible test results with patients before a and were mild or moderate in intensity and described
scan is ordered can help allay anxiety when results as “squeezing, like a hat that’s too tight.” She felt cog-
come back. When ordering an MRI scan of the brain, nitively “fuzzy” during headaches but had no nausea,
for example, we routinely tell patients that it is not vomiting, or sensitivity to light or sound. Headaches
uncommon to find “white bright spots” on the MRI. responded well to ibuprofen. The headaches were get-
We tell them that the importance of these findings is ting better and her preference was not to add addi-
not known for sure, but that we do not believe they tional treatment. She had been referred because her
are anything to worry about. primary care doctor ordered an MRI scan of the head
It is not clear why white matter lesions occur in to assess her headaches. The scan (Figure 4.2) showed
migraineurs. Evidence is conflicting about whether white matter abnormalities that were felt by the radi-
the presence of a right-to-left cardiac shunt (such as ologist to be unusual for her age.
patent foramen ovale) is associated with a greater
burden of white matter lesions. Some theorize that
the lesions are caused by transient localized hypo- What is the differential diagnosis of this
perfusion during migraine or migraine aura, possibly pattern of white matter abnormalities?
caused by oligemia of the deep penetrating vessels, or
The list of conditions that can produce white matter
by neuronal hyperexcitability. An alternative hypothe-
abnormalities is very long. These lesions can be seen in
sis is that migraine and migraine aura are associated
demyelinating conditions; with inflammatory or neo-
with increased permeability of small meningeal ves-
plastic processes; as a result of trauma; or in associ-
sels, perhaps triggered by cortical spreading depres-
ation with congenital, metabolic, toxic, degenerative,
sion, the phenomenon generally accepted as the patho-
or vascular disorders. This patient has no hyperten-
physiologic process underlying migraine aura. This
sion, is not a smoker, and does not have diabetes. Test-
breakdown in the blood–brain barrier could theoret-
ing for Lyme disease and coagulopathies was nega-
ically allow neurotoxic substances into the surround-
tive. Her neurologic examination was normal with the
ing white matter, which might act in unknown ways to
exception of a blunted affect and some subtle cognitive
cause white matter lesions.
The patient reported that her 71-year-old mother
had experienced troublesome headaches for many
Diagnosis years, and had recently developed cognitive prob-
Migraine with aura. lems. Her mother had radiation therapy for a

Chapter 4: Imaging and lumbar puncture

Figure 4.3 The MRI scan of the patient’s mother, showing

extensive white matter lesions of a size and distribution compatible
with a diagnosis of CADASIL.
Figure 4.2 An MRI showing extensive white matter lesions in a
44-year-old woman.
CADASIL, however, by associated physical examina-
pituitary tumor over a decade ago. Two years ago tion findings and laboratory testing. In this case, the
she had developed confusion and the family had her mother’s scan (Figure 4.3) shows extensive white mat-
evaluated for dementia. A meningioma was found ter lesions of a size and distribution compatible with a
and removed, but the cognitive problems persisted. diagnosis of CADASIL.
The patient remarked that her mother “also had some
other abnormalities on her MRI scan” but could not
recall what they were.
CADASIL is a genetically determined disorder caused
by a NOTCH3 mutation on chromosome 19. Ini-
What is the next step? tial symptoms can include migraine-like headaches
This patient’s white matter lesions are extensive for her or vague psychiatric symptoms. The disorder usually
age. This in conjunction with her mild cognitive abnor- presents with brain dysfunction but it is actually a
malities and the report of headache and MRI abnor- systemic disorder, due to abnormal accumulation of
malities in her mother raise the possibility of a famil- the NOTCH3 protein in vascular smooth muscle and
ial syndrome. The next step in this case was to obtain the brain. A single-gene, autosomal recessive form of
the mother’s MRI scan for comparison. The list of arteriopathy with similar clinical manifestations has
genetic disorders that manifest as white matter abnor- recently been discovered and termed CARASIL (cere-
malities in conjunction with headache is quite short, bral autosomal recessive arteriopathy with subcortical
and consists of mitochondrial abnormalities (such infarcts and leukoencephalopathy).
as mitochondrial encephalomyopathy, lactic acido- Molecular genetic testing can be done for
sis, and stroke-like episodes [MELAS] and myoclonic CADASIL to identify the NOTCH3 gene. Skin biopsy
epilepsy with ragged red fibers [MERFF]) or neuro- can also be diagnostic, since NOTCH3 protein accu-
cutaneous disorders such as neurofibromatosis. These mulates at the cytoplasmic membrane of smooth
disorders can usually be easily distinguished from muscle in peripheral as well as cerebral blood vessels.

Chapter 4: Imaging and lumbar puncture

The biopsy specimen, however, needs to be deep Diagnostic criteria for post-dural puncture
enough to include material from the border zone headache do not require confirmation of the diagnosis
between the deep dermis and the upper subcutis. This with neuroimaging or other testing in the setting of
patient was offered referral for genetic counseling a known dural puncture that is followed within five
and possible testing for CADASIL but declined the days by a postural headache. The criteria specify that
referral. She preferred to adopt a “watchful waiting” the headache should worsen within 15 minutes of
approach. standing up, and improve within 15 minutes of lying
down. One of the following must also be present:
nausea, photophobia, hypacusia, tinnitus, or neck
Diagnosis stiffness. Finally, the headache must improve sponta-
Probable CADASIL. neously within a week or within 48 hours of effective
treatment of the leak.
Tip Thus, neuroimaging was not required to make the
diagnosis in this case. It did, however, show charac-
Extensive white matter abnormalities in conjunc-
teristic abnormalities that supported a diagnosis of
tion with a family history of headache and neuro-
post-dural puncture headache, in this case brain sag-
logic abnormalities should prompt consideration of a
ging and diffuse pachymeningeal enhancement on an
genetic syndrome.
MRI with gadolinium. Similarly, a demonstration of
low cerebrospinal fluid (CSF) pressure on LP is not
A young man with headache following required to make a diagnosis of post-dural puncture
headache. If there is no history of a dural puncture,
lumbar puncture however, then confirmation of a diagnosis of low CSF
pressure headache requires at least one of the following
Case to be demonstrated: pachymeningeal enhancement on
A 28-year-old man who had an LP developed a MRI, evidence of a CSF leak (with myelography or cis-
headache within 12 hours of the procedure. It wors- ternography), or a CSF opening pressure of less than
ened within a few minutes when he stood upright and 6 mm Hg in the sitting position on LP.
improved rapidly when he lay flat. It was associated
with nausea and photophobia. The pain was severe.
He was treated conservatively with bed rest, hydra-
tion, and caffeine for several days but the headache Since the epidural blood patch was only
did not improve. On the basis of his clinical presenta- temporarily successful, should this patient
tion, a diagnosis of post-dural puncture headache was
made. He underwent a blood patch which provided be treated with an epidural saline infusion
some relief of headache for a day, but it then returned or intravenous caffeine?
full force. His neurologic examination was benign. A Epidural saline infusions, intravenous caffeine, and the
gadolinium-enhanced MRI of the brain showed flat- use of abdominal binders are all treatments that have
tening of the pons and pachymeningeal enhancement. been reported helpful in cases of low CSF pressure
(Similar to changes depicted in Figure 4.5, although headache. Before resorting to such measures, though,
that is not the scan of this patient.) it is appropriate to repeat the epidural blood patch.
A recent Cochrane review confirmed the benefits of
Did this patient need neuroimaging or epidural blood patches for treatment of post-dural
puncture headache. Epidural blood patches should be
additional testing? done with at least 20 mL of autologous blood, which
This patient had postural headache that occurred is typically injected one segment below the level of the
immediately following dural puncture. He had no dural puncture. The blood is thought to form a fibrin
additional features such as fever or focal neurologic clot over the dural tear so that healing can occur. If a
findings that suggested an ominous cause of headache. first blood patch does not succeed, a second one should
The most likely diagnosis was post-dural puncture be performed, and many experts would repeat this a
headache due to a persistent spinal fluid leak. third time if needed. A day or so of bed rest following

Chapter 4: Imaging and lumbar puncture

the patch is often recommended although there is no A college student with malaise and a
strong evidence that this improves outcomes.
Pharmacologic therapy can be helpful when change in headache pattern
an epidural blood patch is ineffective. A recent
Cochrane review concluded that caffeine is more Case
effective than placebo for treating post-dural puncture A young male college student and athlete with a his-
headache. Theophylline, gabapentin, and hydrocor- tory of episodic migraine without aura sought eval-
tisone decreased pain severity scores compared with uation for a recent change in headache pattern. He
usual care or placebo. Evidence was uncertain for the reported malaise and a three-day generalized, non-
effectiveness of sumatriptan or adrenocorticotropic descript headache of moderate severity. He contin-
hormone (ACTH). The role of intravenous fluids in ued to attend class but was not concentrating well and
treatment of post-dural puncture headache is also was not able to exercise because it made his headache
uncertain. worse. On examination, he was afebrile but his neck
was slightly stiff. Because he was very muscular, the
physician thought this might be due to an inability to
relax his neck muscles during the examination, which
Discussion was otherwise normal. The physician considered per-
The decreased volume of CSF that results from a forming an LP but recalled the patient he had seen ear-
spinal fluid leak causes headache through traction lier in the day who developed a post-dural puncture
and tugging on pain-sensitive structures such as the headache. She had been angry and reported that none
large vessels of the brain. It also results in pachy- of her prior headaches had been as bad as “the one you
meningeal enhancement on MRI because of com- gave me.” The physician advised the college student
pensatory engorgement of venous structures, which that his symptoms were likely to be from viral menin-
expand to fill the void caused by low levels of CSF. gitis, were probably benign, and should respond to rest
Almost a quarter of patients with low CSF pressure and analgesic medication.
documented on LP, however, have normal imaging
The use of smaller gauge pencil-point needles When should a lumbar puncture be
(rather than cutting needles) when performing an LP performed for the evaluation of headache?
appears to reduce the incidence of post-dural punc- The broad answer is that LP should be considered
ture headache. Routine bed rest following LP does in the evaluation of any patient where the physician
not appear to decrease the likelihood of post-dural suspects a possible hemorrhage, infection, malig-
puncture headache. Although blood patches are effec- nancy, or alteration in spinal fluid pressure as a pos-
tive for treating post-dural puncture headache once it sible cause of headache. When an LP is performed,
has occurred, evidence does not support using blood it is routine to order laboratory testing of the fluid
patches prophylactically to prevent the problem. that includes a cell count on the first and last (usu-
ally the fourth) tubes of spinal fluid obtained, a spinal
fluid protein level, and glucose determinations as well
Diagnosis as infection-related testing. Additionally, the physician
Post-dural puncture headache. may order other special studies depending on the indi-
vidual patient’s situation. In our view, an opening pres-
sure should be obtained routinely in all patients under-
going LP for the evaluation of headache, and the result
Tip should be recorded in the chart; a closing pressure can
Neuroimaging is not required to make a diagnosis of be obtained as well if warranted, which is usually in the
low CSF pressure headache in the setting of a known setting of elevated intracranial pressure.
dural puncture followed by postural headache. If a first It can be challenging to decide when there is suf-
epidural blood patch is not successful, it should be ficient suspicion of one of these secondary headaches
repeated with a larger volume of blood up to three to warrant performing an LP. The possibilities of
times. adverse effects of the LP, particularly post-dural

Chapter 4: Imaging and lumbar puncture

puncture headaches, weigh heavily in this decision. Table 4.2. Selected risk factors for headache following
dural puncture
The overall severity of the clinical situation and an
assessment of how sick a patient is are both commonly
r Female sex
taken into account. When the physician is still unde- r Age between 31 and 50 years
cided, a clinical pearl that has been passed on from r Prior history of post-dural puncture headache
senior to younger physicians is that “the indication for r Use of a cutting needle (e.g. Quinke)
r Needle size
an LP is whenever you think of it.” In other words, r Cutting needle bevel orientation perpendicular to the
experienced physicians have learned that it is impor- long axis of the spinal column during the procedure
tant to maintain a high index of suspicion for danger-
ous illnesses that can only be diagnosed by LP.
The patient described in this vignette had a wit- It is unknown whether post-dural puncture
nessed seizure at home later that night. He was taken headaches in migraineurs last longer or are more
to the emergency department where a CT of the head difficult to treat than in non-migraineurs, but our
was negative but an MRI of the brain showed subtle clinical experience suggests this might be the case.
bitemporal enhancement. His spinal fluid test results Fortunately, there is no evidence to suggest that
were consistent with a viral process. Antiviral therapy the development of post-dural puncture headache
was started. Polymerase chain reaction (PCR) testing will have a lasting negative effect on the patient’s
for herpes simplex virus (HSV) was later reported as headache pattern. Thus, the decision to perform an
positive. The patient was diagnosed with herpes sim- LP in situations that warrant it should not depend
plex encephalitis and eventually recovered. on whether the patient has a history of migraine.
The physician in this vignette was clearly hoping
to avoid adding to this patient’s existing burden of
Was the hesitance to perform a lumbar headache by avoiding a post-dural puncture headache.
puncture in this patient warranted? Unfortunately, his reluctance to perform an indicated
Post-dural puncture headache is a complication of LP diagnostic procedure led to a delay in diagnosis of a
that occurs in up to 30% of procedures. The likelihood serious condition.
of this problem, and physician memories of unhappy It is our impression that, for a variety of reasons,
patients who have experienced this complication, may LP is an underused diagnostic procedure in a refer-
make some physicians reluctant to perform the pro- ral headache clinic and even emergency department
cedure, especially when secondary pathology is not settings. Careful attention to optimal procedural tech-
highly suspected and the study is being considered nique, as described in the next vignette, could improve
simply to “rule out” other more serious conditions. On both the clinician’s and patient’s experience with LP
the other hand, in this particular case, the physician and decrease reluctance to perform it when indicated.
involved probably weighed the risk of post-dural punc-
ture headache too highly based on his recent and eas- Diagnosis
ily recalled adverse experience with another patient. Herpes simplex encephalitis.
This is another example of availability bias, in which
memorable events are easily recalled and erroneously
thought to be more likely or common than is the case. Tip
Indicated diagnostic procedures such as LP should
not be avoided in migraineurs because of a fear of
Discussion complications.
Risk factors for the development of post-dural punc-
ture are listed in Table 4.2. Some, such as the choice Optimizing the performance of
of needle used in the procedure, are modifiable, but
others, such as patient sex, are not. Migraineurs may lumbar puncture
be at increased risk for post-dural puncture headache,
so it is especially important that the pre-procedure Case
informed consent discussion includes a review of post- A physician was covering the neurology consulta-
dural puncture headache. tion service of his local hospital one weekend. A

Chapter 4: Imaging and lumbar puncture

2. Explain the major steps of the procedure, how the

patient will be positioned, and review aftercare
3. Have the patient lie on an examining table or bed
in the lateral decubitus position with knees curled
up to the chin, or if seated ask them to lean over a
spinal cord solid surface on which they rest their elbows,
since these positions spread the vertebrae and
epidural space create maximum space between the interspinous
4. Palpate for landmarks over the spinous processes
at L4–5, L3–4, or L2–3 levels. If these cannot be
subarachnoid palpated, locate the sacral promontory, which
space marks the interspace of L5–S1. Use this
reference to locate L4–5 for the entry
5. Sterilize and drape the area after identifying
Figure 4.4 Anatomy of a lumbar puncture.
landmarks. Use lidocaine 1% with or without
epinephrine to anesthetize the skin and the
deeper tissues under the insertion site.
first-year resident approached him and explained that
6. Assemble needle and manometer. Attach the
she needed to perform an LP on a patient she had just
3-way stopcock to manometer.
admitted. She was not confident of her ability to do the
procedure and asked if the physician could “walk her 7. Insert LP needle through the skin and advance
through” the technique. through the deeper tissues. A slight pop or give is
felt when the dura is punctured. The angle of
insertion should be slightly towards the navel,
between the vertebrae. If you hit bone, partially
How do you advise the resident? withdraw the needle, reposition, and
LP is considered a routine medical procedure and most re-advance.
physicians gain experience performing LPs in their 8. When CSF flows, attach the 3-way stopcock and
first post-graduate year. As with the resident in the manometer. Measure intracranial pressure, which
vignette, though, most residents rely on advice and should be 20 mm H2 O or less. Remember that
guidance from senior physicians to learn how to per- any pressure reading performed when the patient
form the procedure, which they then perfect through is sitting will be unreliable.
trial and error. The use of simulation-based training 9. If CSF does not flow, or you hit bone, withdraw
techniques has been shown to improve the perfor- needle partially, recheck landmarks, and
mance of LP in clinical settings, but not all training re-advance.
programs provide simulator training. Figure 4.4 illus- 10. Once the pressure has been recorded, remove the
trates the anatomic structures and approach to doing 3-way stopcock, and begin filling collection tubes
an LP. There are many descriptions in the medical lit- 1–4 with 1–2 mL of CSF each:
erature of the techniques for performing LPs. The fol-
lowing is a description of our preferred approach. a. Tube 1: glucose, protein, protein
electrophoresis, cell count
b. Tube 2: Gram’s stain, bacterial and viral
Procedure cultures
1. Obtain informed consent; discuss with the patient c. Tube 3: cell count and differential
the possibility of complications, including d. Tube 4: reserve tube for any special tests.
post-dural puncture headache, bleeding, or
infection, and how those complications would be 11. After tap, remove needle, and place a bandage
handled. over the puncture site.

Chapter 4: Imaging and lumbar puncture

a. Document the procedure in the patient’s The headache was absent when he awakened in the
medical record. morning but would recur within ten minutes after the
b. Include in your note a brief history and patient got out of bed. A noncontrast head CT scan
physical examination of the patient, the was normal. Low CSF pressure headache was consid-
reasons for performing the LP, and consent. ered as a diagnosis but excluded when a normal CSF
Note in particular a brief examination of the opening pressure was found on LP and a normal CSF
cranial nerves, presence or absence of tracer study showed no apparent leak. The patient was
papilledema, or any other lateralizing diagnosed with cervical degenerative disease and cer-
neurologic finding. Also include a brief note vicogenic headache. He unsuccessfully tried a vari-
of examination of the patient’s spine with ety of treatments, including nerve blocks and facet
attention to any obvious spinal deformity. joint injections, and when seen in follow-up was no
c. Document position of patient during the better.
procedure, opening pressure, and clarity/color
of the CSF. Once results of the CSF analysis
are available, they can be appended to your
Is cervicogenic headache the best diagnosis
note. in this patient?
Certainly this story is more suggestive of a secondary
headache than of migraine, and given the patient’s
Are there guidelines for the performance of age it was almost a certainty that testing would show
some osteoarthritis of the neck. There is, however,
lumbar puncture? very little specific history or diagnostic information
The American Academy of Neurology has published in his case to suggest cervicogenic headache as a
recommendations for the performance of LP. These likely diagnosis. The term cervicogenic headache is
suggest: (1) using the smallest needle size possible; (2) used to describe a disorder of the cervical spine or
using noncutting atraumatic 22 gauge Sprotte spinal adjacent structures that produces pain referred to the
needles to reduce the frequency of post-dural puncture head or face. Apart from the finding of degenerative
headache. If noncutting needles are used, the stylet changes consistent with cervical spondylosis, there is
should be replaced before the needle is withdrawn little here to suggest cervical pathology as the cause of
from the intradural space. (3) Keeping bevel orienta- the patient’s headache. Furthermore, cervical spondy-
tion parallel to the long axis of the spinal canal if cut- losis and osteochondritis, since they are so commonly
ting needles are used. (4) Bed rest or fluid loading have present asymptomatically, are explicitly mentioned
not been shown to reduce the risk of post-dural punc- in International Classification of Headache Disorders
ture headache, so it is not necessary to limit activity (ICHD)-3 beta as conditions that “may or may not be
after the procedure. a cause of cervicogenic headache.” In this patient, who
has no physical findings of decreased cervical range
Misinterpreting the workup in low CSF of motion or worsening of headache with provoca-
tive neck maneuvers, cervicogenic headache does not
pressure headache seem a likely diagnosis. In addition, the patient did
not improve with blockade of nerves supplying cer-
Case vical structures, which makes the diagnosis even less
A 68-year-old male presented with a persistent likely.
headache of six months’ duration. There was no past
personal or family history of migraine. There was
no history of trauma or instrumentation. Initially he What features of the history are of most
had noticed a brief intermittent global headache trig- value in establishing the correct diagnosis in
gered by coughing and at times associated with neck
pain and tinnitus. The headache progressed to become this case?
daily with or without trigger but had cleared com- Historical features of note in this case included the
pletely within five minutes when the patient lay down. positional nature of the headaches, which began

Chapter 4: Imaging and lumbar puncture

Table 4.3. Conditions that can be missed on CT Table 4.4. Headache attributed to spontaneous intracranial
hypotension: clinical and diagnostic characteristics
r Vascular diseases including:
Aneurysms A headache, often but not always positional in nature and
Arteriovenous malformations (especially posterior fossa) usually accompanied by stiffness of the neck or hearing
Subarachnoid hemorrhage abnormalities
Arterial dissections
r CSF pressure of less than 60 mm H2 O (in the sitting position) or
Acute infarcts
r radiologic evidence of a CSF leak
Venous thrombosis
Subdural and epidural hematomas
r Neoplastic diseases including:

Posterior fossa neoplasms Discussion
Meningeal carcinomatosis
Small pituitary tumors The clinical characteristics of headache due to sponta-
r Infections: neous intracranial hypotension are listed in Table 4.4.
Paranasal sinusitis The headache is usually diffuse or dull and often wors-
Brain abscess ens within 15 minutes after standing. It commonly
r Other: remits, sometimes slowly, after the patient lies flat for
Low CSF pressure some time. Neck stiffness, tinnitus, diminished hear-
ing acuity, photophobia or nausea may be reported.
Diagnostically, we like to see that one or more of the
following are present: MRI evidence of low pressure,
radiographic evidence of a CSF leak, OR a CSF open-
shortly after rising each day, and their rapid disappear- ing pressure of less than 60 mm H2 O in the sitting posi-
ance when the patient lay flat. This pattern is most tion. Although this is not the typical position for mea-
consistent with a diagnosis of headache attributed surement of CSF pressure we find that a measurement
to low CSF pressure. Practitioners often assume that obtained in the recumbent position may be normal in a
there must be a history of a prior LP or other instru- low CSF pressure headache. In patients who are known
mentation in order to consider such a diagnosis, but to have had a procedure that might have punctured the
there is ample evidence that spontaneous CSF leaks dura or who have MRI evidence of a low CSF pres-
can develop and lead to low pressure or volume sure headache, an LP to establish a low pressure is not
headache. needed prior to treatment.
Although this diagnosis was considered, it was Even when the history strongly implicates a low
apparently discarded based on the normal CT report, CSF pressure headache, it can be difficult or impos-
the normal CSF opening pressure, and the normal sible to actually detect the leak. In fact some patients
tracer study. However, the clinical presentation is so may not have an actual leak at all: patients who
characteristic of low CSF pressure headache that it is have multiple dural diverticulae in the lumbar region
important to consider the possibility of false-negative may sequester CSF in the diverticulae when they
or inaccurately interpreted test results. The CSF open- are upright, leading to low CSF pressure without an
ing pressure may be normal in low CSF pressure actual leak. This is sometimes termed a “compliance”
headache when it is measured with the patient recum- headache.
bent; taking a pressure reading with the patient in a Table 4.3 lists conditions in addition to low CSF
sitting position will often show the pressure to be low. pressure headaches that may be missed by CT but
Tracer studies are notoriously likely to be negative in which are usually detected by MRI. CT may be
patients who have a slow leak. And finally, CT scan- suboptimal compared to MRI in the evaluation of a
ning is known to be inferior to MRI for demonstrating number of central nervous system conditions; a report
the pachymeningeal enhancement that is seen in many of a normal head CT scan may be reassuring by indi-
(but not all) patients with low CSF pressure headaches cating the absence of a large mass or bleed but does
(Table 4.3). Taken as a whole, the best explanation of little to clarify some of the other conditions listed,
this patient’s headache presentation remains headache including low CSF pressure headache, and should not
due to low CSF pressure – despite his apparently nor- be relied upon to confirm the diagnosis. In many
mal test results. neurologic conditions, including suspected low CSF

Chapter 4: Imaging and lumbar puncture

Table 4.5. MRI features of low pressure headache

r Diffuse pachymeningeal enhancement after gadolinium
r Cerebellar tonsillar descent
r Downward descent (sagging) of the brain/brainstem
r Crowding of the posterior fossa
r Pituitary enlargement
r Bilateral convexity subdural hygroma
r Paraspinal extra-arachnoid CSF collection

spinal MRI. Changes are enhanced with the use of


Headache attributed to spontaneous intracranial

Overreliance on the results of the wrong diagnostic
Figure 4.5 Gadolinium enhanced T1-weighted axial MRI showing study can obscure the clinical picture and lead to pre-
pachymeningeal enhancement. mature, and unwarranted, diagnostic conclusions and
unsuccessful management. Selecting the correct study
for the suspected condition is essential.
pressure headache, the evaluation may be considered
incomplete in the absence of MRI scanning, although Further reading
even MRI can produce false negatives. In one study, the
sensitivity of MRI with and without gadolinium was Radiation risks from scanning
83%. Thus, about a fifth of patients with low CSF pres- Brenner DJ, Hall EJ. Computed tomography – an increasing
sure headaches will have normal MRI scans. source of radiation exposure. N Engl J Med. 2007;357:
In this patient, an MRI scan was ordered 2277–84.
(Figure 4.5) and showed clear evidence of pachy- Mettler FA Jr., Huda W, Yoshizumi TT, Mahesh M.
meningeal enhancement consistent with a low CSF Effective doses in radiology and diagnostic nuclear
pressure headache. He was treated with a 20 mL autol- medicine: a catalog. Radiology. 2008;248:254–63.
ogous epidural blood patch and experienced complete
resolution of his headaches almost immediately after
Does testing relieve anxiety?
the procedure. Fitzpatrick R. Telling patients there is nothing wrong. BMJ.
The characteristic MRI features of low CSF pres-
sure headache (Table 4.5) are likely to be related to the Frishberg, BM. The utility of neuroimaging in the
suspected pathophysiology of the condition; in the set- evaluation of headache patients with normal neurologic
examinations. Neurology. 1994;44:1191–7.
ting of reduced CSF volume and support in an upright
patient, contents of the skull may respond to gravi- Howard L, Wessely S, Leese M, et al. Are investigations
anxiolytic or anxiogenic? A randomised controlled trial
tational force producing the sagging appearance with
of neuroimaging to provide reassurance in chronic daily
crowding of the posterior fossa and reduced size of headache. Neurol Neurosurg Psychiatry. 2005;76:1558–
the cisterns. Downward traction on meningeal struc- 64.
tures has been associated with increased volume of McDonald IG, Daly J, Jelinek VM, Panetta F, Gutman JM.
the epidural space, leading to increased venous vol- Opening Pandora’s box: the unpredictability of
ume, producing the appearance of pachymeningeal reassurance by a normal test result. BMJ. 1996;313:
enhancement, and to increased epidural fluid, lead- 329–32.
ing to hygroma formation. Diagnostic changes likely Petrie KJ, Muller JT, Schirmbeck F, et al. Effect of providing
related to pressure changes may also be apparent on information about normal test results on patients’

Chapter 4: Imaging and lumbar puncture

reassurance: randomized controlled trial. BMJ. 2007;334: CADASIL

352–4. Federico A, Di Donato I, Bianchi S, et al. Hereditary
van Ravesteijn H, van Dijk I, Darmon D, et al. The cerebral small vessel diseases: a review. J Neurol Sci.
reassuring value of diagnostic tests: a systematic review. 2012;322(1–2):25–30.
Patient Educ Couns. 2012;86:3–8. Fukutake T. Cerebral autosomal recessive arteriopathy with
subcortical infarcts and leukoencephalopathy
(CARASIL): from discovery to gene identification.
Defensive medicine J Stroke Cerebrovasc Dis. 2011;20(2):85–93.
Birbeck GL, Gifford DR, Song J, et al. Do malpractice Reimschisel T. Ethical perspectives in neurology. Accessed
concerns, payment mechanisms, and attitudes influence August 29, 2012 at:
test-ordering decisions? Neurology. 2004;62(1): assets/5585.pdf.
DeKay ML, Asch DA. Is the defensive use of diagnostic tests
good for patients, or bad? Med Decis Making. 1998;18(1):
Post-dural puncture headache
19–28. Basurto Ona X, Martı́nez Garcı́a L, Solà I, Bonfill Cosp X.
Drug therapy for treating post-dural puncture headache.
Cochrane Database Syst Rev. 2011;(8):CD007887.
Thunderclap headache Ghaleb A, Khorasani A, Mangar D. Post-dural puncture
Ducros A, Bousser M-G. Thunderclap headache. BMJ. headache. Int J Gen Med. 2012;5:45–51.
2013;346:e8557. Schievink WI, Maya MM, Louy C, Moser FG, Tourje J.
Schwedt, TJ. Thunderclap headaches: a focus on etiology Diagnostic criteria for spontaneous spinal CSF leaks and
and diagnostic evaluation. Headache. 2013;53(3):563– intracranial hypotension. AJNR Am J Neuroradiol.
9. 2008;29:853–6.
Sencakova D, Mokri B, McClelland RL. The efficacy of
epidural blood patch in spontaneous CSF leaks.
White matter lesions in migraine Neurology. 2001;57(10):1921–3.
Agostoni E, Rigamonti A. Migraine and small vessel Sudlow C, Warlow C. Posture and fluids for preventing
diseases. Neurol Sci. 2012;33(Suppl 1):S51–4. post-dural puncture headache. Cochrane Database Syst
Debette S, Markus HS. The clinical importance of white Rev. 2002;(2):CD001790.
matter hyperintensities on brain magnetic resonance Vakharia SB, Thomas PS, Rosenbaum AE, Wasenko JJ,
imaging: systematic review and meta-analysis. BMJ. Fellows DG. Magnetic resonance imaging of
2010;341:c3666. cerebrospinal fluid leak and tamponade effect of blood
Kruit MC, van Buchem MA, Hofman PA, et al. Migraine as patch in post-dural puncture headache. Anesth Analg.
a risk factor for subclinical brain lesions. JAMA. 2004; 1997;84(3):585–90.
Kruit MC, van Buchem MA, Launer LJ, Terwindt GM. Lumbar puncture
Migraine is associated with an increased risk of deep
Barsuk JH, Cohen ER, Caprio T, et al. Simulation-based
white matter lesions, subclinical posterior circulation
education with mastery learning improves residents’
infarcts and brain iron accumulation: the population-
lumbar puncture skills. Neurology. 2012;79(2):132–
based MRI CAMERA study. Cephalalgia. 2010;30(2):
Ellenby MS, Tegtmeyer K, Lai S, Brano DAV. Lumbar
Kurth T, Diener H-C, Buring JE. Migraine and
puncture. N Engl J Med. 2006;355:e1 (video).
cardiovascular disease in women and the role of aspirin:
subgroup analyses in the Women’s Health Study. Straus SE, Thorpe KE, Holroyd-Leduc J. How do I perform
Cephalalgia. 2011;31(10):1106–15. a lumbar puncture and analyze the results to diagnose
bacterial meningitis? JAMA. 2006;296(16):2012–22.
Kurth T, Mohamed S, Maillard P, et al. Headache, migraine,
and structural brain lesions and function: population
based Epidemiology of Vascular Ageing-MRI study. Low CSF pressure headache
BMJ. 2011;342:c7357. Agarwal P, Menon S, Shah R, Singhal BS. Spontaneous
Palm-Meinders IH, Koppen H, Terwindt GM, et al. intracranial hypotension: two cases including one
Structural brain changes in migraine. JAMA. treated with epidural blood patch. Ann Indian Acad
2012;308(18):1889–96. Neurol. 2009;12(3):179–82.

Chapter 4: Imaging and lumbar puncture

Chung SJ, Kim JS, Lee MC. Syndrome of cerebral spinal Schievink WI. Spontaneous spinal cerebrospinal fluid leaks
fluid hypovolemia: clinical and imaging features and and intracranial hypotension. JAMA. 2006;295(19):
outcome. Neurology. 2000;55(9):1321–7. 2286–96.
Rando TA, Fishman RA. Spontaneous intracranial Watanabe A, Horikoshi T, Uchida M, et al. Diagnostic value
hypotension: report of two cases and review of of spinal MR imaging in spontaneous intracranial
the literature. Neurology. 1992;42(3 Pt 1):481– hypotension syndrome. AJNR Am J Neuroradiol. 2009;
7. 30(1):147–51.

Pitfalls in diagnostic testing: blood, urine,

5 and other tests

In addition to imaging studies and lumbar puncture, the need for every test. Testing should be ordered after
a variety of other tests can be useful in the evalua- consideration of the patient’s individual clinical situa-
tion or treatment of patients with headache disorders. tion, rather than ordered as part of a routine workup.
Blood and urine tests are the most commonly ordered
additional tests in the headache clinic, but electrocar- Hormone testing in women with
diograms, electroencephalograms, and other tests may
also be used. Such tests are most helpful in assessing menstrual headaches
the possibility of secondary causes of headache but are
only rarely required for such a diagnosis. Case
Of the hundreds of headache types described in A 36-year-old woman sought treatment because of
the International Classification of Headache Disor- severe headaches associated with her menstrual cycle.
ders (ICHD), only four secondary forms of headache Her description of headaches was consistent with a
require blood or urine tests for a definitive diagno- diagnosis of migraine and her neurologic examina-
sis: urine testing is necessary to make a diagnosis tion was normal. The patient reported that although
of headache due to preeclampsia or eclampsia, while she had headaches at other times of the month, they
blood tests are required to make a definitive diagno- were more likely to occur just before the onset of men-
sis of headache due to human immunodeficiency virus strual bleeding. She was treating headaches with over-
(HIV) infection or hypothyroidism. For a large num- the-counter ibuprofen with little success. A friend had
ber of other disorders, blood or urine testing is help- suggested that her problems might be due to hor-
ful in making a diagnosis but not absolutely required. monal abnormalities. The patient had her estrogen lev-
A diagnosis of giant cell arteritis, for example, can be els checked by a naturopathic physician who told her
based solely on suggestive symptoms in conjunction they were “completely off” and recommended treat-
with a positive temporal artery biopsy. ment with add-back bio-identical estrogen during the
Laboratory testing is frequently needed to monitor perimenstrual period. The patient had tried this for
headache treatment. Lithium and verapamil, the two several months, but treatment was expensive and while
most commonly used preventive treatments for clus- she thought perhaps it had been helpful for her period-
ter headache, require careful monitoring to minimize associated headaches, she still had an average of five or
the chance of adverse events from treatment. Patients six other migraines a month.
on chronic opioid therapy also require regular moni-
toring with blood or urine tests to ensure adherence
to recommended treatment and minimize the likeli-
What is the patient describing and how can
hood of drug diversion or other aberrant drug-related that diagnosis be confirmed?
behavior. For some drugs, monitoring of renal or liver Migraine headaches that occur in predictable rela-
function tests may be advisable. tion to the menstrual cycle are commonly referred
The judicious use of laboratory and other forms of to as “menstrual migraines.” According to diagnostic
testing is an important part of the practice of headache criteria in the appendix of ICHD-3 beta, menstrual
medicine. Because abnormal results on testing can cre- migraine attacks must occur during a five-day win-
ate anxiety and lead to further, potentially harmful, dow of time beginning two days before and extend-
interventions, careful consideration should be given to ing to three days after the onset of vaginal bleeding.

Chapter 5: Blood, urine, and other tests

Furthermore, this must occur on average in two of inherited a migraine-prone nervous system are unusu-
every three cycles. The diagnostic criteria suggest that ally sensitive to normal fluctuations in steroid hor-
these attacks are without aura, and say that headaches mone levels. There is no clinical reason to check serum
occurring with aura are unlikely to be associated with levels of estrogen in women with menstrual migraine.
menstruation. In this case, an emphasis on menstrual-associated
The diagnostic criteria define “pure menstrual headaches meant that the patient’s many non-
migraine” as a situation in which migraine attacks menstrual headaches were ignored. The patient was
occur only during this five-day menstrual window and begun on 10 mg of amitriptyline daily and asked to
at no other time of the month. In contrast, “menstru- use eletriptan 40 mg for up to two headaches a week.
ally related migraine” is diagnosed when headaches At a follow-up visit she reported that her headache
occur predictably during this five-day window but also frequency was down to two headaches a month
at other times of the month. and that those headaches were effectively treated by
Confirmation of a diagnosis of menstrual migraine eletriptan.
requires objective evidence of the timing of headaches
in relation to the patient’s menstrual cycle. In general,
we recommend basing the diagnosis on at least three
months of headache diary records. In our clinical expe- Menstruation is a highly noticeable, culturally signifi-
rience, it is very important to take the time to do this, cant recurrent event. For these reasons it is what soci-
to identify correctly a menstrual trigger for headaches. ologists sometimes refer to as a “magnet explanation.”
Patient self-report of menstrual migraine is not reli- In other words, events that occur in association with
able. One study showed that roughly 20% of women menstruation are highly likely to be attributed to it,
who self-reported a diagnosis of menstrual migraine sometimes mistakenly. This is an example of the cog-
did not actually meet criteria for a diagnosis of men- nitive error or bias known as “illusory correlation,” in
strual migraine when headache diaries were kept. which coincidental events are falsely assumed to be
In this case, the patient’s headache diary showed causally related, often as a result of strongly held prior
migraine attacks during the five-day menstrual win- beliefs.
dow during all three months of recording. The patient In women with frequent migraine attacks,
also had attacks at other times of the month, so she met headaches may occur in conjunction with menstru-
criteria for a diagnosis of menstrually related migraine. ation simply by chance. It is important to verify
even strongly held assumptions about the connection
between headaches and hormonal events, ideally using
Is it useful to check hormone levels in a headache diary. Exclusive emphasis on hormonal
women who have menstrual migraine? causes of headaches may limit attention to other
Good evidence suggests that estrogen withdrawal is explanations for headache or may lead to misguided
the likely cause of menstrual attacks of migraine, attempts at treatment. In the case described in the
although other hormonal changes at this time of the vignette, treatment aimed solely at hormonal causes
menstrual cycle may contribute to migraine suscep- of headache did nothing to help the patient with
tibility. Estradiol levels are high following ovulation. her many nonhormonal headaches. In contrast, all
These high levels are sustained until the late luteal of her headaches were successfully managed using
phase of the cycle, when they fall quickly. A simi- traditional nonhormonal migraine therapies.
lar fall in estrogen levels also occurs and can provoke
headache in susceptible women during the placebo pill Diagnosis
week of some oral contraceptive regimens, following Menstrually related migraine.
childbirth, and with certain types of noncontinuous
hormone replacement regimens.
Many patients with menstrually triggered migraine Tip
attacks wonder whether there is “something wrong” Headache diaries, not blood or urine hormone levels,
with their hormone levels. There is no evidence, how- are most helpful in making a diagnosis of menstrually
ever, that treatable hormonal abnormalities are the related migraine attacks, and are necessary to avoid
cause of menstrual migraine. Rather, women who have falsely attributing headaches to a hormonal cause. A

Chapter 5: Blood, urine, and other tests

hormonal trigger for headaches does not mean that for control of headaches. One recommended titration
treatment must also be hormonal. strategy is to increase the drug by 80 mg a day every
two weeks until headache control is gained or intoler-
Monitoring verapamil treatment in able side effects occur.
These high doses increase the likelihood and
cluster headache severity of side effects, which can include constipa-
tion, peripheral edema, fatigue, and heart block. The
Case patient in the vignette has achieved excellent con-
A 40-year-old man presented to the headache clinic for trol of his cluster headaches with verapamil. Although
follow-up care. He reported daily 90-minute episodes this patient could be experiencing an unrelated or
of 10/10 right-sided retro-orbital pain with right-sided smoking-related respiratory condition, the best option
nasal congestion and ptosis. He had been seen in the in this case before continuing verapamil would be to
clinic ten years earlier and diagnosed with episodic assess the patient for verapamil-related cardiac con-
cluster headache. Treatment with verapamil was effec- duction delay, which may appear even after long peri-
tive and he returned to the care of his personal physi- ods of uneventful use.
cian who continued to prescribe verapamil.
The patient’s cluster headache bouts typically How common are electrocardiographic
occurred for just three months every year but he accu-
mulated a stockpile of verapamil by filling prescrip- abnormalities with verapamil treatment for
tions year-round. When his personal physician retired cluster headache?
he used medication from this stockpile to manage In one study, electrocardiographic abnormalities were
his annual bout of cluster headaches. Through trial detected in about a third of patients treated for cluster
and error he found that a dose of 720 mg daily was headaches with high doses of verapamil. Bradycardia
more effective than the lower doses of medication rec- (a heart rate of ⬍ 60 beats per minute) was the most
ommended by the headache clinic and his personal common abnormality, followed by first-degree heart
physician. block (PR interval ⬎ 0.2 seconds). More severe degrees
At this appointment the patient said that his annual of heart block were rare. In this study, electrocardio-
bout of cluster headaches had begun. His neurologic graphic changes were seen with mean verapamil doses
examination was normal, but on review of systems he of around 1000 mg daily and did not occur with doses
complained of fatigue, constipation, and shortness of lower than 800 mg a day. Although these abnormalities
breath, noting that recently he had barely been able sometimes required cessation of verapamil, in many
to get through his workday. He attributed these symp- other cases they resolved with a decrease in dose. A
toms to cigarette smoking and planned to discuss them reasonable monitoring strategy is to obtain an electro-
when he obtained an appointment with a new primary cardiogram before verapamil treatment is started, and
care physician. He was about to run out of verapamil after each dosage increase or with the appearance of
and requested a refill. symptoms that could have a cardiac cause.
The patient in the vignette had no history of heart
Should this patient’s verapamil prescription disease and an electrocardiogram done two years ago
be refilled? had been normal. However, on examination at this
visit his resting heart rate was 48 beats per minute.
Verapamil is the drug of choice for prevention of An electrocardiogram showed first-degree heart block
cluster headaches, although the US Food and Drug
with a PR interval ⬎0.2 seconds. The dose of verapamil
Administration (FDA) has not approved it for that pur-
was reduced to 80 mg three times daily. The patient’s
pose. In our experience, the dose of verapamil required headaches did not recur and his symptoms of fatigue
to suppress cluster headaches successfully is higher
and shortness of breath resolved.
than the doses that are typically used to treat other
conditions such as hypertension. Most experts begin
with a dose of 240 mg a day, either given as a once- Discussion
daily, sustained release preparation or as 80 mg three Cluster headache is characterized by attacks of severe
times daily. Many patients require an even higher dose or very severe, strictly unilateral orbital, peri-orbital,

Chapter 5: Blood, urine, and other tests

or temporal pain that last from 15 minutes to 3 hours. Diagnosis

During cluster periods or bouts, headaches may occur
Episodic cluster headache and verapamil-induced car-
every other day up to several times a day. Attacks
diac conduction delay.
are associated with ipsilateral cranial autonomic fea-
tures such as tearing, ptosis, miosis, or rhinorrhea. In
episodic cluster headache, cluster bouts can last for Tip
weeks or months and often recur yearly at roughly the High doses of verapamil may be required for success-
same time. Many patients have long periods of remis- ful prevention of cluster headache. Regular electrocar-
sion during which they are entirely free of headaches. diographic monitoring is needed to avoid the adverse
In contrast, patients with chronic cluster headache do effects of cardiac conduction delay.
not experience remissions, or have remissions that are
very short. Urine drug testing in patients
Injectable sumatriptan or inhalation of 100% oxy-
gen are effective treatment for individual attacks of taking opioids
cluster headache, but their use on a daily or near-
daily basis is impractical. Additionally, while wait- Case
ing for them to take effect patients will experi- A 37-year-old woman returned to the headache clinic
ence a short period of incapacitating pain, which for a follow-up visit. The patient had severe post-
can severely disrupt daily life activities. Thus, essen- traumatic head and face pain following a gunshot
tially all patients with cluster headache should also wound. The pain had not responded to treatment with
receive preventive treatment, which is given daily dur- gabapentin, amitriptyline, or numerous other medica-
ing a cluster bout to decrease or eliminate cluster tions for neuropathic pain. Because it was relieved by
attacks. opioid medications, the patient had been started on
Although verapamil is the first-line preventive sustained release morphine. Unfortunately, intermit-
treatment for cluster headache, other preventive treat- tent bouts of vomiting from pancreatitis made it dif-
ments including lithium, melatonin, or the use of ficult to use oral medications. At her last visit the oral
occipital nerve blocks can be tried. Successful use of morphine was discontinued and she was started on a
verapamil may require titration to fairly high doses; a fentanyl patch. The patient reported that her pain was
common reason for apparent failure of this treatment well controlled on the fentanyl. Along with pancre-
is the use of inadequate doses. In many instances doses atic enzymes and a multivitamin she was using cariso-
over 360 mg daily are required up to and including the prodol intermittently for episodes of muscle spasm.
recommended maximum daily dose of 960 mg, doses Because her usual physician was on maternity
that are far higher than those used in the management leave, the patient was seen by a resident who was rotat-
of cardiac conditions. ing in the clinic. After consultation with a supervis-
Once successful control of cluster headaches is ing physician, the resident refilled the patient’s fen-
achieved, many patients are reluctant to discontinue tanyl prescription. He noticed that at her last visit a
preventive treatment. Some patients with episodic pain treatment urine drug screen had been obtained,
cluster headache prefer to stay on preventive treat- so before the patient left he repeated this test. Later that
ment such as verapamil year-round even if they have day the resident reviewed the test results. He was star-
attacks for only three or four months a year. While this tled to see that the urine drug test results did not men-
preference is understandable in view of the severity tion the presence of fentanyl but instead reported that
of cluster headache pain, we discourage this practice. meprobamate, a tranquilizer, had been detected.
Instead, we recommend that patients gradually lower
the dose of preventive medication as they approach the What are the possible explanations for these
expected time when their cluster bout should be end-
ing. If headaches recur when the dose is lowered, a urine findings?
higher dose can be resumed for several more weeks, Fentanyl is not reliably detected by urine or blood drug
when another attempt should be made to taper the testing, making it challenging to monitor the long-
drug. In our experience, this is an often-overlooked term use of this drug. Urine drug testing is still use-
aspect of managing cluster headaches. ful, though, in order to monitor the patient’s possible

Chapter 5: Blood, urine, and other tests

use of other nonprescribed substances. In this case, oid therapy it is important to obtain informed con-
the finding of meprobamate in the urine might be sent from the patient for treatment and to outline in
misinterpreted as a sign that the patient is taking a detail the rules of your practice about opioid treat-
nonprescribed, possibily illicitly obtained, substance. ment and monitoring. Most experts recommend the
However, carisoprodol is metabolized to meprobamate use of “treatment contracts” that outline the conditions
and thus the presence of meprobamate in this patient’s under which the medication will be prescribed, the
urine sample is not unexpected. It is important to be methods that will be used to monitor the treatment,
aware of substances that can produce a false-positive and the circumstances under which lost prescriptions
urine drug test result. It is advisable to consult with will be replaced or therapy discontinued. Sample treat-
the laboratory that you use to process the samples, to ment contracts can easily be located online. These
learn what tests they use, the limits of detection, and treatment contracts do not have legal force but do
any other unique characteristics of the laboratory. stand as a clear record of the understanding you had
If this patient’s urine sample had contained evi- with the patient about treatment. They are valuable
dence of tetrahydrocannabinol (THC), an indication if questions or disputes later arise about the patient’s
of marijuana use, the physician would have had more treatment.
reason to be concerned. THC use is correlated with the Appropriate selection of patients for chronic opioid
use of other illicit substances and some studies sug- therapy may help reduce the risks of diversion or mis-
gest it is an indicator of high risk for abuse of opi- use, although there is no strong evidence about this. A
oids. An exception to this would be a patient who number of tools can be used to identify patients who
lives in a state where the use of “medical marijuana” is are at high risk of aberrant drug use, including the opi-
allowed. Such patients should have received the drug oid risk tool, the Pain Medication Questionnaire, and
through approved channels, however, and should have others.
informed the physician about its use prior to drug
How else might this patient’s use of fentanyl The use of chronic opioid therapy in patients with non-
malignant pain syndromes is controversial. Its long-
be monitored? term efficacy is uncertain, and the benefits and harms
The purposes of monitoring in patients on chronic opi- are often finely balanced. Risks of chronic opioid ther-
oid therapy are two-fold: (1) to ensure the patient is apy include the development of addiction or depen-
benefitting from treatment; and (2) to detect prob- dence syndromes, opioid-induced hyperalgesia, and
lems with drug diversion or misuse. The benefits of the potential for diversion or misuse of prescribed
treatment can be established through pain ratings that opioids. These problems have received a great deal of
show meaningful reductions in pain intensity related attention from the popular media, politicians, and reg-
to opioid treatment. Functional improvement is also ulators. For all of these reasons, patients on chronic
important. Many physicians find it useful to mon- opioid therapy must be closely monitored, with regu-
itor such things as return to work, reduced use of lar office visits and efforts to minimize the misuse or
breakthrough pain medications, and fewer emergency diversion of such drugs.
department visits for poorly controlled pain. When Urine or blood drug testing is nearly always an
urine or blood tests cannot reliably detect an opioid, important part of a monitoring plan for patients on
as is the case in this vignette, other ways of monitor- opioid maintenance therapy. In most cases, we recom-
ing compliance with long-term opioid therapy can be mend a baseline urine drug test before starting therapy,
used. For example, the patient’s skin can be inspected as well as regular random tests. The frequency of such
at each visit to ascertain that the expected patches testing depends on the patient’s circumstances. In gen-
are in place. Patients can also be asked to bring their eral, high-risk patients should be monitored more fre-
unused patches along to appointments to show that quently, possibly at every visit. For long-term patients
these are in their possession and to ensure that the who appear to be stable and at low risk of diver-
proper number of patches remains. sion or misuse, we still recommend random testing
Patients may perceive these forms of monitoring as once or twice a year. One study of headache patients
burdensome or intrusive. Before starting chronic opi- receiving opioid maintenance therapy showed that

Chapter 5: Blood, urine, and other tests

unexpected urine drug test results suggesting aberrant The headache specialist obtained an electrocardio-
drug-related behavior occurred in a sizeable propor- gram, which showed a PR interval of 0.2 seconds. She
tion of apparently stable patients, sometimes even after asked the patient to decrease his verapamil dose to
long periods of trouble-free use. 240 mg/day, and suggested that he add lithium car-
Patients can be referred to independent laborato- bonate, 300 mg twice a day, to his treatment regi-
ries for such testing, but many physicians obtain the men. She also recommended subcutaneous sumatrip-
urine samples in their own offices. These samples are tan in place of oral zolmitriptan to treat individual
not usually handled with the same strict “chain of cus- attacks of headache. She asked the patient to return a
tody” procedures applied in professional laboratories. week later to have his lithium level and an electrocar-
diogram checked. The electrocardiogram showed the
PR interval had returned to normal, but the patient’s
Diagnosis lithium level was low at 0.3 mmol/L (normal range
Post-traumatic head and facial pain, treated with for this laboratory was 0.5–1.3 mmol/L). The doc-
chronic opioid therapy. tor asked the patient to increase his lithium dose to
300 mg of lithium carbonate three times daily. When
checked a month later, his lithium level was still low at
Tip 0.3 mmol/L and his headaches were no better. The doc-
Urine or blood drug testing does not reliably detect tor increased the lithium dose again, but a week later a
some opioids, such as fentanyl, but such tests are still third lithium level was 0.2 mmol/L.
valuable to detect use of other nonprescribed sub-
stances. Knowledge of cross-reactive substances that
can produce false-positive urine drug tests is impor-
tant to avoid inaccurately accusing patients of non- What are the possible explanations for this
compliance with treatment standards. patient’s lithium level?
This patient’s lithium level was low when he began
Lithium levels in a patient with taking lithium and remained essentially unchanged
despite two increases in the prescribed dose of lithium.
cluster headache It is unlikely that these multiple test results were
incorrect. Although lithium can interact with a num-
Case ber of medications, this patient’s medication regimen
A 45-year-old man presented to the headache clinic was otherwise unchanged during the period when his
with a ten-year history of right retro-orbital attacks of lithium dosage had been increased. Thus, the most
10/10 pain. These lasted 45 minutes to an hour and likely explanation for these consistently low levels was
were associated with right-sided ptosis, rhinorrhea, that the patient was nonadherent with the prescribed
and tearing of the right eye. During attacks the patient dose of lithium.
was restless and sometimes banged his head on the There are many reasons for nonadherence to drug
wall. Initially these attacks occurred infrequently and therapy, which are discussed below. In this case, the
he was diagnosed with a possible sinus problem. He patient’s physician approached the matter in a non-
underwent sinus surgery with no improvement, and judgmental way, since she knew that patients are often
attacks began to increase in frequency. He was referred reluctant to admit they have not adhered to ther-
to a neurologist who diagnosed cluster headache and apeutic recommendations. She told the patient that
treated him with verapamil after he failed to respond to his lithium level had remained unchanged despite the
a course of oral steroid therapy. Over the last year and recommended increases in dose; she then “normal-
a half, the headache attacks had become daily despite ized” the behavior of nonadherence by remarking that
continued treatment with verapamil, which had been “many patients find it difficult to take their medicine
increased to 320 mg/day. The patient was also using regularly” and asking the patient if that was true for
oral zolmitriptan and hydrocodone to treat individual him. The patient appeared relieved and responded that
attacks of headache, but reported they were not always his busy workday made it difficult to remember his
effective. morning dose of medication, and that he sometimes

Chapter 5: Blood, urine, and other tests

forgot to take his evening dose because of the demands lithium would have been inaccurately regarded as inef-
of his three children when he returned home at night. fective for his cluster headache.
The patient and physician then strategized about
ways in which he might remember to take his medica- Discussion
tion, and the patient decided to program an alarm on The FDA has not approved any drugs for preven-
his cellphone. The doctor also reviewed the expected tive treatment of cluster headache, but verapamil and
benefits of lithium and the fact that alternative treat- lithium are widely used off-label. Verapamil is the pre-
ments that might be used in its place had less evidence ferred initial drug because it seems to have a quicker
of benefit. Both she and the patient were pleased when onset of action and fewer side effects. Lithium is used
at his next visit his headaches were greatly diminished alone when verapamil is ineffective or not tolerated, or
and his lithium level had increased to 0.8 mmol/L. as an add-on to verapamil when combination therapy
is needed. The scientific evidence supporting its use is
modest, but clinically it is perceived to be quite effec-
How common is nonadherence to drug tive, particularly in chronic cluster headache. A typi-
therapy in headache patients and why does cal starting dose is 300 mg of lithium carbonate given
it occur? twice a day, which can be increased in 300 mg incre-
ments based on treatment response and blood levels.
Nonadherence or poor compliance to recommended
Most patients who respond to the drug do so at doses
treatment is an important problem with all types of
between 600 and 1200 mg/day.
medical treatment, particularly with complex treat-
Nuisance side effects such as tremor, nausea, or
ment regimens for chronic illnesses. There is no evi-
diarrhea are relatively common, and at higher doses
dence that nonadherence is more common among
lithium toxicity can occur. This can produce kidney
headache patients than the general medical popula-
dysfunction, ataxia, or convulsions that may be fatal.
tion. Failure to adhere to recommended treatment can
Sodium depletion, which can occur with the use of
take the form of overuse of prescribed or over-the-
some diuretics, may increase the likelihood of lithium
counter treatments, or underuse, which may consist of
toxicity. Nonsteroidal anti-inflammatory medications
never filling a prescription at all, using less than the
may increase lithium levels. Verapamil has been
recommended dose, or premature cessation of treat-
reported both to increase and decrease lithium levels.
ment. One study of headache patients found that only
Because lithium has a narrow therapeutic win-
slightly more than a third of patients were fully com-
dow, periodic monitoring of blood levels is necessary
pliant with their prescribed treatment regimens.
to avoid toxicity. Levels should be checked after any
There are many reasons for nonadherence to pre-
dosage increase, when symptoms related to toxicity
scribed therapy. One study found that medication
occur, or when other medications that might affect
costs, patient uncertainty that the drug was indicated,
lithium levels are started or stopped. Blood levels do
or fears of side effects were the most common reasons
not, however, correlate with response to the drug, so
for nonadherence.
there is no “target” blood level. In patients on long-
Poor adherence is also related to the complexity
term lithium therapy most physicians also check yearly
of the treatment regimen. The more frequently a drug
kidney function and thyroid function tests.
must be taken, for example, the less likely patients are
to adhere to the regimen. When feasible, switching to
once-daily treatments can improve the likelihood of
adherence to drug regimens. Nonadherence to pre- Chronic cluster headache with nonadherence to
scribed medications is probably an underappreciated lithium treatment.
reason for the apparent failure of headache treatment
regimens. In the case described in this vignette, the Tip
patient happened to be taking one of the few headache Blood levels should be periodically monitored in
drugs that require monitoring of serum drug levels. patients taking lithium. Although principally intended
If blood tests had not been obtained, it is likely that to avoid toxicity, blood levels may also be used to assess
his nonadherence would have gone undetected and adherence to the prescribed treatment regimen.

Chapter 5: Blood, urine, and other tests

A young man with body aches, rash, HIV infection. In our experience, such patients do
not usually come to the headache clinic but are more
and headache likely to present in an emergency department or urgent
care setting, given the abrupt nature and relatively
Case short duration of symptoms. The reason for headache
A 24-year-old man sought treatment in the emergency or other symptoms is not fully understood, but may
department for a chronic headache. He described involve release of cytokines.
the headache as dull and generalized. He also com- A more chronic form of headache has been
plained of body aches and sensitivity to light. He described in people with stable HIV infection. Once
was afebrile and his physical examination was nor- superimposed infections related to immunosuppres-
mal with the exception of a macular red rash over the sion have been eliminated as a cause of headache, a
torso and some cervical lymphadenopathy. The emer- diagnosis of headache attributed to human immuno-
gency department physician made a diagnosis of pos- deficiency syndrome/acquired immunodeficiency
sible mononucleosis and sent the patient home with syndrome (HIV/AIDS) can be considered. This diag-
instructions to see his primary care physician if the nosis is contained in the appendix of ICHD-3 beta
symptoms did not remit. rather than the main body of the classification because
A week later the patient saw his primary care doc- the validity of the diagnostic criteria has not been
tor because of continued headache and body aches. firmly established. This is largely due to the fact that it
The doctor asked the patient whether he had other can be very challenging to distinguish the low-grade
reasons for seeking treatment and what he thought headache likely to be attributable to HIV from other
might be the cause of his problems. The patient con- forms of primary headache commonly reported by
fided that his new girlfriend had been an intravenous HIV patients. Additionally, some antiretroviral drugs
drug abuser; over the last few months they had had can produce headache.
many unsafe sexual encounters and he worried that Nonetheless, most experts believe that low-grade
he might have “some kind of infection.” The physi- headache is a common feature of HIV infection. It
cian recognized that the patient was at high risk for a is typically a dull, bilateral headache. This diagno-
variety of sexually transmitted infections and ordered sis should only be made after blood testing confirms
appropriate testing, including an HIV antibody test infection with HIV and/or immunodeficiency, in con-
and an HIV RNA (viral load) test. The HIV anti- junction with demonstration of disease-related patho-
body test was positive and the patient’s viral load was physiology “likely to cause headache.” Blood or oral
⬎ 100 000 copies/mL. fluid samples can be tested with rapid tests but any
positive results require confirmation using immuno-
Can headache be a symptom of assay procedures. Because antibody tests may be nega-
tive early in the infectious process, an HIV RNA (viral
HIV infection? load) test should also be ordered in settings where
This case is a reminder that a half to two-thirds of peo- infection may have been recent.
ple with newly acquired HIV infection will experience
a constellation of signs and symptoms as they sero- What are some other possible causes of
convert, usually about two to eight weeks following
infection. Headache may be a prominent symptom of headache in people with HIV infection?
this influenza-like event, which is often termed “acute A high index of suspicion for secondary causes of
retroviral syndrome” and lasts an average of about a headache is appropriate in HIV-positive patients, par-
month. In one series of 218 patients with acute HIV ticularly when immunosuppression is severe. A large
infection, headache was reported by 51%. Only fever number of opportunistic infections such as Crypto-
(77%), lethargy (66%), and skin rash (56%) were more coccus or toxoplasmosis can produce headache, and
common. Other neurologic symptoms of photophobia patients with HIV also have an increased risk of central
and meningismus (stiff neck) were reported by 12% nervous system neoplasms.
and 9% of patients, respectively, making headache by The degree of immunosuppression is positively
far the most common neurologic symptom of acute correlated with the likelihood of these secondary

Chapter 5: Blood, urine, and other tests

causes of headache, and can be useful in deciding illness who have risk factors for sexually transmitted
whether or not an HIV-positive patient with headache diseases. To maximize the chance of detecting recent
needs a brain scan. One study found that impor- infection, both an HIV antibody test and an HIV RNA
tant abnormalities on CT scans of the brain were far (viral load) test should be ordered.
more likely in patients whose CD4 counts were below
200 cells/␮L, and recommended this as a threshold
below which imaging should be considered.
A young woman with headache and
multiple laboratory abnormalities
Discussion Case
A high degree of suspicion should be maintained for
A 19-year-old woman was seen in the headache clinic
the possibility of acute HIV infection in people with
on referral from her primary care physician. She
headache and risk factors for sexually transmitted dis-
had developed headaches a few months earlier. These
eases. Individuals with acute HIV infection may have
occurred on average twice a month and were very
very high viral loads and be particularly infectious
severe. She described them as pounding and usually
to others. Following the period of symptoms during
located over her right temple. These headaches could
seroconversion, infected people may experience a long
last up to two days at a time and were associated with
interval during which they have relatively few signs
severe nausea and vomiting and sensitivity to noise.
or symptoms, and diagnosis may be delayed. Thus the
She was unable to function during the headaches and
correct recognition of the acute retroviral syndrome
retreated to her darkened bedroom, although between
may allow early diagnosis of a potentially fatal illness
headaches she was active and well.
and the opportunity to minimize the risk of transmis-
The patient’s primary care physician thought that
sion to others. Unfortunately, because the symptoms of
migraine was the most likely diagnosis. However, the
acute retroviral syndrome are variable and nonspecific,
patient’s parents were very concerned and insisted on
this opportunity is often missed.
a “complete workup” to rule out dangerous causes of
In the case described in the vignette, the emer-
headache. A CT scan of the brain was normal. A large
gency department physician missed an opportunity to
number of blood tests had also been ordered, and sev-
identify the patient’s HIV infection, but fortunately
eral of these had been abnormal. Her antinuclear anti-
the patient again sought treatment. His primary care
body (ANA) test was positive at 1:80 so the patient
physician correctly recognized that the patient had risk
had been seen by a rheumatology consultant, who after
factors for HIV infection through his exposure to a for-
extensive further testing did not find evidence that the
mer intravenous drug abuser. The prevalence of HIV is
patient had systemic lupus erythematosus (SLE) or any
high in intravenous drug users, and their sexual part-
other rheumatic disease. Her thyroid-stimulating hor-
ners are at risk of infection with this and other blood-
mone (TSH) level was also slightly elevated, and the
borne infections such as hepatitis. A recent study sug-
patient had been seen by an endocrinologist who, after
gests that a short course of antiretroviral treatment
further testing, did not find any evidence of thyroid
early in HIV infection may delay disease progression,
dysfunction, but did note a small 1 cm thyroid nodule,
with important benefits for individual patients and
which proved to be benign on fine-needle aspiration.
public health, making it even more important to detect
infection as early as possible.
Was it a mistake to order all of these tests?
Diagnosis In a word, yes. Laboratory testing should be limited to
Headache related to acute retroviral syndrome (HIV patients who have a reasonable pre-test probability of
seroconversion). the disease the test is meant to detect. The physician
in this case meant well, but both he and the patient’s
parents failed to appreciate one of the common pitfalls
Tip of diagnostic testing, which is that testing should be
Acute retroviral syndrome should be considered in ordered based on the clinical context. In trying to
patients presenting with headache and influenza-like make sense of this concept, it is helpful to understand

Chapter 5: Blood, urine, and other tests

the concepts of sensitivity, specificity, and positive inations were normal. Thus, the likelihood of a sinis-
and negative predictive values. Together, these are ter cause of headache was very low. She had no other
the operating characteristics of a test. Sensitivity is the signs or symptoms that might suggest thyroid abnor-
proportion of patients with a disease who have a pos- malities or an autoimmune disorder or any of the other
itive test result. Specificity is the proportion of patients diseases for which she was tested. Specifically, she had
without a disease who have a negative test result. The not experienced recent fatigue, constipation, skin rash,
predictive value of a test is the likelihood that a patient or any other systemic indicators of underlying disease.
does or does not have a disease given a particular The extensive testing that was ordered was not done
test result. Positive predictive value is the likelihood because of any suspicion that a particular disease was
that a patient with a positive test result really has the present, but rather to appease the patient’s parents and
disease in question, and negative predictive value is ease their anxiety about a possible dangerous cause of
the likelihood that a patient with a negative test result headache.
really is disease-free. Although blood tests are not invasive and are com-
The sensitivity and specificity of a test do not vary monly perceived as harmless, a good case can be made
from patient to patient, but the positive and nega- that indiscriminate testing harmed the patient in this
tive predictive values of a test can differ dramatically vignette. Not only was attention to her presenting
depending on the prevalence of the disease in the problem – migraine – delayed by additional fruitless
population that is being tested. In the patient in the investigation, she was exposed to radiation from a
vignette, the likelihood that she had lupus was low – CT scan of the head and underwent invasive aspira-
for the purposes of demonstration, let’s assume it was tion needle biopsy for an incidental thyroid nodule.
on the order of 1%. In other words, assume that in This sort of additional investigation triggered by an
100 otherwise healthy young women like our patient abnormal finding is often referred to as the “medical
who have intermittent severe headaches, one truly has cascade.” Additional invasive diagnostic procedures
SLE. If an ANA test for SLE has a sensitivity of 100% always carry a risk of harm, and certainly the wait for
and a specificity of 90%, the positive predictive value results creates anxiety. These harms and the costs of
of the test is only 9%. In other words, there was still additional testing may be low for an individual patient,
a 91% chance the patient did not have SLE even with but if such unnecessary testing occurs on a widespread
a positive test. (If you are interested in the details of scale the iatrogenic harms and costs can be substantial
this calculation, positive predictive value is calculated on a population level. Ordering a large number of tests
by dividing the true-positive rate by the true-positive in the absence of specific indications for testing does
plus false-positive rate; in other words, 100%/[100% + not represent high value medical care.
10%], which equals roughly 9%.) The bottom line is Patients (and sometimes physicians) commonly
that when the prior probability of disease is low, the overestimate the value of laboratory or other forms of
positive predictive value of a test is poor. Addition- testing and have unrealistic ideas about the accuracy of
ally, when a large number of tests are ordered, some many tests. No test is perfect: not all positive or abnor-
are likely to be abnormal on the basis of chance alone. mal test results indicate disease – instead, some are
false positives. Similarly, not all negative or normal test
results mean that a patient does not have the disease
Discussion in question – these are false negatives. In situations
Laboratory and other tests are most helpful when where the likelihood of disease is low, as in the patient
ordered based on the clinical context of an individual in this vignette, positive results on a test are more likely
patient’s situation. They should be ordered selectively to be false positives than true indicators of disease. In
and interpreted very cautiously. It is difficult to defend other words, their positive predictive value (the likeli-
the practice of routinely ordering a large panel of tests hood that a patient with a positive test actually has the
for all patients with a particular problem without con- disease in question) is low.
sideration of the individual clinical situation.
The incidence (new onset of disease) of migraine
is very high in young women, and this patient’s Diagnosis
headaches met criteria for that disorder. She was oth- Migraine and iatrogenic harm from unnecessary
erwise healthy and her physical and neurologic exam- testing.

Chapter 5: Blood, urine, and other tests

Tip May A, Leone M, Afra J, et al. EFNS guidelines on the

treatment of cluster headache and other trigeminal-
A shotgun approach to testing in patients with autonomic cephalalgias. Eur J Neurol. 2006;13:1066–
headache is inadvisable. In settings where the prior 77.
probability of a particular disorder is low, most Rains JC, Penzien DB, Lipchik GL. Behavioral facilitation of
abnormal test results will be false positives, evalua- medical treatment for headache – Part II: Theoretical
tion of which can lead to unnecessary anxiety and models and behavioral strategies for improving
possible harm from unneeded additional testing or adherence. Headache. 2006;46:1395–403.
interventions – the so-called “medical cascade.” Tfeldt-Hansen PC, Jensen RH. Management of cluster
headache. CNS Drugs. 2012;26(7):571–80.
Further reading
Acute retroviral syndrome
Menstrual migraine
Cohen HS, Gay CL, Busch MP, Hecht FM. The detection of
Loder E, Rizzoli P, Golub J. Hormonal management of
acute HIV infection. J Infect Dis. 2010;202(Suppl 2):
migraine associated with menses and the menopause: a
clinical review. Headache. 2007;47(2):329–40.
Graham CB, Wippold FJ, Pilgram TK, Fisher EJ, Smoker
Marcus DA, Bernstein CD, Sullivan EA, Rudy TE. A
WRK. Screening CT of the brain determined by CD4
prospective comparison between ICHD-II and
count in HIV-positive patients presenting with
probability menstrual migraine diagnostic criteria.
headache. AJNR Am J Neuroradiol. 2000;21:451–4.
Headache. 2010;50:539–50.
The SPARTAC Trial Investigators. Short-course
Somerville BW. The role of estradiol withdrawal in the
antiretroviral therapy in primary HIV infection.
etiology of menstrual migraine. Neurology. 1972;
N Engl J Med. 2013;368:207–17.
Vanhems P, Allard R, Cooper DA, et al. Acute human
Verapamil treatment of cluster headache immunodeficiency virus type 1 disease as a
mononucleosis-like illness: is the diagnosis too
Cohen AS, Matharu MS, Goadsby PJ. Electrocardiographic restrictive? Clin Infect Dis. 1997;24:965–70.
abnormalities in patients with cluster headache on
verapamil therapy. Neurology. 2007;69:668–75. Vanhems P, Dassa C, Lambert J, et al. Comprehensive
classification of symptoms and signs reported among
Leone M, D’Amico D, Frediani F, et al. Verapamil in the 218 patients with acute HIV-1 infection. J Acquir
prophylaxis of episodic cluster headache: a double-blind Immune Defic Syndr. 1999;21:99–106.
study versus placebo. Neurology. 2000;54:1382–5.
Nesbitt AD, Goadsby PJ. Clinical review: cluster headache.
BMJ. 2012;344:e2407. Iatrogenic harms of unnecessary testing
Gough J, Scott-Coombes D, Palazzo FF. Thyroid
Chronic opioid therapy incidentaloma: an evidence-based assessment of
Manchikanti L, Abdi S, Atluri S, et al. American Society of management strategy. World J Surg. 2008;32(7):
Interventional Pain Physicians (ASIPP) guidelines for 1264–8.
responsible opioid prescribing in chronic non-cancer Loder E, Cardona L. Evaluation for secondary causes of
pain, Parts 1 and 2. Pain Physician. 2012;15:S1–100. headache: the role of blood and urine testing. Headache.
Owen GT, Burton AW, Schade CM, Passik S. Urine drug 2011;51:338–45.
testing: current recommendations and best practices. Qaseem A, Alguire P, Dallas P, et al. Appropriate use of
Pain Physician. 2012;15:ES119–33. screening and diagnostic tests to foster high-value,
cost-conscious care. Ann Intern Med. 2012;156(2):
Lithium for cluster headache and nonadherence to 147–9.
medication Ulvestad E, Kanestrøm A, Madland TM, et al. Evaluation
of diagnostic tests for antinuclear antibodies in
Hedenrud T, Jonsson P, Linde M. Beliefs about medicines
rheumatological practice. Scand J Immunol.
and adherence among Swedish migraineurs. Ann
Pharmacother. 2007;42:39–45.

When historical or examination findings are

6 missed or misinterpreted

There are no biomarkers for benign headache disor- tic for the patient and instructive for the physician.
ders such as migraine, and laboratory or imaging stud- Many patients will spontaneously describe features
ies serve mainly to exclude other causes of headache. such as photophobia or vomiting that are required for
Thorough patient evaluation, especially a careful his- headache diagnosis; the physician can query missing
tory and physical examination, is therefore essen- information later.
tial for the accurate diagnosis of headache problems. It is important to note that although most patients
Headache diagnosis has important treatment impli- should be allowed to tell their story in an unfettered
cations, since management that is effective for one way, some patients may require different degrees of
form of headache may be ineffective for another. Over- active guidance from the provider. This can help retain
looked or misinterpreted historical or examination a focus on the most salient details, the result being a
findings can result in missed diagnoses and lost oppor- narrative crafted from a balance of conversational flow
tunities for appropriate therapy. One common exam- and what could be termed “supportive” redirection on
ple is a patient with cluster headache who has been the part of the provider: for example, “Let’s put your
misdiagnosed with migraine and treated ineffectively visit to the emergency department aside for now and
with anti-migraine drugs for years. go back to what you were saying about your headache.”
Most seasoned headache experts recognize that the With only some practice and without any special
patient history is usually more informative than the or expensive technology, the provider can, after an
physical examination. In fact, the physical examina- effective history and physical examination, expect to
tion is expected to be normal in most patients with arrive at a correct headache diagnosis much of the
primary headache disorders. Despite this, additional time. When the process breaks down or shortcuts are
findings such as occipito-nuchal tenderness, abnormal taken, various complications can occur, as illustrated
neck posture, or temporomandibular joint dysfunc- by the following cases.
tion can identify problems that might be contribut-
ing to a primary headache disorder. Documentation A young man with visual changes and
of these findings also can be helpful to patients for
insurance reasons or disability determination. The fact a sore head
that physical or neurologic abnormalities are uncom-
monly encountered in headache patients may lull doc- Case
tors into a false sense of security. Thorough examina- A 28-year-old man with a prior history of hypertension
tion of all patients with headache is important since, was seen in the emergency department with a com-
as cases in this chapter illustrate, subtle findings can plaint of sudden visual changes. The patient described
change a diagnosis and influence treatment. a “flashing” in his peripheral vision that started in one
Considerable skill is sometimes required to obtain eye and spread to both over a period of about ten min-
a thorough history of headache problems, especially utes. This gradually evolved into what he described as
in cases of long-standing headache with many failed a “flashing floater” along the edges of his vision bilat-
treatment attempts. We find it helpful to begin with the erally. This was followed by blurred vision in the orig-
open-ended question “Tell me about your headaches.” inally affected eye.
This allows the patient uninterrupted time to tell The visual symptoms, which had lasted a total of
their story. The resulting narrative is often therapeu- 20 minutes, gradually disappeared while the patient

Chapter 6: Missed historical features

Table 6.1. Features suggestive of retinal detachment Table 6.2. Selected features that distinguish the visual aura of
migraine from ischemic events
Acute visual loss (this is an ophthalmologic emergency)
Visual aura of migraine Ischemic events
Recent onset of visual “floaters”
Visual phenomena begin in Sudden onset of visual loss
Sudden onset of flashes of light
the periphery and develop
Darkening of vision or development of a curtain or shadow gradually
across vision
Both positive and negative Negative visual symptoms
visual phenomena are present predominate
Reversible, with duration of Short duration (typically no
event no longer than an hour more than 15 minutes)
was on the way to the emergency department. There
was no associated headache. The emergency depart- If other aura symptoms are If other neurologic symptoms
present, they occur in are present, they occur
ment physicians were concerned about the possibil- sequence simultaneously with visual
ity of a transient ischemic attack and he underwent phenomena
an evaluation that included a carotid ultrasound, a CT Similar events on multiple No prior history of similar
scan of his head, and consultation by an ophthalmol- occasions events
ogy resident. The results of these investigations were Symptoms are followed by Headache is not a common
normal and he was discharged home with instructions headache (usually within associated feature
15 minutes)
to take 81 mg of aspirin daily and follow up in the out-
patient neurology clinic.
He was seen in the clinic several days later where
a neurologist elicited a history of several similar sient ischemic attacks from more benign causes of
episodes of visual disturbance when the patient was in visual symptoms such as migraine aura, but several
his teens. Those were described as periods of “fluttery” historical features, which are listed in Table 6.2, can
peripheral vision lasting 5–25 minutes and followed by help.
a period of mild head “soreness.” These episodes had The patient described in the case experienced a
occurred sporadically over a period of several years mixture of positive and negative visual features that
and then disappeared. The patient had sought ophthal- developed gradually and were followed by headache.
mologic examination at the time but no abnormalities These are all more characteristic of visual aura than of
were found. transient ischemic attack. Patients often find it difficult
to describe visual aura. Obtaining a previous history
of such attacks can be very helpful. We make a spe-
What is the differential diagnosis of these cial point of inquiring about such attacks at a younger
visual symptoms? age, since in our experience patients may not sponta-
Isolated visual symptoms can be related to intraoc- neously report them.
ular pathology, transient ischemic attacks, migraine
aura, retinal detachment, or seizures. Visual phenom- What bedside test may have helped clarify
ena are often described as “positive” or “negative.” Pos-
itive visual phenomena include false visual images or the diagnosis?
distortions of normal vision, such as sparkling lights, Almost all patients who have any type of migraine
lines, or geometric shapes, spots of color, or shimmer- aura also have visual aura. Thus, in order to recog-
ing vision. Negative visual phenomena, on the other nize patients with aura, it is only necessary to establish
hand, include visual deficits such as dark or blind the presence of visual aura. This can be done using the
spots. Visual Aura Rating Scale (VARS), which is a validated
The differential diagnosis for visual symptoms tool developed for this purpose. The point system used
that resolve spontaneously includes seizures, migraine by the VARS is described in Table 6.3. The VARS is sim-
aura, and transient ischemic attacks. Symptoms that ple to administer at the bedside or in the office. Points
persist beyond an hour or two require urgent investiga- are assigned for the presence of any of five character-
tion for intraocular pathology such as retinal detach- istics of visual aura: duration from 5 to 60 minutes,
ment; features characteristic of this diagnosis are listed gradual onset, visual scotoma, zig-zag lines, or uni-
in Table 6.1. It can be difficult to distinguish tran- lateral location of the visual phenomena. The patient

Chapter 6: Missed historical features

Table 6.3. The Visual Aura Rating Scale for the diagnosis of Table 6.4. Clinical characteristics of aura
migraine with aura
One of the following features should be present:
Feature Points 1. Fully reversible (usually homonymous) visual symptoms
including positive features (e.g. flickering lights, spots, or lines)
Duration 5–60 minutes 3 points
and/or negative features (e.g. loss of vision)
Develops gradually ࣙ 5 minutes 2 points 2. Fully reversible (usually unilateral) sensory symptoms including
positive features (e.g. pins and needles) and/or negative
Presence of associated scotoma (blind spot) 2 points features (e.g. numbness)
Zig-zag lines (fortification spectra) 2 points 3. Fully reversible dysphasic speech disturbance
Unilateral location 1 point
A VARS score of 5 or more diagnosed migraine with aura with a
sensitivity of 96% (95% confidence interval [CI] 92–99%) and a the evolution of visual or sensory symptoms during an
specificity of 98% (95% CI 95–100%).
Reproduced with permission from: Eriksen MK, Thomsen LL, Ole-
sen J. The Visual Aura Rating Scale (VARS) for migraine aura diag-
Aura symptoms usually develop gradually over five
nosis. Cephalalgia. 2005;25(10):801–10. or more minutes, and different aura symptoms, when
present, usually develop in succession, each lasting
between 5 and 60 minutes. Associated loss of vision
described in this case would have scored five of ten or blurred vision may be present, but blurred vision
possible points. The use of the VARS or elicitation of alone is not sufficient to make a diagnosis of visual
the history of similar episodes when he was younger aura. Table 6.4 lists the typical features of aura.
might have spared him an unnecessary workup for Aura may occur without a subsequent headache or
transient ischemic attack, and unnecessary treatment can be followed by a mild, non-migrainous headache.
with aspirin. In the past, some physicians prescribed With age, patients may lose characteristic migraine
low-dose aspirin therapy for patients with aura, but headache features or may not experience headache at
there is no evidence that this is beneficial, and some all after an aura. The diagnosis of visual aura with-
recent evidence suggests it may be harmful. Thus, we out headache thus occurs more frequently in older
do not recommend the use of low-dose aspirin therapy patients.
in patients with migraine aura.
Discussion Typical aura with headache.
Aura occurs in about a third of patients with migraine.
Visual aura is by far the most common type of aura, Tip
occurring in 99% of subjects with aura. This is fol- Failure to elicit historical features consistent with
lowed by sensory aura (31%) and aura involving lan- migraine aura can lead to an erroneous diagnosis of
guage symptoms such as aphasia (18%). Motor aura is cerebral ischemia in patients with aura.
rare, and this special form of aura is separately recog-
nized as hemiplegic migraine. Some forms of hemi- A young woman with a headache that
plegic migraine are autosomal dominant single gene
disorders; these have provided an important opportu- was “worse than ever”
nity to study the genetics of migraine.
Migraine aura is thought to be due to the phe- Case
nomenon of cortical spreading depression, which is an A 30-year-old obese female presented to a headache
orderly wave of depolarization that spreads anteriorly clinic with records from her initial evaluation at
across the cerebral cortex. This is followed by a pro- another clinic for the same symptoms. These records
longed period during which neurons struggle to repo- indicated that she had initially reported several
larize (hence the term “depression”) and is followed months of a “worse than ever” daily headache
by a small reduction in regional cerebral blood flow. sometimes associated with “black spots” in her
The wave of neuronal depolarization spreads at a rate vision, nausea, vomiting, and photo and phonopho-
of 2–5 mm per minute, which correlates closely with bia. She was taking a combination estrogen–progestin

Chapter 6: Missed historical features

contraceptive pill, and was a smoker. Her examination unfounded: certainly the patient’s nicotine addiction
was normal. The physician, hearing that she was hav- deserves attention.
ing “the worst headache ever,” was concerned about The picture that emerged of this patient’s
a secondary headache. He sent her for an emergent headaches appeared very different when consid-
MRI followed by an unsuccessful attempt at lumbar ered in the overall context of her headaches. She
puncture. A second lumbar puncture was attempted reported childhood carsickness and subsequent onset
under fluoroscopy and was finally successful. This was of headaches with migrainous features in her 20s.
followed by an MRV. The results of these tests were These headaches slowly progressed in frequency and
normal. severity over many years. The change in her headache
pattern in the preceding four months before the cur-
rent evaluation was probably the point at which the
Was the physician’s level of concern patient recognized that her headaches were becom-
ing chronic and more disabling. Because the initial
appropriate? headache history did not elicit this story of gradual
The phrase “worst headache ever” was understandably progression, unnecessary diagnostic testing and treat-
worrying to this patient’s doctor, who undoubtedly was ment ensued. Unnecessary diagnostic investigation is
concerned about secondary causes of headache such as often uncomfortable for the patient, provokes anxiety,
cerebral venous thrombosis or stroke. However, when and creates the potential for iatrogenesis.
viewed in the context of the patient’s history this sit-
uation is not so alarming. In the headache clinic, the
patient was asked to provide a chronologic history that
included childhood headache symptoms and the evo- In most cases the patient history is the most impor-
lution of her headaches from onset to their current pat- tant part of headache evaluation and sets the stage
tern. This revealed a strong family history of migraine, for decision-making about testing and treatment. Con-
and a personal history of years of episodic migraine ventional teaching is that the patient history is a step-
headache with a gradual transformation to a chronic wise process that begins with the history of the present
situation. illness, proceeds through a review of what brought
This pattern is very typical of the evolution of the patient in for evaluation at that moment (the
episodic to chronic migraine, caught around the time chief complaint), and then the details of the current
of transformation of the headache pattern. Further- problem.
more, the patient’s sex and obesity are factors asso- In the evaluation of headache, however, it is often
ciated with an increased risk of transformation from wise to begin with an open-ended question such as
episodic to chronic migraine. When this patient’s “tell me about your headaches” and then fill in details
headache history is considered as a whole, the exten- using a time-based or chronologic approach. This
sive diagnostic workup that was undertaken seems begins with questions about family history and child-
superfluous, costly, and unnecessarily invasive. hood headache symptoms, followed by detailed lon-
gitudinal questioning about migraine-related symp-
toms during the teenage years and then in each decade
beyond the teens, with appropriate attention devoted
Two evaluations of the same historical to headache changes associated with such events as
information produced divergent outcomes. menarche, pregnancy, marriage, or job changes. The
What went wrong? headache history can then be considered in the context
of the overall history of headaches, which provides a
The initial examiner failed to take into account the
more complete picture and is likely to produce a more
context into which the current headache situation
accurate diagnosis.
fits. This is another example of the cognitive error
of premature closure. The doctor settled on a likely
diagnosis of a dangerous secondary headache based Diagnosis
on some details but minus other useful information. Migraine without aura, with transformation to chronic
This does not mean that his concern was completely migraine.

Chapter 6: Missed historical features

Table 6.5. Types of adverse or traumatic experiences

Failure to consider a patient’s current headache prob- Physical abuse or neglect
lem in the context of the overall headache history can Sexual abuse
lead to unnecessary testing and treatment. Intimate partner violence
Emotional abuse or neglect

Intractable headache and psychiatric Alcohol or drug abuser in the home

No parents in the home or single parent
illness in a young woman Chronically physically or mentally ill person in the home

Case Witnessing violence in the home or social environment

A 24-year-old female had been followed in adult neu-

rology since age 16, and before that in a pediatric neu- such events. Table 6.5 lists some common forms of
rology clinic, for debilitating and limiting headache adverse or traumatic experiences. The prevalence of
associated with multiple psychiatric comorbidities. childhood sexual or physical abuse or neglect is high in
Over the span of eight years, five neurologists had eval- the general population and is not confined to a particu-
uated her for her headaches. She was also receiving lar ethnic or socioeconomic group. For example, some
treatment from a psychiatrist and psychologist for anx- sources estimate that as many as one in four girls and
iety, major depression, and suicidal gestures. She had one in six boys experience sexual abuse in childhood.
been hospitalized twice for management of psychiatric Abuse is not confined to children, however. Intimate
problems, and had periodically been unable to attend partner violence or other traumatic events frequently
school for weeks at a time. Both her headaches and her occur in adulthood, and those with a childhood his-
psychiatric problems had not improved despite multi- tory of adverse experiences may be at high risk for later
ple trials of treatment. revictimization.
When seen at her most recent visit, the patient told The prevalence of abuse or trauma is even higher
her neurologist that she had read a magazine article in patients with chronic headache disorders than in
about the possible connection between headaches and the general population. A large multicenter study
abuse. She told the doctor that she had been sexually performed in specialty headache clinics found that
abused as a child by a relative. She said she had just roughly a quarter of patients with migraine acknowl-
mentioned this to her therapist and was beginning to edged a history of physical or sexual abuse or neglect,
explore the issue. A recent psychiatry chart note con- and over a third reported a history of emotional abuse
firmed that the abuse history had “been revealed for or neglect. There was substantial overlap of the differ-
the first time.” Review of the patient’s record showed ent forms of maltreatment, and almost half of those
no documentation of any previous questions or dis- who experienced abuse in childhood reported experi-
cussions about physical, emotional, or sexual abuse or encing abuse in adulthood.
trauma, with the exception of a note in the psychiatry In our experience, it is prudent to maintain a
intake form that stated: “Not evaluated at this time.” high index of suspicion about abuse or trauma in
patients with difficult-to-treat headaches. This patient
Should the patient’s doctors have suspected had intractable, debilitating headaches from a young
age, associated with severe, difficult to treat psychiatric
a prior history of abuse or trauma? disease.
This patient’s doctors probably should have been more It is difficult to prove a causal relationship between
alert to the possibility that this patient had experienced maltreatment and headache occurrence or intractabil-
prior maltreatment, since a history of abuse is very ity, but several lines of evidence suggest this connec-
common in patients with refractory headache disor- tion. For example, several studies have demonstrated
ders. Traumatic experiences in childhood are linked a dose–response relationship between the number of
with an elevated risk of a large number of child- adverse childhood experiences and the prevalence and
hood and adult medical problems, including headache. risk of frequent headaches in both sexes. Some evi-
Trauma can result not only from the personal expe- dence suggests that childhood trauma can produce
rience of violence or abuse, but also from witnessing permanent changes in the hypothalamic–pituitary

Chapter 6: Missed historical features

axis, while other work implicates the development of on to whether they feel safe in their current living sit-
poor coping strategies, somatization, or dissociation. uation, finally asking whether they have ever been in
a situation that was unsafe or abusive. As with other
How and when should questions about sensitive information, it is helpful to frame the ques-
tions carefully. We try to “normalize” the disclosure of
abuse or trauma be asked? such information by telling patients that many people
Good quality evidence supports an association with chronic headache have experienced prior sexual
between maltreatment and chronic headache, but or physical abuse or trauma. We then let them know
there is no consensus about whether all patients that we routinely ask all patients about abuse, and fin-
should be questioned about a history of abuse. At ish with more specific questions.
present recommendations for screening are limited
to particular subgroups. For example, the United
States Preventive Services Task Force (USPSTF) has How should doctors respond when patients
reviewed available evidence and recommends that
“clinicians screen women of childbearing age for disclose abuse or trauma?
intimate partner violence, such as domestic violence, Most physicians are subject to mandatory report-
and provide or refer women who screen positive ing requirements if abuse is suspected in vulnerable
to intervention services . . . The USPSTF concludes patients such as minor children or the elderly, and
that the current evidence is insufficient to assess some states have additional reporting requirements.
the balance of benefits and harms of screening all These vary and physicians are encouraged to stay up
elderly or vulnerable adults (physically or mentally to date with local regulations. In the case of competent
dysfunctional) for abuse and neglect.” adult patients without immediate injuries, the physi-
This patient was prompted to tell her caregivers cian usually is legally obligated to respect the privacy
about a previous history of childhood sexual abuse of information disclosed in the doctor–patient rela-
after reading about its possible connection with tionship. The clinician’s role in these cases is limited to
chronic headaches. Her providers were distressed to helping the patient make decisions and ensuring they
find that this history had remained undocumented for have access to information about available help and
so long. Few patients are likely to be as forthcoming resources.
as this patient was. Our belief is that it makes sense to We tell adult patients who disclose abuse or trauma
ask about abuse routinely in patients seeking treatment that we are glad they told us about the maltreatment.
for headache problems, especially those whose prob- It is also important to tell the patient that “you do not
lems have resisted treatment or occur in the context of deserve to be treated like this.” The next step is to eval-
other psychiatric disorders. Psychiatrists often include uate the immediate safety of patients who are currently
a history of abuse in their patient evaluations, but as involved in abusive situations, and to give the patient
this case illustrates, this cannot be assumed. information about resources such as abuse hotlines,
There are several ways in which information about legal resources, and healthcare resources. We keep a
maltreatment can be obtained. If patients are asked to variety of information pamphlets on hand for this
fill out written surveys or questionnaires prior to being purpose.
seen, the question can be asked in writing. Validated In the case of patients who are involved in abusive
instruments such as the Childhood Trauma Question- relationships, many clinicians are tempted to advise
naire can be used. In our practice we include a writ- the patient to leave the relationship or ask them why
ten question about a history of trauma or abuse in the they put up with such treatment. Although well inten-
information we ask patients to complete before being tioned, such questions or advice may imply that you
seen. However, since patients may be reluctant to dis- hold the patient responsible for the abusive situation.
close this information on a form we also follow up with In some cases, such advice if acted upon may put the
a routine question during our initial meeting with the patient or any involved children in danger from the
patient. abuser. It is important to respect the patient’s choices
We usually ask these questions when inquiring about whether or when to end a relationship. We tell
about other aspects of the social history, first asking patients experiencing abuse that we are very concerned
patients about their living situation, and then moving about their welfare, recommend that they contact local

Chapter 6: Missed historical features

counseling resources, and urge them to make follow- A man with headache, neurologic
up appointments so that we can monitor the situation.
symptoms, and spinal fluid pleocytosis
Discussion Case
Good quality evidence supports a link between var- A 33-year-old male was seen in the emergency depart-
ious forms of childhood abuse (sexual, physical, or ment with headache and neurologic symptoms. There
emotional) and later life pain syndromes, including was a past history of occasional episodic migraine but
migraine. One study found that a past history of emo- the patient had not been troubled by migraines for
tional abuse was almost four times more common several years. He was otherwise healthy and taking
in patients, mostly female, with migraine. Migraine no medications. The patient had developed a constant
patients with a past history of such “adverse childhood mild frontal headache three days prior to evaluation.
experiences” may be more likely to experience not only On the fourth day, he had noted migratory right-sided
chronic migraine in later years but also depression and numbness and paresthesias in the arm, leg, and face.
anxiety. They may also be less compliant with treat- Shortly thereafter he was unable to speak for a short
ment or medical appointments, particularly if the sit- time, and following that he developed a “worst ever”
uation at home remains chaotic. headache with associated nausea and vomiting. This
Because this is a relatively new area of research, prompted urgent evaluation including a plain CT scan
there are no clear implications for patient management of the head followed by lumbar puncture, MRA, and
when a history of abuse is uncovered. We find, how- MRV. Test results were normal except for the find-
ever, that patients who have experienced intimate part- ing of elevated white blood cells (150 cells/␮L) in the
ner or sexual violence may have difficulty with treat- CSF, lymphocyte predominant. He was afebrile. He
ments that remind them of the circumstances of any was advised that the diagnosis was viral meningitis.
abuse. For example, a female patient who has been sex- One week later he noted a return of the numb-
ually abused by an older male may become anxious ness followed by a one-hour episode of confusion and
during physical therapy or relaxation sessions with an a moderately severe headache. He returned to the
older male therapist. emergency department where evaluation was again
It is not clear that the outcome of this patient’s care normal apart from elevated white cells in the CSF
would have been different if her abuse history had been (450 cells/␮L) and a mildly elevated CSF protein at
known. It does, however, seem evident that it is an 113 mg/dL. The diagnosis at that point was possible
integral part of her medical and psychosocial history. viral encephalitis. He was admitted to the hospital and
Knowledge of this part of the patient’s history might placed on seizure prophylaxis and antiviral agents.
have helped her caregivers to understand the patient’s
headaches in the context of her life experiences. It is Was a diagnosis of encephalitis appropriate
important not to underestimate the effect that trau-
matic life experiences can have on a patient’s headache in this case?
problems. At the initial presentation, viral meningitis was a rea-
sonable diagnosis, although the patient did not report
a full range of typical symptoms such as fever, stiff
Diagnosis neck, or other viral features (malaise, muscle aches,
Chronic migraine in a patient with a history of child- etc.). Additionally, the patient had reported the most
hood sexual abuse. severe headache of his life and there was no history
of prior headache or migraine. In this setting, a sec-
ondary headache of some kind was a leading diag-
Tip nostic possibility. During his second presentation the
Do not forget to ask all headache patients, especially patient’s symptoms included confusion, which indi-
those with treatment-refractory headaches, about a cates a process affecting either the brainstem or both
history of sexual, physical, or emotional abuse or halves of the cortex. On the other hand, no localizing
trauma. These high-yield historical questions can clar- neurologic findings were noted and imaging findings
ify a complex situation. remained normal. Although the patient appeared less

Chapter 6: Missed historical features

Table 6.6. Clinical characteristics of headache with neurologic

deficits and CSF lymphocytosis (HaNDL)
In hindsight, what historical and
r Episodes of moderate headache occurring with or quickly examination features might have
followed by transient neurologic deficits
r Episodes of headache and focal neurologic deficits recur over suggested this patient’s condition was
a period of less than 3 months
r Lumbar puncture findings show cerebrospinal fluid benign?
pleocytosis with lymphocytic predominance (⬎ 15 cells/␮L) Although abrupt headaches in combination with focal
r CSF culture and neuroimaging findings are normal
neurologic deficits may indicate a serious disorder,
they can also be due to benign, self-limited problems.
In this patient there were no signs of ischemia on
imaging and it would be unusual for repeated tran-
ill than might be expected with encephalitis, an early
sient ischemic attacks to occur in an otherwise healthy
or incomplete encephalitic process was plausible and
young man. Although CSF lymphocytosis was docu-
management for this potentially severe condition was
mented, there were no other signs of infection. The
patient continued to feel and act better than would be
The patient stayed in the hospital for several days.
expected if he had an ongoing encephalitic or ischemic
He remained afebrile and looked well at the time of dis-
process. Additionally, he was completely well between
charge. However, he returned again to the emergency
episodes and his spinal fluid cultures were negative,
department three days later with a similar presenta-
lowering the likelihood of any ongoing infectious
tion: headache, confusion, and paresthesias, though
now these were all milder than at the time of the ini-
Although physicians worry about missing poten-
tial presentation. He was afebrile, looked well, and had
tially severe but treatable conditions, they also have
a normal examination at this last emergency depart-
a duty to protect patients from unnecessary testing,
ment visit. He underwent a diagnostic workup similar
iatrogenesis, and undue anxiety. In this case, more
to his previous episodes and was again admitted to the
restraint in testing and treatment was probably war-
ranted when, despite the worrisome diagnoses under
consideration, the patient continued to look and do
What is the best diagnosis at this point?
The diagnostic possibilities in this case are becom-
ing clearer as a history emerges of repeated benign Discussion
episodes of headache and CSF abnormalities. This HaNDL is a rare but well-described syndrome that
patient presented with new-onset headache without may be confused with meningoencephalitis or other
migrainous features, with mild reversible focal neu- serious conditions. It is benign and usually resolves on
rologic deficits, and CSF lymphocytosis. Furthermore, its own. There are several case reports of patients with
such episodes recurred. This is the classic description HaNDL who were suspected of having acute stroke and
of headache with neurologic deficits and CSF lympho- treated with systemic thrombolysis. Failure to recog-
cytosis (HaNDL), which was first described in 1981. nize the characteristic historical and examination fea-
Features of the syndrome that are helpful for diagnos- tures of this illness can thus have serious consequences.
tic purposes include a finding of CSF lymphocytosis In this case the patient was unnecessarily treated with
along with negative cultures (Table 6.6). antiepileptic and antiviral therapies.
Clinical mimics of this syndrome include migraine HaNDL was initially thought to represent a form
aura, transient ischemic attacks or stroke, as well as of migraine, but current thinking centers on the pos-
viral meningitis or encephalitis. In this case, there sibility of a post-viral syndrome. The reversible focal
were no migraine features to the headache and the deficits have been attributed to transient vasoconstric-
episodes were not stereotyped, as would be expected tion. Some authors report that steroid treatment can
in migraine aura. Although about a third of patients be helpful, although there are no large-scale studies to
with HaNDL have a history of migraine, the clinical support this view. Catheter angiography may worsen
significance of this is uncertain because the prevalence symptoms and should be avoided if this condition is
of migraine is so high in the general population. suspected.

Chapter 6: Missed historical features

Diagnosis Was giant cell arteritis the best initial

Headache with neurologic deficits and CSF lymphocy- explanation of the headache pattern?
tosis (HaNDL).
Headache associated with giant cell arteritis has been
described as generalized and throbbing, intermittent
or constant; but it can also be nonspecific, as it was in
Tip this case. Temporal artery tenderness and enlargement
Prompt recognition of benign historical and examina- may be found on physical examination. In this case, a
tion findings can save patients from invasive and dan- temporal artery biopsy was positive and the patient ini-
gerous treatments or multiple hospitalizations. tially improved with corticosteroid treatment. When
her headache recurred, the physician assumed the
symptoms were due to the original diagnosis but failed
to consider other possible explanations for the contin-
An elderly woman with giant cell ued headache. This is an example of “anchoring bias,”
in which new historical information is interpreted in
arteritis and persistent headache the light of an established diagnosis, without consid-
on steroids eration of other explanations. In hindsight, although
this patient certainly did have giant cell arteritis, it is
Case unclear how much of the initial headache report was
A 73-year-old female presented for evaluation of a explained by the arteritis. It is plausible that the symp-
mild to moderately severe headache of three months’ toms of low CSF volume prompted her initial evalua-
duration. She described an achy global headache asso- tion. The possibility that this patient had two causes of
ciated with mild photophobia, ear pain and nausea. headache was not considered until relatively late in her
She did not have distended temporal arteries or any course.
head soreness. There were no constitutional or jaw
claudication symptoms. As part of her evaluation, an When should a diagnosis be reconsidered?
erythrocyte sedimentation rate (ESR) and C-reactive
In our experience, headaches attributable to giant
protein (CRP) were obtained. Both were moderately
cell arteritis typically improve promptly with corti-
elevated, and a subsequent temporal artery biopsy
costeroid treatment, although there are occasional
reportedly showed giant cell arteritis.
patients where headaches linger. Worsening of
Her headache initially improved on initial doses of
headache is sometimes attributable to incomplete
prednisone but later returned during the prednisone
treatment of the arteritis and an elevation in steroid
taper. The physician attributed this worsened headache
dose is a reasonable first step. However, failure to
to ongoing arteritis and recommended an increase in
respond as expected to treatment (in this case to an ele-
the patient’s corticosteroid dose. At the patient’s next
vation of the corticosteroid dose) should raise concern
visit she reported that this increased dose had not
about alternative or additional diagnostic possibilities.
improved her symptoms. The physician asked her to
As was done by the physician in this case, it is help-
give a more complete description of the headache. She
ful in such situations to put aside assumptions and re-
reported that the headache was absent on waking each
evaluate the headache history from the beginning. It
day, but would begin after becoming upright for 10–15
can also be useful to discuss the case with a colleague,
minutes and progress throughout the day to a severity
since presenting a case to someone forces a reconsid-
of 6 on a 0–10 pain scale. The headache was dull and
eration of the entire clinical picture.
bilateral. Her MRI was reviewed and, although it had
originally been read as normal, showed subtle find-
ings consistent with low CSF volume. Her symptoms Discussion
improved with use of an abdominal binder when she The physician in this case initially did not recognize
was upright and a regimen of daily caffeine tablets. She that the historical evolution of the patient’s headache
was eventually sent for an epidural blood patch and her deviated from the clinical course expected with her ini-
headache cleared completely. tial diagnosis. Upon revisiting the history, however, he

Chapter 6: Missed historical features

elicited important information about postural features diagnosis of hemicrania continua and recommended
of headache. This eventually led to the diagnosis of an treatment with indomethacin.
additional cause of headache: low CSF pressure, pre- The physician performed a brief examination but
sumably from a spontaneous leak. the patient declined to remove his sunglasses for a fun-
It is important to be vigilant for giant cell arteri- duscopic inspection. As the patient was about to leave,
tis in older patients, since it is a medical emergency. however, the patient’s wife asked the doctor why her
The clinical picture of giant cell arteritis can be highly husband was “squinting” and noted “It’s been like that
variable, and early treatment with corticosteroids may since the headache started.” Upon further examina-
prevent permanent visual loss. Thus, a low threshold tion, the patient was noted to have right-sided ptosis
for biopsy and treatment is appropriate. It is useful to and miosis, with warmer skin on the right side of his
remember, however, that corticosteroid treatment may forehead. Further examination revealed a suggestion
produce temporary improvement in a wide variety of of a slight left-sided central facial palsy and mild left
headache types, so response to steroid treatment is not upper extremity weakness.
What is the differential diagnosis, and what
Diagnosis tests should be performed next?
Giant cell arteritis; low CSF pressure headache. Autonomic features are frequent accompaniments of
many types of headache, most notably the trigemi-
nal autonomic cephalgias and hemicrania continua.
Tip They can also be seen in migraine. Thus, the differen-
Patients can have more than one cause of headache. tial diagnosis of side-locked headache in association
When patients do not respond as expected to treat- with ipsilateral autonomic features includes benign
ment, the history should be carefully reviewed for headaches such as cluster headache, hemicrania con-
overlooked clues. tinua, or even migraine, but also dangerous conditions
such as giant cell arteritis, carotid dissection, or cervi-
A man with unilateral headache and cal facet disease.
Neuroimaging should be considered in all patients
subtle eye findings with side-locked headaches, even when they appear
to meet criteria for a benign headache disorder. MRI
Case of the cervical spine and MRA of the extracranial
A 62-year-old man was seen for evaluation of a carotid circulation are indicated when there are abnor-
headache that had started about three weeks ago. mal examination findings, as there were in this case. In
He had a prior history of occasional bilateral, mild older patients an ESR and CRP should also be consid-
headaches that lasted a few hours at most, but this ered, since giant cell arteritis can present as a unilateral
headache was located behind the right eye and was headache.
continuous since onset. The patient reported that the This patient had a normal ESR and CRP, but MRA
headache began after he had been target shooting for showed dissection of the extracranial part of the right
several hours. This required him to close his left eye internal carotid artery with significant stenosis. The
frequently and hold his neck in a particular posture. patient was treated with aspirin.
The evening of target practice and for two days after,
the patient had experienced blurry vision. His vision What is the cause of the physical
then normalized, but he developed a continuous dull
right retro-orbital headache that radiated to the angle examination findings in this patient?
of his jaw on the right. He had photophobia and wore The combination of ptosis and miosis is character-
sunglasses, but denied any associated features such as istic of Horner’s syndrome. A lesion of sympathetic
nausea or vomiting. He mentioned that he felt as if he fibers that innervate the face and eye causes this
had some grit or sand in his right eye. Based on the uni- problem. Other findings in Horner’s syndrome may
lateral location of the headache, the physician made a include anhidrosis (decreased sweating) on the side

Chapter 6: Missed historical features

of the lesion. Flushing and conjunctival injection are foreign body sensation in the eye. The physician in this
less frequent. First-order neurons from the poste- case probably reasoned that she was unlikely to find
rior hypothalamus travel to the upper cervical and anything on physical examination and so skipped this
lower thoracic segments of the spinal cord, where step.
they synapse with second-order preganglionic neu- When autonomic features occur in close associa-
rons. These travel over the apex of the lung, pass tion with discrete episodes of headache pain, and espe-
around the subclavian artery, and finally end in the cially if lacrimation or rhinorrhea also occur, the cause
superior cervical ganglion. Here they synapse with is most likely to be benign activation of the autonomic
third-order postganglionic neurons, some of which nervous system by a headache. It is easy to under-
travel with the internal carotid artery through the cav- stand how the physician in this case made a diagnosis
ernous sinus to innervate the pupil and muscles of the of hemicrania continua, especially since the feeling of
eyelid, others of which travel with the external carotid “grit” in the eye is a common complaint in that disor-
artery to the facial sweat glands. der. More careful examination, however, prompted by
Suspicion should be high that patients with new- the wife’s parting question about “squinting,” revealed
onset Horner’s syndrome and headache may have a subtle neurologic signs that raised suspicion of a more
dangerous secondary explanation for their problems. sinister explanation for the patient’s problem.
This is true even when, as in this case, the headaches Physicians may perform brief physical exams for
appear to meet criteria for a benign form of headache. many reasons: a shortage of time or a belief that the
A Horner’s syndrome may occur from a lesion any- diagnosis is clear based upon a characteristic history.
where along the sympathetic pathway. If needed, spe- They may reason that therefore the physical examina-
cial pupillary testing can establish whether lesions are tion is an afterthought, or wish to spare the patient dis-
pre- or postganglionic. One frequent cause that also comfort. In this case, several of these factors were likely
presents with head or neck pain is carotid dissec- at play. This case is a reminder of the importance of a
tion. As described above, the sympathetic fibers lie thorough physical exam even if it means putting the
just outside the carotid artery and a dissection may patient through temporary discomfort, such as expos-
cause stretching of the fibers sufficient to disrupt func- ing a photophobic patient to a brief ophthalmologic
tioning. For this reason, a patient who presents with exam.
headache and a Horner’s syndrome, or neck pain and
a Horner’s syndrome, should be evaluated with carotid
imaging. This could be accomplished with duplex Discussion
sonography, a CTA of the neck, or an MRA with fat Headache is present in roughly two-thirds of patients
suppression sequences. with carotid dissection, and is usually ipsilateral to the
dissection. Horner’s syndrome is also common. A his-
tory of previous minor head or neck trauma, as in this
What accounts for the diagnostic “near case, is also common. Dissection is thought to occur
miss” in this case and how can mistakes like when small tears in the intimal layer of the artery allow
arterial blood (which is under high pressure) to pen-
this be avoided? etrate into the medial layers of the arterial wall. The
The physician in this case did not perform a careful resulting hematoma can produce partial or complete
physical examination, perhaps because she fell prey to occlusion of the artery, and emboli arising in this area
the cognitive error of “premature closure.” No addi- may also result in ischemia.
tional diagnoses were considered once a verdict that Cardiac risk factors such as hypertension or hyper-
“fits the facts” had been found. The physical exam- cholesterolemia may raise the risk of dissection, as
ination findings in this patient were subtle, and his do connective tissue disorders. Chiropractic manipu-
sunglasses probably masked the eye abnormalities. lation of the neck has also been associated with dis-
He refused funduscopic examination, but that is not section of cranial arteries, and this is one reason we
unusual in patients with primary headache disorders do not recommend it for the treatment of headache.
who are sensitive to light. In addition, the patient’s Dissection is usually treated with either anticoagula-
history was highly suggestive of a particular form of tion or antiplatelet agents. A recent nonrandomized
benign headache, including the “classic” feature of a study did not show clear benefit for either approach

Chapter 6: Missed historical features

so in view of the potential risks of full anticoagula- hematoma is also a possibility, but again there are no
tion, many physicians opt for treatment with aspirin focal symptoms to support these diagnoses. An inflam-
or other antiplatelet regimens. matory process also seems unlikely.

Diagnosis What laboratory or other investigation

Carotid dissection with Horner’s syndrome. should this patient have?
A careful workup to exclude the dangerous diag-
Tip noses discussed above is appropriate in this situation.
A careful ophthalmologic examination is important The patient in this case underwent a full battery of
in all patients who present with headache. A cursory blood tests, including an ESR and CRP. A computed
examination can miss subtle but important findings. tomographic scan of the head with and without con-
trast showed no abnormalities, and a lumbar puncture
An older woman with focal headache showed clear fluid with a normal pressure. The patient
was diagnosed with recurrent migraine. She was given
Case a prescription for acetaminophen with codeine and
advised to see her primary care physician in a few days
A 68-year-old woman was seen in the urgent care
if the symptoms did not improve.
department of the hospital because she had awakened
The pain continued over the next day and was only
earlier that day with a headache. The pain was burn-
partially relieved by acetaminophen with codeine. On
ing, throbbing, and located over the left temple and
the afternoon of the second day, the patient began
forehead. She rated it 8 on a pain scale of 0–10. The
to experience itching as well as pain and noticed the
headache was not associated with additional symp-
development of a vesicular rash over her left forehead.
toms such as nausea, vomiting, photo or phonophobia.
She saw her primary care physician who diagnosed
She had a remote history of episodic migraine without
herpes zoster (shingles).
aura but had not experienced any migraine headaches
for the last 12 years. Her physical examination was nor-
mal with the exception of elevated blood pressure of How should this patient’s herpes zoster
178/96 mm Hg. Her neurologic examination was also be treated?
normal, with no bruits or tender or indurated tempo-
Soothing topical lotions and analgesics are appropriate
ral blood vessels.
symptomatic treatments for the rash of herpes zoster.
Eye involvement occurs in 10–25% of patients with
What is the differential diagnosis? zoster, when zoster affects the ophthalmic division of
There is a long list of diagnostic possibilities that the trigeminal nerve. This patient should be referred
should be considered in an older patient with the sub- for ophthalmologic evaluation since zoster affecting
acute onset of a new type of headache. This patient the eye can pose a threat to vision. Ophthalmologists
has a history of migraine but her current headache has may prescribe antiviral eye drops, or ointments may be
few migrainous features. It would be unusual for severe helpful.
migraine to recur abruptly at this age. Because the pain of zoster can be intense, many
Some sort of vascular problem is at the top of the patients require the use of opioid analgesics. Most doc-
list of differential diagnoses in this case. While sud- tors also treat patients with systemic antiviral agents
den, severe elevations in blood pressure can provoke such as acyclovir or famciclovir. These may reduce the
headache, this patient’s modest degree of hypertension duration of the attack and speed healing, but the drugs
is an unlikely explanation for her headache. Could this are likely to be most effective when started as soon
be a cerebral hemorrhage or infarct? This patient is as possible after symptoms begin, so time is of the
alert and has no focal neurologic deficits, making this essence.
possibility less likely. Giant cell arteritis is a possibil- Timely use of antiviral agents may decrease the
ity, but this patient does not have additional systemic chance that patients will develop persistent post-
signs or symptoms suggestive of that disorder. A mass herpetic pain, and additional treatment with cortico-
lesion of some sort or a slowly expanding subdural steroids may help some patient subgroups. Many

Chapter 6: Missed historical features

doctors thus use both antiviral agents and steroids to A 27-year-old woman with weakness
treat herpes zoster. Postherpetic neuralgia occurs in
10–18% of patients with zoster, and is frustrating to and headache
treat. The usual treatment is amitriptyline, but results
are often disappointing and anticholinergic side effects Case
may limit its use in elderly patients. Carbamazepine A 27-year-old woman presented to the headache clinic
and gabapentin are also used. for a new patient visit after moving to the area for work.
She had developed migraines at the age of six but was
otherwise healthy. Headaches had varied in frequency
Discussion over the years and were now occurring two or three
Herpes zoster results from spontaneous reactivation of times a month. The headaches were typical of migraine
the varicella zoster (chickenpox) virus, which remains and a previous brain MRI had been normal.
latent in the sensory ganglia after the initial chicken- The patient reported that all of her headaches were
pox infection. It is a common viral infection, with an preceded by right-sided numbness and weakness. She
incidence of about 4 per 1000 people per year. The could tell when a headache was likely because she
elderly and immunocompromised individuals are at began to drop things and experienced some difficulty
highest risk. Zoster is easy to diagnose when the char- walking. The patient had multiple family members
acteristic vesicular rash is present but, as illustrated by with migraine, but no one else had experienced pre-
this case, pain may precede the appearance of the rash ceding symptoms like hers. These symptoms typically
by several days. In those patients, the severe localized lasted about an hour and cleared as the headache
pain can easily be mistaken for other illnesses such as began.
migraine. Her previous neurologist had diagnosed hemi-
A zoster vaccine has recently become available and plegic migraine and told her that she should stop using
is recommended for use in those over the age of 60. sumatriptan, which she had taken since she was 18
It is intended to prevent zoster, and is not a treatment and which had worked well for her severe headaches.
for existing cases of shingles or postherpetic neural- Instead, he prescribed a butalbital-containing combi-
gia. One clinical trial suggested that vaccination of nation analgesic but the patient reported it was not
1000 patients of age 60 or older would prevent one very effective.
case of postherpetic neuralgia over the subsequent Her new neurologist asked the patient to call for an
three years. Unfortunately, vaccination is least effec- urgent appointment the next time she experienced an
tive in patients who need it most: the very elderly and episode of weakness. On examination during such an
those who are immunocompromised. Nonetheless, we episode, the patient had dense numbness of the right
strongly recommend to all of our patients over the hand, arm, foot, and calf. Her motor strength, however,
age of 60 that they consider vaccination. There is no was 5/5 throughout. Because of this, her diagnosis was
evidence that pre-existing headache disorders such as revised to that of migraine with typical sensory aura
migraine increase the risk of developing zoster or its rather than hemiplegic migraine. She resumed the use
complications, but we think it is worth minimizing the of sumatriptan, which was again effective.
chance that patients with an existing headache disor-
der will also have to cope with postherpetic neuralgia.
What were the diagnostic possibilities based
Diagnosis on the history?
Herpes zoster. This patient had a long history of headaches that
met criteria for migraine and responded well to
migraine-specific therapy with triptans. She also expe-
Tip rienced stereotyped focal neurologic deficits prior
The diagnosis of herpes zoster is easily missed in to her headache episodes that from her description
patients who present with pain and no rash. The pos- sounded like transient right hemiparesis. The list of
sibility of herpes zoster infection should be considered diagnostic possibilities in this case includes migraine
in elderly or immunocompromised patients with the with aura; familial or sporadic hemiplegic migraine; a
subacute onset of severe unilateral head pain. seizure disorder; vascular disorders such as arteritis or

Chapter 6: Missed historical features

transient ischemic attacks; mitochondrial myopathy, Other features of this patient’s presentation were
encephalopathy, lactic acidosis, and stroke-like symp- not wholly compatible with a diagnosis of hemiplegic
toms (MELAS); and cerebral autosomal dominant migraine. She did have a family history of migraine,
arteriopathy with subcortical infarcts and leukoen- but no one else in the family had similar neurologic
cephalopathy (CADASIL). accompaniments to headache. This makes a history
Many of these diagnoses can be immediately of hemiplegic migraine less likely since in most cases
excluded based on the patient’s history. She had no there is a strong family history compatible with auto-
cognitive impairment or complex motor behavior somal dominant transmission of the disorder. Speech
associated with her episodes, so a seizure disorder was disturbances are also common in hemiplegic migraine,
unlikely. MELAS or CADASIL were likewise improb- but this patient had no speech component to her aura.
able based on a normal brain MRI that showed no evi- Additionally, this patient’s aura symptoms preceded
dence of prior infarcts or other abnormalities. Like- her headache and faded as the headache began. Most
wise, vascular disorders are an unlikely explanation for patients with hemiplegic migraine experience aura
headaches in a healthy young woman who has experi- symptoms throughout the headache and occasionally
enced no permanent problems despite multiple similar beyond.
attacks over two decades.
About a third of patients with migraine expe-
rience focal neurologic deficits before a headache. How does the management of migraine
Typical aura symptoms can include visual, sensory, with typical aura differ from that of
or speech phenomena. Patients who experience true
motor weakness in association with migraine are clas- hemiplegic migraine?
sified separately as having hemiplegic migraine. This is Vasoconstrictive medications such as triptans that
usually an inherited autosomal dominant disorder and are routinely used for the treatment of patients with
a number of single-gene mutations have been iden- migraine without aura and migraine with typical aura
tified that are associated with hemiplegic migraine. are contraindicated in patients who have familial
Non-familial cases also occur and are attributed to hemiplegic migraine. This contraindication is based
spontaneous mutations. The patient’s history is sug- on fears that triptans or ergots might be more likely
gestive of transient hemiplegia in association with to cause vasoconstrictive complications in patients
headaches, so the differential diagnosis in this case is who have complex forms of aura such as hemiplegic
between migraine with typical aura and hemiplegic migraine. There is no firm scientific foundation for
migraine. these worries, but formal package labeling contraindi-
cates the use of these agents. Because of this, most
physicians are understandably reluctant to allow their
What historical and physical examination use in patients who carry a diagnosis of hemiplegic
findings were at odds with the diagnosis of migraine.
Most patients with hemiplegic migraine have rel-
hemiplegic migraine? atively infrequent attacks. Depending upon the sub-
It can be challenging to discriminate between typical type, these are sometimes triggered by head injury,
migraine aura and hemiplegic migraine. These disor- emotional stress, or hormonal fluctuations. In most
ders share many features of migraine including severe cases of hemiplegic migraine the neurologic deficits
headache and neurologic accompaniments to those are reversible, but there are cases in which neuro-
headaches. One common pitfall is accepting at face logic deficits have progressed over time or become per-
value a patient’s report of motor weakness. Patients manent. Because attacks are usually infrequent, most
can have difficulty distinguishing between sensory and patients do not require preventive medications. There
motor symptoms. Numbness can make it difficult to is a clinical impression, but no firm evidence, that
grasp or hold objects, and the resulting clumsiness may calcium channel antagonists or acetazolamide may
be interpreted as due to weakness. In this case, the be more desirable preventive agents for hemiplegic
patient’s new physician correctly recognized that it was migraine than for migraine with typical aura. Many
important to examine the patient during an episode to physicians believe it is prudent to avoid beta-blockers
assess for objective signs of true motor weakness. in patients with hemiplegic migraine.

Chapter 6: Missed historical features

Discussion Tietjen GE, Brandes JL, Peterlin BL, et al. Childhood

maltreatment and migraine. Headache. 2010;50(1):
Hemiplegic migraine is a diagnosis of exclusion that 20–51.
can have serious implications for patients. In this case
Tietjen GE, Khubchandani J, Herial NA, Shah K. Adverse
the patient was prevented from using highly effec- childhood experiences are associated with migraine and
tive treatment for her migraine based on a mistaken vascular biomarkers. Headache. 2012;52:920–9.
diagnosis of hemiplegic migraine. This could have Teitjen GE, Peterlin BL. Childhood abuse and migraine:
been avoided if the physician who made the diagno- epidemiology, sex differences, and potential
sis had made efforts to verify the patient’s report of mechanisms. Headache. 2011;51:869–9.
Diagnosis Bartleson JD, Swanson JW, Whisnant JP. A migrainous
syndrome with cerebrospinal fluid pleocytosis.
Migraine with typical (sensory) aura. Neurology. 1981;31:1257–62.
Berg MJ, Williams JS. The transient syndrome of headache
Tip with neurologic deficits and CSF lymphocytosis.
Neurology. 1995;45:1648–54.
Patients can easily confuse numbness and weakness.
Krause T, Nolte CH. The syndrome of transient headache
Patient reports of motor weakness in association with and neurological deficits with cerebrospinal fluid
headache should be verified by examination during an lymphocytosis (HaNDL) as an acute ischemic stroke
episode. mimic leading to systemic thrombolysis: a case report.
Clin Neurol Neurosurg. 2012;114(6):689–90.
Further reading Nakashima K. Syndrome of transient headache and
neurological deficits with cerebrospinal fluid
Typical aura with headache lymphocytosis: HaNDL. Intern Med. 2005;44(7):690–1.
Hansen JM, Lipton, RB, Dodick, DW, et al. Migraine
headache is present in the aura phase: a prospective Low CSF pressure headache
study. Neurology. 2012;79:2044–9. Mokri B. Spontaneous low cerebrospinal pressure/volume
Russell MB, Olesen J. A nosographic analysis of the headaches. Curr Neurol Neurosci Rep. 2004;4(2):117–
migraine aura in a general population. Brain. 24.
1996;119:355–61. Tseng YL, Chang YY, Lan MY, Wu HS, Liu JS. Spontaneous
intracranial hypotension in a patient with reversible
Transformed migraine pachymeningeal enhancement and brain descent. Chang
Scher AI, Stewart WF, Buse D, Krantz DS, Lipton RB. Gung Med J. 2003;26(4):293–8.
Major life changes before and after the onset of chronic
daily headache: a population-based study. Cephalalgia. Carotid artery dissection
2008;28(8):868–76. Kennedy F, Lanfranconi S, Hicks C, et al.; CADISS
Scher AI, Stewart WF, Ricci JA, Lipton RB. Factors Investigators. Antiplatelets vs anticoagulation for
associated with the onset and remission of chronic dissection: CADISS nonrandomized arm and
daily headache in a population-based study. Pain. meta-analysis. Neurology. 2012;79(7):686–9.
2003;106(1–2):81–9. Parwar BL, Fawzi AA, Arnold AC, Schwartz SD. Horner’s
syndrome and dissection of the internal carotid artery
Headache and maltreatment after chiropractic manipulation of the neck. Am J
Ophthalmol. 2001;131(4):523–4.
Nelson HD, Bougatsos C, Blazina I. Screening women
for intimate partner violence: a systematic review Patel RR, Adam R, Maldjian C, et al. Cervical carotid artery
to update the U.S. Preventive Services Task Force dissection: current review of diagnosis and treatment.
recommendation. Ann Intern Med. 2012;156(11): Cardiol Rev. 2012;20(3):145–52.
796–808, W-279, W-280, W-281, W-282.
Norman RE, Byambaa M, De R, et al. The long-term health Herpes zoster
consequences of child physical abuse, emotional abuse, Harpaz R, Ortega-Sanchez IR, Seward JF; Advisory
and neglect: a systematic review and meta-analysis. PLoS Committee on Immunization Practices (ACIP) Centers
Med. 2012;9(11):e1001349. for Disease Control and Prevention (CDC). Prevention

Chapter 6: Missed historical features

of herpes zoster: recommendations of the Advisory Hemiplegic migraine

Committee on Immunization Practices (ACIP). MMWR Lafrenière RG, Rouleau GA. Identification of novel genes
Recomm Rep. 2008;57(RR-5):1–30. involved in migraine. Headache. 2012;52:107–10.
Sanford M, Keating GM. Zoster vaccine (Zostavax): a Russell MB, Ducros A. Sporadic and familial hemiplegic
review of its use in preventing herpes zoster and migraine: pathophysiological mechanisms, clinical
postherpetic neuralgia in older adults. Drugs Aging. characteristics, diagnosis, and management. Lancet
2010;27(2):159–76. Neurol. 2011;10(5):457–70.

Errors in management of acute headache

Nearly every patient who experiences headache will patients treat headaches first with a nonspecific agent
use some form of therapy to treat his or her acute such as an NSAID, and then go on to a specific ther-
headaches. Also known as abortive or symptomatic apy (usually a triptan) only if the NSAID is not effec-
therapies, these treatments are intended to relieve the tive. In stratified care, patients use nonspecific agents
pain of headache, as well as accompanying symptoms as first-line therapy for less severe and specific thera-
such as nausea. Acute therapies can range from non- pies as first line for more severe headaches; in other
pharmacologic treatments such as topical ice or heat, words, the treatment is tailored to the patient’s indi-
to nonspecific over-the-counter and prescription anal- vidual circumstances. A randomized trial comparing
gesics, to treatments specific to the headache type. stepped care with stratified care in the treatment of
Triptan therapy for treatment of migraine falls into this acute migraine showed that stratified care was more
last category. effective and associated with higher patient satisfac-
In addition to the particular pitfalls illustrated tion. For this reason, we strongly recommend that
by the cases in this chapter, a few general princi- patients with troublesome migraine should be treated
ples should be kept in mind. Treatments for acute with stratified care approaches rather than required to
episodes of headache are generally more effective when fail therapies that are usually ineffective before being
taken early in the headache, before central sensitiza- allowed to use highly effective therapy.
tion develops. One study showed that triptans were
less effective once cutaneous allodynia (a marker for
central sensitization) had developed. Unfortunately,
A migraineur who does not respond to
many patients wait until a full-blown headache has triptan treatment
developed before taking a triptan, so education about
the importance of treating early can by itself improve Case
response rates. A 30-year-old woman was referred for consultation
When a single treatment for headache is not because of recurrent headaches for several years. She
effective, a combination of treatments may be more was in good health and her neurologic examination
effective. For example, some patients respond well was normal. Her headaches met diagnostic criteria for
to a combination of triptans and nonsteroidal anti- migraine without aura and occurred on average twice
inflammatory drugs (NSAIDs) but not to either med- a month. She had successfully treated headaches with
ication given alone. Likewise, patients who develop 550 mg of naproxen sodium until two years ago when
nausea early in the course of their migraines will she underwent gastric bypass surgery for obesity. Post-
likely have a better response to oral triptan ther- operatively, the surgeon told her that she should not
apy if they are also treated with an antiemetic. The use naproxen or other nonsteroidal anti-inflammatory
phenothiazine antiemetics (such as promethazine and drugs. To replace naproxen she was given a prescrip-
prochlorperazine) probably have some intrinsic anti- tion for 1 mg naratriptan tablets. These were not effec-
migraine effects as a result of their anti-dopaminergic tive, so her physician prescribed 2.5 mg of zolmitrip-
effects. tan, which also did not help.
Lastly, it is worth mentioning the debate about Eventually the patient was given a prescription for
whether stepped or stratified care is a better approach oxycodone. This partially relieved her headache pain
to treatment of acute headaches. In stepped care, but produced nausea and made it impossible for her to

Chapter 7: Errors in management of acute headache

function at her job as a nurse. The patient tried several mulations of triptans or that treatment has not been
migraine preventive drugs that were not tolerable or properly timed. In our experience, these are the most
effective. She does not want to take medicine every day common explanations for an apparent lack of response
for something that happens only twice a month. to triptan treatment.
Clinical trials suggest that the maximum response
Why should NSAIDs be avoided after rate to triptans is approximately 70%, although in
clinical practice they seem to be slightly less effec-
bariatric surgery? tive, with response in about 60% of treated attacks.
Gastric ulceration is a well-known side effect of The likelihood of a good response to a triptan seems
NSAIDs. Bariatric surgery does not necessarily to depend on the dose and formulation of the drug,
increase the risk of gastric ulceration. Rather, ulcer- and the timing of treatment. A recent meta-analysis
ations that do occur can be more difficult to detect of sumatriptan trials showed the 100 mg dose was
after bariatric surgery and may produce more serious effective for a larger percentage of patients than the
complications. Thus most bariatric surgeons recom- 50 mg dose, which in turn was more effective than
mend that patients avoid this class of drugs following the 25 mg dose. This dose–response relationship has
surgery. been demonstrated for other triptans. The patient in
The most common type of gastric bypass surgery this vignette was treated with the lowest commercially
in the United States is the Roux-en-Y gastric bypass, available doses of naratriptan and zolmitriptan. It is
which sections off a portion of the stomach into a possible that higher doses of these drugs (e.g. 2.5 mg of
small pouch to limit food intake. The small intestine naratriptan or 5 mg of zolmitriptan) would have been
is repositioned and connected to the pouch at the top effective.
of the stomach, where its narrow opening slows stom- Most patients prefer oral migraine treatments but
ach emptying. After this procedure the stomach cannot parenteral drugs have higher bioavailability. The ther-
easily be visualized with gastroscopy, making it diffi- apeutic gain for subcutaneous sumatriptan, for exam-
cult to detect ulcers if they are suspected. Nonsteroidal ple, is higher than for any of the oral triptans.
anti-inflammatory drugs also may increase the risk of Intranasal formulations of sumatriptan and zolmitrip-
anastomotic leakage. Finally, because the active por- tan are available, but these are not more effective than
tion of the stomach is very small, any ulceration that oral triptans. Only a small amount of these intranasal
does occur covers a larger proportion of the total sur- triptans is absorbed through the nasal mucosa; rather,
face of the stomach. most of the liquid drips down the back of the throat,
A common misconception among patients and is swallowed, and absorbed through the gastrointesti-
some physicians is that avoiding oral ingestion of non- nal tract in the same way as orally administered drugs.
steroidal anti-inflammatory medications will circum- Likewise, the orally disintegrating formulations of
vent this problem. Nonsteroidal anti-inflammatory rizatriptan and zolmitriptan are not well absorbed sub-
drugs can be administered intranasally, parenterally, lingually but dissolve and are swallowed. The patient in
or as rectal suppositories, but these formulations do this case was treated with oral triptans, and it is possi-
not reduce the risk of stomach ulceration. The risk of ble that parenteral sumatriptan would have been more
gastric ulceration is related to systemic inhibition of effective.
prostaglandins and not to direct irritation of the gastric Finally, the timing of drug administration may
mucosa, so non-oral administration of NSAIDs does spell the difference between successful triptan treat-
not reduce the risk. ment of a migraine attack and apparent drug fail-
ure. Some research suggests that triptan treatment is
What are the possible explanations for this less likely to be effective when delayed until headache
intensity is severe or the attack is well established. This
patient’s failure to respond to triptans? lack of benefit may be due to the development of cen-
Triptans are very effective but do not work for every- tral sensitization. We do not know at what point in
one. One possibility is that this patient is truly refrac- her headaches this patient treated her attacks, but it is
tory to triptans. On the other hand, it is possible that possible that early treatment when the headaches were
she has been treated with subtherapeutic doses or for- mild would have been more effective.

Chapter 7: Errors in management of acute headache

Table 7.1. Selected strategies for optimizing triptan treatment

of acute headache
It is a mistake to conclude that patients do not respond
Use adequate doses
Try a different triptan to triptans without ensuring that they have used an
Try combination therapy adequate dose and formulation early in a headache.
– Triptans with NSAIDs
– Triptans with antiemetics
Treat early
Use alternative formulations (nasal spray, injection, orally
A young woman with chest pressure
dissolving tablet) from sumatriptan
Discussion Case
A 19-year-old woman with episodic migraine with-
Most patients with migraine or cluster headache
out aura was seen in the emergency department (ED)
respond to a triptan if the right dose and formulation
with a headache and nausea. The headache was sim-
are used. Treatment is most effective when adequate
ilar to her previous attacks but had not responded to
doses of the drug are used early in a headache. Thus,
her usual treatment of 1000 mg of aspirin and 10 mg
our position is that “no patient has failed a triptan
of metoclopramide. Her neurologic examination was
until they have failed a full dose of an injectable trip-
normal. She had no other major medical problems
tan given early in a headache.” In practice, this means a
or illnesses and was taking no regular medications.
6 mg dose of subcutaneous sumatriptan administered
She was given a subcutaneous injection of 6 mg of
as soon as the headache begins. In our experience, the
sumatriptan. Her headache disappeared and nausea
subcutaneous formulation of sumatriptan is remark-
improved, but she experienced chest pressure and neck
ably underused and is often effective in patients who
tightness lasting 15 minutes. She had no associated
have apparently “failed” triptan therapy. It is admin-
symptoms. She had an electrocardiogram that was nor-
istered using an auto-injector that can hold cartridges
mal and was referred for outpatient cardiovascular
of 6 or 4 mg of sumatriptan. The dose can be repeated
at any point after two hours if needed. The maximum
daily dose is 12 mg, which translates to two 6 mg or
three 4 mg injections in a 24-hour period. Were the electrocardiogram and
If triptans alone are not effective, or are only par-
tially effective, we commonly recommend that patients
cardiovascular investigation indicated?
use them in combination with an NSAID. (Obviously, No. This young patient at low risk of cardiovascular
however, that is not an option for the patient in the disease experienced short-lived chest pressure in close
vignette.) Naproxen sodium is available in a fixed-dose temporal relation to an injection of sumatriptan. Non-
tablet formulation with sumatriptan, but this combi- serious side effects such as flushing, paresthesias, and
nation pill is expensive. Unless patients prefer the con- transient neck or chest tightness are well-recognized
venience of a single tablet, we prefer to provide this triptan side effects. They occur often enough that they
combination of treatments using separate pills. This are referred to as “triptan sensations.” They usually
also allows for customization of the dose of both trip- begin within a few minutes to half an hour of drug
tan and the NSAID. For patients who have prominent administration and can last up to an hour, although
nausea, we may even add a third anti-nausea drug to in our experience they are typically of relatively short
this regimen, such as 10 mg of metoclopramide or a duration. For unclear reasons, triptan sensations seem
promethazine rectal suppository, a regimen we refer to to be most common in younger women such as the
as “triple therapy.” Strategies for improving treatment patient in the vignette.
results of acute headache using triptans are summa- Triptan sensations may frighten patients and doc-
rized in Table 7.1. Many of these are generalizable to tors because of their similarity to more serious car-
other forms of treatment for acute migraine as well. diac symptoms. Although cardiovascular events, some
serious, have been reported with triptan use, they are
not common. Most have occurred in patients with
Diagnosis many risk factors for coronary artery disease. Research
Migraine without aura. suggests that the clinically used doses of triptans do

Chapter 7: Errors in management of acute headache

Table 7.2. Triptan characteristics.

Generic Brand
name name Formulations Dosing Comments
Almotriptan Axert Tablet 12.5 mg
Eletriptan Relpax Tablet 20/40 mg
Frovatriptan Frova Tablet 2.5 mg Longest half-life
Naratriptan Amerge Tablet 1/2.5 mg Second longest half-life
Rizatriptan Maxalt Tablet 5/10 mg Decrease dose when used with
Orally disintegrating tablet 5/10 mg propranolol
Sumatriptan Imitrex Tablet 25/50/100 mg Also available in a fixed-dose
Nasal spray 5/20 mg combination tablet containing
Subcutaneous injection 4/6 mg sumatriptan and naproxen
Zolmitriptan Zomig Tablet 2.5/5 mg
Nasal spray 5 mg
Orally disintegrating tablet 2.5/5 mg

not cause coronary artery constriction sufficient to problem recurred with the use of oral sumatriptan, and
provoke myocardial ischemia. Furthermore, studies was troubling to the patient, it would still be worth-
of patients who have chest or neck pain with triptan while to try a different oral triptan. Although triptan
administration have not shown electrocardiographic sensations can occur with any triptan, for individual
or other evidence of reductions in myocardial perfu- patients the adverse effect profile of the different trip-
sion. Based on this evidence, we do not believe that car- tans may vary.
diovascular evaluation is warranted in young patients Since there are relatively few classes of drugs that
like the one in this vignette who have transient triptan can be used to treat individual migraine attacks, it is
sensations. important to maximize the chances of success when
a triptan is first prescribed. We find it useful to warn
Should this patient be advised to patients about the possibility of transient triptan sen-
sations before they try the drug for the first time. This
avoid triptans? reduces the likelihood that they will be frightened if
There is no safety reason for this patient to avoid the such symptoms occur and then refuse to use the drug
use of triptans in the future. Sumatriptan was effec- again – thus foreclosing use of a highly effective treat-
tive for a severe headache that had not responded ment option. For patients who are very anxious about
to nonsteroidal anti-inflammatory treatment. She is the possibility of triptan side effects, we sometimes
likely to have similar bad headaches in the future. It is suggest that they try the drug for the first time when
important not to foreclose the use of an effective treat- they do not have a headache. This allows them to expe-
ment unnecessarily. We find that many of our patients rience the effects of the drug without the added anxiety
believe the benefits of sumatriptan outweigh short- produced by a bad headache, and makes it possible to
lived side effects such as chest pain, especially once separate the side effects of the drug, if any, from the
they are reassured these side effects are not serious. symptoms of a headache.
Furthermore, most patients who are prone to triptan
sensations do not experience them every time they
take a triptan, and those who do often come to view Discussion
them as a sign that “the drug is working.” Triptans have been in widespread clinical use since
Triptan sensations may be more common in their introduction in the early 1990s, and have a good
patients who receive parenteral rather than oral trip- record of safety and tolerability. Table 7.2 lists the char-
tans, so one strategy for this patient might be to use an acteristics of the seven triptans that are available in the
oral formulation of sumatriptan. In one study about United States. All are US Food and Drug Administra-
half of patients treated with subcutaneous sumatrip- tion (FDA) approved for treatment of acute migraine
tan reported such side effects compared with roughly and sumatriptan injections are FDA approved for
a quarter who were treated with the oral drug. If the treatment of acute cluster headache. Triptans are

Chapter 7: Errors in management of acute headache

appropriate treatment for patients whose attacks do teens. Attacks were frequently related to her menstrual
not reliably respond to simple analgesics, which was cycle. During each of two pregnancies, she had been
the case for this patient. They provide good relief of free of headache. Since her second delivery, however,
pain and headache-related symptoms such as nausea. she had noted headaches with every menstrual period
For most patients, triptans have a more favorable that were moderate to severe in intensity, lasted for two
side effect profile than other choices for treatment days each, and limited her activities. Each headache
of acute migraine. Unlike opioids and barbiturate- began with a prodrome of cervical muscle spasm after
containing medications, triptans do not generally pro- which she developed unilateral throbbing frontal and
duce sedation or cognitive impairment, nor are they retro-orbital pain, increasing with physical activity and
associated with the development of abuse and depen- associated with photo and phonophobia and rare nau-
dence syndromes. Unlike NSAIDs, triptans do not sea. NSAIDs and oral sumatriptan were of incomplete
impose a risk of gastrointestinal bleeding. Overuse of benefit. She had previously been advised to try suma-
triptans may produce medication overuse headache, triptan by injection but was reluctant since she recalled
but evidence suggests that it is simpler to treat medica- being told that it was contraindicated because she was
tion overuse headache that results from triptans than also taking a low dose of sertraline. Her primary care
that resulting from overuse of opioids or barbiturates. physician had prescribed the sertraline for situational
These are important long-term considerations. depression after a recent divorce.
Unlike minor, transient side effects such as triptan sen- The headache specialist told the patient that in his
sations, these more serious adverse events must be opinion the risk of using a triptan in combination with
taken into account when deciding what treatment to a selective serotonin reuptake inhibitor (SSRI) was low.
use for individual migraine attacks. Although treat- He recommended early use of the sumatriptan injec-
ment for individual headaches is used intermittently, tion in the hopes of reducing her typical level of dis-
migraine is a chronic condition for which such treat- abling symptoms. The pharmacist, however, refused to
ment will be used repeatedly over many years. Thus, fill the prescription without contacting the physician,
adverse effects that are inconsequential when treat- citing the potential for a severe interaction between the
ment is used for a single attack can be very important sumatriptan and the sertraline. The patient was told
with long-term regular use. Conversely, side effects that this was a “class one interaction.” The patient, con-
such as triptan sensations which are more unpleas- vinced that her prior worries had been warranted and
ant within the context of an individual headache may now suspicious of the medical advice she had received,
nonetheless not be as important in the larger perspec- left the pharmacy without the prescription and can-
tive of treatment decisions. celled her return visit to the headache clinic.

Diagnosis Were the physician’s treatment

Migraine without aura and transient “triptan sensa- recommendations appropriate?
tions” with sumatriptan. A 2006 FDA warning advised caution in the use of trip-
tans with other serotonergic drugs because of worries
Tip about serotonin syndrome. Experts have called the evi-
Transient triptan sensations such as chest pressure, dence used as the basis for this warning into question.
paresthesias, or flushing are common, nonserious side For one thing, the concomitant use of SSRIs and trip-
effects of triptans. When they occur in patients at low tans is very common, yet an epidemic of serotonin syn-
risk for cardiovascular disease they do not warrant car- drome has not emerged. One study suggested that as
diac investigation or require avoidance of triptans. many as 65 million patients in a one-year period in
the United States were using an SSRI along with a trip-
tan with no documented cases of serotonin syndrome
Worries about serotonin syndrome identified among them. An American Headache Soci-
ety Position Paper published in 2010 stated that “The
Case currently available evidence does not support limiting
A 45-year-old female presented for evaluation of the use of triptans with SSRIs or selective serotonin–
headache that had begun when she was in her late norepinephrine reuptake inhibitors (SNRIs), or the use

Chapter 7: Errors in management of acute headache

of triptan monotherapy, due to concerns for serotonin Table 7.3. Selected symptoms of serotonin syndrome
syndrome (Level U). However, given the seriousness Tachycardia
of serotonin syndrome, caution is certainly warranted Hypertension
and clinicians should be vigilant to serotonin toxicity
symptoms and signs to insure prompt treatment.”
Many physicians and patients believe that men- Clonus (inducible or spontaneous)
strual migraine is often more severe and resistant to Ocular clonus (slow horizontal eye movements)
treatment than migraines occurring at other times Tremor
of the month. Early and aggressive management can Hyperreflexia
therefore be appropriate and necessary. In patients Agitation or restlessness
who do not respond to oral therapy, such as this
Hypertonicity (muscle rigidity)
one, treatment can include recommendations for an
injectable triptan. As in this case, the goal is to reduce Autonomic signs: diaphoresis, dilated pupils, flushing, increased
bowel activity
disability and improve the ability to function during
this monthly event.
above, these “stops” can lead to disruption of care for
the patient and contribute to distrust of the medical
Is there anything that the provider could establishment.
have done differently to avoid the patient The serotonin syndrome results from enhanced
leaving care? serotonergic activity in the brain usually as the result
of use of medications which can increase serotoner-
Possibly. The provider could have warned the patient to
gic activity, either alone or in combination. This results
expect advice of an interaction from the pharmacy and
in a constellation of symptoms and signs listed in
discussed the basis of his advice with her. The phar-
Table 7.3 that include restlessness, confusion, auto-
macist, if concerned, could have discussed the mat-
nomic instability, muscular irritability, tremor or
ter with the physician to understand more fully the
myoclonus, and even coma.
reasons for use of the agents before approaching the
patient; a more balanced discussion may have resulted.
Discussion Menstrually related migraine.
Many headache specialists, feeling confident about the
low risk of serotonin syndrome with concomitant use Tip
of SSRIs/SNRIs and triptans, and with limited time Good quality evidence suggests that the risk of sero-
to spend with patients, may omit discussion of this tonin syndrome is low when triptans are prescribed in
matter. However, patients who search for information combination with other serotonergic drugs. It is pru-
about these drugs online will encounter multiple sites dent to discuss the balance of benefits and harms with
which describe a “major interaction” that is “rare but patients prior to prescribing this combination of treat-
serious and potentially fatal” and typically warn that ments, however, in order to avoid misunderstandings
concomitant use should be avoided. later.
Decision support software embedded in many elec-
tronic medical records will often show a warning when A patient with sulfa allergy and
such agents are prescribed together, and presumably
pharmacy software also produces similar warnings. migraine
Accumulated evidence is quite compelling, however,
that the risk of serotonin syndrome with concomi- Case
tant use of triptans and selective serotonin reuptake A young woman with episodic, infrequent migraine
inhibitors is very low. Even after years of experience with aura not controlled with over-the-counter med-
with triptans, however, pharmacists continue to con- ications presented asking for better treatment of her
tact our office frequently to discuss the possibility individual attacks of headache. She was given a pre-
of serotonin syndrome. Unfortunately, as in the case scription for sumatriptan to manage her occasional

Chapter 7: Errors in management of acute headache

Figure 7.1 A sulfonamide group. otics and non-antibiotics may explain the lack of
S R3
R1 N More technically, the sulphonamide-containing
antibiotics are sulfonylarylamines, in which the sul-
fonamide moiety is attached to a benzene ring, and
have an amine at the N4 position. The triptans are
attacks, and appropriate use and side effects of the simple sulfonamides, in which the sulfonamide moi-
drug were discussed. The next day, the pharmacy called ety is not connected to a ring structure. Incidentally
the office to notify the provider that the patient had the intermediate group, which includes carbonic anhy-
a documented allergy to sulfa. Because the suma- drase inhibitors, has the sulfonamide moiety attached
triptan molecule contains a sulfa group (depicted in to the ring structure but without the N4 amine.
Figure 7.1), the pharmacist wanted to know if the It does appear that patients allergic to sulfon-
patient should be switched to a different triptan. amide antibiotics may in general have a greater risk
of allergy to other medications, thus creating a general
increased risk of hypersensitivity but not a specific risk
Should the patient be switched to in relation to the sulfonamide derivatives. Given the
a different medication? described structural differences, and the lack of clini-
Of the seven commercially available triptans, nara- cal reports of cross-reactivity between sulfa antibiotics
triptan, eletriptan, and almotriptan all contain sulfon- and triptans, the risk of sumatriptan use in a patient
amide groups. These sulfa moieties are folded within with sulfa allergy is likely to be low.
the molecule and are not exposed, however. They do In this case, the patient was reassured and went on
not appear to produce cross-reactivity in patients who to use sumatriptan without incident.
are allergic to sulfonamide antibacterial drugs. The
package insert for sumatriptan notes that hypersen- Diagnosis
sitivity reactions to sumatriptan are rare but may be Episodic migraine; sulfa allergy.
more likely in patients with hypersensitivity to mul-
tiple medications. In our experience, most patients
with a history of sulfa allergy can use sulfonamide- Tip
containing non-antibacterial medications without risk Although sumatriptan contains a sulfa moiety, it is
and it would be reasonable to advise the patient in this not attached to a benzene ring. Cross-sensitivity in
case to fill her prescription. If her previous sulfa reac- patients allergic to sulfonamide antibiotics is thus not
tion was anaphylactic in nature, however, we would err likely to occur.
on the side of caution and consider prescription of a
triptan such as rizatriptan which does not contain a Adverse effects associated with regular
sulfa moiety. Otherwise, the patient could be advised
not to be alone the first time she takes the medication.
long-term NSAID use
Discussion A 45-year-old woman presented with headaches that
A number of studies have examined the question had begun when she was in graduate school and which
of cross-reactivity in patients who are allergic to had gradually increased in frequency. They were now
sulfonamide-containing antibiotics, to which allergy is present more days than not. They were bifrontal, vise-
relatively common (3%), and other medications that like in quality, moderate in severity, and associated
contain sulfa moieties. The mechanism of the allergy with low-grade phonophobia but not photophobia or
to the antibacterial drugs is not completely under- nausea. The headaches were not worse with exercise
stood and all types of hypersensitivity reaction have and there was no aura. These headaches generally
been described. Many studies, however, have failed responded well to ibuprofen 400–800 mg, depending
to show clear cross-reactivity with other sulfonamide- on severity, and she had been taking ibuprofen at least
containing medications and suggest that structural dif- four days per week for as long as she could remem-
ferences between the sulfonamide-containing antibi- ber, sometimes requiring multiple doses a day. She had

Chapter 7: Errors in management of acute headache

previously been diagnosed with tension-type headache lowing NSAID withdrawal, the patient would bene-
and advised to continue ibuprofen, which she was told fit from preventive therapy to reduce the frequency
was the best treatment for tension-type headache. A of headaches and the need for treatment of individual
friend of hers recently experienced liver damage as headaches.
a result of excessive use of acetaminophen. This had Ideally, this patient’s chronic headaches would have
caused her to wonder about whether it was safe to con- been recognized as a problem many years ago, and
tinue taking ibuprofen. she might have been offered preventive therapy for
them, perhaps when she initially presented to her
physician’s office and had been diagnosed with chronic
How would you advise this patient, and is tension-type headaches. Although NSAIDs are first-
any further evaluation indicated? line therapy for treatment of many acute headache
This patient has been taking NSAIDs for many years, disorders, they are less appropriate in patients who
and is therefore at risk for some of the health con- have chronic headaches. Their use should be care-
sequences of chronic NSAID use. Gastritis is the fully monitored in patients with frequent headaches
most common adverse effect associated with NSAIDs, because daily or near-daily use can lead to medication
and all patients taking NSAIDs should be queried overuse headaches and other complications. If patients
about symptoms of heartburn or other signs of gastro- are maintained on long-term NSAID treatment, regu-
esophogeal reflux disease. Patients who are on chronic lar monitoring for development of gastric or renal side
NSAIDs should probably also have renal function effects is usually in order.
checked periodically. Lastly, NSAID use may increase
the risk of cardiovascular events in patients with heart
disease, and risk factors for heart disease should be Discussion
evaluated periodically. NSAIDs are commonly used to treat many types
Upon further questioning, this patient reported of headache. Ibuprofen and naproxen are often
progressively worsening heartburn over the last sev- used for tension-type headache or migraine, while
eral years. She did not connect it with ibuprofen use indomethacin is a first-line therapy for the rela-
as she had been using ibuprofen for many years and tively uncommon syndromes of hemicrania continua,
the heartburn had only developed recently. She had paroxysmal hemicranias, and exertional headaches.
not been to a doctor in many years because she was NSAIDs are also one of the most common over-the-
healthy aside from her headaches, and she had not had counter treatments for headaches, particularly among
lab work in the last five years. A creatinine level was patients with milder headache or those who have
obtained at her presenting visit and was 1.4 mg/dL. not sought medical treatment. Despite their nonpre-
scription status, NSAIDs are associated with sub-
stantial risks, particularly when used long term or
What treatment options are available for frequently. Important risks include gastrointestinal
this patient’s headaches, and what could events, including gastritis, peptic ulcer disease, and
have been done to prevent her kidney gastrointestinal bleeding, and an increased risk of car-
diovascular events. These two risks are listed in a black
problems? box warning on the prescription form of ibuprofen,
This patient has experienced two possible compli- and both of these risks may increase with prolonged
cations of chronic NSAID use, namely gastritis and duration of use. Indomethacin seems particularly
mild chronic renal failure. Because of these adverse prone to produce gastroduodenal toxicity.
events, she should be advised to discontinue the use of Another potential adverse effect from long-term
NSAIDs. In addition to these adverse events, the pos- NSAID use is renal papillary necrosis, which can lead
sibility that her chronic headaches are due to medica- to renal failure. It is suspected that this is related to
tion overuse should also be considered. If that is the prostaglandin inhibition by NSAIDs; prostaglandins
case, her headaches may ultimately improve once she are responsible for pre-renal vasodilation and thus
has discontinued the daily use of analgesic medications help maintain renal blood flow. Patients with pre-
(although they may worsen temporarily as a result of existing renal insufficiency, the elderly, and those tak-
drug withdrawal). If headaches do not improve fol- ing diuretics or angiotensin-converting enzyme (ACE)

Chapter 7: Errors in management of acute headache

inhibitors seem to be at the highest risk of renal com- within 15 minutes by a similar episode involving the
plications with NSAIDs. Many patients return to base- right side.
line renal function after discontinuation of NSAIDs,
but this may be more likely if the renal dysfunction Given the clinical description provided,
is discovered early. In this case, it appeared that the
patient’s renal function had been compromised for what is the likely cause of this patient’s
some time, because her creatinine remained elevated symptoms? Was the additional
despite discontinuation of NSAIDs.
prochlorperazine a good idea?
This patient’s agitation and uncontrolled muscular
Diagnosis movements were typical of extra-pyramidal side effects
NSAID-induced gastritis and mild chronic renal fail- that can occur with neuroleptic medications. Two
ure; chronic tension-type headache. types of extra-pyramidal side effects are relevant to
migraine treatment: neuroleptic-induced acute dys-
tonia results in abnormal muscle contractions with
Tip spasm or twisting of the head and abnormal move-
Chronic use of NSAIDs can be associated with gas- ments of the trunk or limbs. Episodes of dystonia
tritis, renal failure, and cardiovascular events. Risks usually last 20–30 minutes and often produce pain.
increase with duration of use and appropriate moni- One well-described form of dystonia is the oculo-
toring for safety is necessary. gyric crisis, in which the eyes deviate together to one
side. Neuroleptic-induced akathisia is characterized
by severe motor restlessness and agitation. This can
Agitation after intravenous treatment be more difficult to recognize than dystonia. Patients
for acute headache may pace the room or rock from side to side if in
bed. Dystonia and akathisia may occur separately or
Case together; typically dystonic symptoms precede those of
A 37-year-old woman with migraine was brought to akathisia, although that was not so in this case.
the ED by her boyfriend. She had been suffering from The correct diagnosis of extra-pyramidal side
a typical migraine headache that had begun the pre- effects in patients receiving neuroleptics for the treat-
vious morning. It had not responded to several doses ment of migraine is a challenge. The differential diag-
of her usual treatment of 6 mg subcutaneous suma- nosis also includes neuroleptic malignant syndrome,
triptan and oral metoclopramide. For prevention of serotonin syndrome, tricyclic antidepressant overdose,
migraines, she also took venlafaxine 75 mg daily, topi- or cocaine intoxication, so a careful review of all pos-
ramate 50 mg twice daily and lisinopril 10 mg daily. She sibilities is in order. An important management pit-
had no neurologic findings on examination and fur- fall is that the agitation and discomfort associated
ther testing was not felt necessary. An intravenous line with extra-pyramidal symptoms are easily mistaken
was placed and she was hydrated with normal saline for anxiety or for residual pain, as occurred in the case
and given 30 mg of intravenous ketorolac and 10 mg presented. Patients may then be given additional neu-
of intravenous prochlorperazine. roleptic medication in the mistaken belief that pain
About an hour after receiving this treatment, the is the problem. In hindsight, the additional 5 mg of
patient became restless and agitated. When asked how prochlorperazine was a mistake.
she was feeling she reported that she felt “terrible . . . I
just can’t explain it. I feel so bad and I can’t lie still.” How should this patient’s symptoms be
Assuming that the patient was restless because of pain, treated? Can neuroleptic-induced
an additional dose of 5 mg of intravenous prochlor-
perazine was administered. Ten minutes later the extra-pyramidal symptoms be prevented?
patient experienced a prolonged tightening of her right The traditional approach to treating neuroleptic-
sternocleidomastoid muscle. Her chin was drawn induced extra-pyramidal symptoms involves discon-
down towards the right shoulder and she developed tinuation of the offending neuroleptic drug. In addi-
slurred speech. This slowly resolved but was followed tion, anticholinergic agents such as diphenhydramine

Chapter 7: Errors in management of acute headache

or benztropine, or benzodiazepines, are often admin- migraine therapy or where such therapy is contraindi-
istered in the belief that they will cause more rapid cated. Intravenous treatment with prochlorperazine
resolution of symptoms. Anticholinergic agents are 10 mg or metoclopramide 20 mg, given with diphen-
believed to work because extra-pyramidal reactions hydramine 25 mg, is more effective than placebo for
involve an imbalance between central dopaminer- acute migraine, and three-quarters of patients receiv-
gic and cholinergic systems. The acute blockade of ing these treatments say they would want them for
dopaminergic systems by the neuroleptic drug is the- their next headache. In fact, one study suggested
orized to result in an excess of cholinergic activity. that the combination of prochlorperazine and diphen-
The dose of diphenhydramine is typically in the range hydramine was more effective than sumatriptan for
of 0.5–1.0 mg/kg and for benztropine it is 0.010– migraine in the emergency setting.
0.015 mg/kg, both administered intramuscularly. Neuroleptic-induced extra-pyramidal symptoms
Good quality evidence suggests that anticholiner- are, however, important potential side effects that must
gic medications are effective in treating neuroleptic- be kept in mind when using these drugs. These effects
induced dystonia. Evidence is conflicting, however, are more common with prolonged administration but
about whether diphenhydramine or benzodiazepines can also complicate short-term therapy. In fact, they
should be used to treat akathisia. In our view, both can appear within minutes of treatment initiation.
drugs probably work but the evidence is stronger and Their reported prevalence varies, but a reasonable esti-
more consistent for benzodiazepines. A study com- mate is that they may occur in about 20% of individ-
paring diphenhydramine and midazolam for the treat- uals who receive intravenous neuroleptic medications
ment of acute akathisia due to metoclopramide found for migraine treatment.
that midazolam was more effective than diphenhy- Diagnosis may be difficult because of the variety of
dramine, although sedation was a prominent side symptoms that can occur and their similarity to symp-
effect. A recent Cochrane Collaboration review also toms displayed with anxiety or pain. Extra-pyramidal
supports the view that benzodiazepines are effective symptoms are severely distressing for patients. Patients
for treatment of acute akathisia. who have experienced inadequately treated dystonia or
Given the discomfort associated with neuroleptic- akathisia commonly refuse to use these drugs in the
induced extra-pyramidal side effects, it is reasonable future, with the result that an important therapeutic
to wonder whether these symptoms can be prevented. option is lost.
A randomized double-blind controlled trial compared
midazolam 1.5 mg IV, diphenhydramine 20 mg IV, Diagnosis
or placebo on the occurrence of akathisia related to
Neuroleptic-induced dystonia and akathisia.
metoclopramide used for the ED treatment of nau-
sea or headache. Midazolam was effective in prevent-
ing akathisia but diphenhydramine was no more effec- Tip
tive than placebo. Thus, benzodiazepines appear to Neuroleptic medications are useful for severe
be superior to diphenhydramine for both acute treat- headaches, but clinicians should be careful not to
ment and prevention of neuroleptic-induced extra- misinterpret the extra-pyramidal side effects of dys-
pyramidal symptoms. Another study suggested that a tonia or akathisia as being due to anxiety or pain
slower infusion rate of metoclopramide (15 minutes and mistakenly treat the patient with additional
versus 2 minutes) lowered the incidence of akathisia neuroleptics.
without reducing the effectiveness of the drug for
nausea. A woman with intractable headache
and vomiting
Neuroleptic medications such as prochlorperazine, Case
metoclopramide, and chlorpromazine successfully While on call, a physician received a message to con-
relieve pain and nausea associated with acute attacks tact a patient. This 36-year-old woman had a long
of migraine. They are the drugs of choice for severe history of occasional migraine headaches that usu-
headaches that have not responded to specific anti- ally responded well to 10 mg of oral rizatriptan. The

Chapter 7: Errors in management of acute headache

previous day she had developed a typical headache, but Table 7.4. Commonly used rescue therapies
she was getting over a bout of norovirus and could not Triptans:
keep rizatriptan down. Her headache had lasted almost – Sumatriptan SC 4–6 mg
two days and she wanted to know if she should go to NSAIDs:
the ED. – Indomethacin 50 mg PO or PR (PR may be more effective)
– Ketorolac nasal spray/PO 15–30 mg

When should someone with a bad headache Sedating phenothiazines:

– Promethazine 25 mg PO
go to the emergency department? – Prochlorperazine 5–10 mg PO or 25 mg PR (PR may be more
Even patients with a well-established history of pri-
mary headaches such as migraine can have other Ergot derivatives:
– Dihydroergotamine nasal spray 0.5 mg, 1 spray each nostril q15
causes of headache. An ED visit is appropriate when min ×2
there are “red flag” signs or symptoms that sug-
gest a dangerous cause of headache. These include a – Prednisone, dexamethaxone, or methylprednisolone
headache that is unusually severe or sudden in onset –
the so-called “thunderclap” headache – or headaches Used with caution due to concern of overuse or addiction
associated with fever or confusion. In this case, the – Butalbital-containing medications
patient reported that her headache was a typical one – Narcotics
for her – what was unusual was that nausea prevented
PO, by mouth; PR, per rectum; SC, subcutaneous
her from taking her usual oral medication. When ques-
tioned further she denied high fever, confusion, or any
other alarm symptoms. She preferred to avoid a trip include injectable medications such as sumatriptan,
to the ED if possible. She was there several years ago which comes in an auto-injector that can be self-
under similar circumstances and ended up having a administered, nasal sprays, and rectal suppositories. In
CT scan of her head and a lumbar puncture, compli- the latter category, we find promethazine suppositories
cated by a severe post-lumbar puncture headache that to be particularly useful. They will usually stop vomit-
did not resolve until she had a blood patch. ing and put a patient to sleep until a headache runs its

What strategy might be helpful in treating

this headache, and prevent similar Discussion
Poorly controlled pain is an acceptable reason to go
situations in the future? to the ED, but many patients with headache dislike
Migraine is a chronic illness. Even patients whose seeking ED care. They report feeling labeled as “drug
headaches typically respond to oral treatment are likely seekers” by medical staff, and dislike the long wait-
to have occasional headaches that do not. Sometimes ing times. Most also dread the fluorescent lights and
the headache begins while they are sleeping and has loud, busy environment of the ED. Then, too, there is
progressed to the point of vomiting when they awaken. the danger of excessive diagnostic testing or investiga-
Perhaps treatment has been delayed if a headache tion as ED staff seek to “rule out” dangerous causes of
developed when they were without medication. Or headache such as subarachnoid hemorrhage. Finally,
perhaps, as in this case, circumstances made it impos- patients who visit the ED with a bad headache are at
sible to use customary treatment. Whatever the reason, risk of receiving nonspecific treatments such as nar-
a number of useful backup or “rescue” treatments may cotics, which may create expectations of future nar-
still be effective even when a patient’s first-line treat- cotic treatment that are difficult to reverse. In our expe-
ment has failed. Often these will prevent the need for rience, it is difficult to predict with certainty when or
an ED visit. whether patients with well-controlled migraine might
As a general rule, rescue treatments should be need to visit the ED for an out-of-control headache.
non-oral. Non-oral treatments take effect quickly and We find it prudent to provide a backup “plan B” rescue
will still be effective in patients who are vomit- treatment for every patient, even those who think they
ing. Table 7.4 lists common rescue treatments. They might not need it. Our advice is to “fill the prescription

Chapter 7: Errors in management of acute headache

and keep it in your medicine cabinet. You may never Table 7.5. Selected treatments for status migrainosus
need it, but if you do you will be glad it is there.” Drug and dose Comments
Dihydroergotamine (DHE) Pretreatment with an
Diagnosis 0.5–1 mg IV as a single dose/
1 mg IV q8 hours ×3 days:
antiemetic such as
metoclopramide is advisable.
Migraine in the setting of recent gastrointestinal this is often referred to as Patients may need to be
illness. the “DHE protocol” or admitted or treated in an
“Raskinizing” a patient (since outpatient infusion center in
Dr. Neil Raskin first devised order to receive the full DHE
this protocol) protocol
Tip Droperidol 2.5 mg IV every Patients should be warned of
Most headache patients benefit from having a “rescue” 30 minutes until three doses possible side effects of sedation
plan for backup treatment when their first-line strate- or patient is completely or and akathisia. Use is limited
gies fail. In most cases this backup treatment should be almost headache-free because of the potential for
prolongation of the QT interval
non-oral so that it will be effective even if the patient
Metoclopramide 10 mg IV Monitor for dystonic reactions
is vomiting.
Prochlorperazine 5–10 mg IV Monitor for hypotension,
sedation, and dystonic reactions
A never-ending migraine
Case ments. In status migrainosus, the ongoing episodes are
A 29-year-old woman presented to the ED with a typical of a patient’s previous migraine attacks, differ-
severe headache. This was her second visit to the ED ing only in duration. Sometimes, as in this case, an
in four days for the same problem. She had a long his- attack may be long-lasting because vomiting interferes
tory of intermittent migraine without aura that usu- with the ability to treat the headache, but even patients
ally responded well to oral almotriptan. A week prior who use parenteral therapies will sometimes have par-
to this visit she had developed a typical headache, but ticularly stubborn attacks of migraine.
her usual treatment was not effective, perhaps because There are no large studies of treatments for sta-
she was vomiting and could not keep the medicine tus migrainosus. Status migrainosus is usually treated
down. After three days of severe headache and persis- with parenteral medication in order to maximize
tent vomiting, she presented to the ED. bioavailability of the drug. Intravenous hydration is
Her neurologic examination was normal, and a CT also helpful to counteract the volume depletion that
scan of the head and lumbar puncture were also nor- may arise after several days of poor appetite or vom-
mal. She was treated with subcutaneous sumatriptan iting. Drugs with longer durations of action are pre-
and intravenous hydration, and the headache resolved. ferred to treat this long-lasting form of migraine. (Of
She returned home, but the headache and vomiting note, the patient in this vignette received subcuta-
recurred. The headache was bilateral, throbbing, asso- neous sumatriptan, which has a half-life of only a
ciated with phono and photophobia, and worse with few hours.) Table 7.5 lists some treatment regimens
movement. She rated her pain 10 on a 0–10 scale. Her commonly recommended for the treatment of status
neurologic examination was normal. migrainosus in the acute setting. Dihydroergotamine
(DHE) is especially useful and the “DHE protocol” of
What is this patient’s diagnosis and how repetitive parenteral administration of DHE-45 is an
especially valuable strategy. Droperidol has also been
should it be treated? employed with substantial clinical success, but its use
Migraine episodes that continue for at least three days is limited by the potential for prolongation of the QT
are termed status migrainosus. Temporary improve- interval.
ment with medication or interruption during sleep is
disregarded in determining the duration of an attack. How could migraine recurrence and return
Status migrainosus is considered a complication of
migraine. Recurrence of migraine depends on many to the ED have been prevented?
things, including the patient’s age, sex, the initial sever- About half of patients who receive ED treatment for
ity of the headache, and the choice and timing of treat- migraine will experience headache recurrence within

Chapter 7: Errors in management of acute headache

the next few days. Some, as the patient in this vignette, A man with cluster headache and
will return to the ED because of relapse. Several meta-
analyses have evaluated clinical trials studying the limited sumatriptan access
use of steroids, mostly dexamethasone, as a treat-
ment that will reduce headache recurrence. The way Case
in which steroids might reduce the risk of headache A 36-year-old man was referred for consultation
recurrence is not clear, but may relate to suppres- because of headaches that awoke him from sleep nearly
sion of neurogenic inflammation. There appear to be every night. The headaches were located behind the
few adverse events associated with single doses of right eye and were described as sharp and stabbing.
dexamethasone. They lasted about an hour and were rated 10 on a 0–10
pain scale. With the pain, he had mild nausea and a
stuffy nose on the side of the headache. The headaches
Discussion occurred daily for the last month. He had a similar
bout of headaches last fall that resolved after three
Even patients whose migraines are well controlled will
months. His primary care physician diagnosed cluster
occasionally have headaches that do not respond well
headache, gave him a prescription for the sumatrip-
to treatment and which last beyond 72 hours. A single
tan auto-injector (4 mg), and arranged for him to have
dose of dexamethasone is well tolerated and is mod-
oxygen to use at home. The patient reported that suma-
erately effective in preventing migraine relapse. One
triptan was effective for the headaches but his insur-
meta-analysis estimated a number needed to treat of
ance would only pay for only four boxes (eight doses)
9. Because steroids take some time to become effec-
a month. He could not afford to pay for it out of pocket.
tive, they are not usually used as stand-alone treatment
He had tried the oxygen but said that it “doesn’t do
in the ED. In fact, there is little evidence to suggest
much and those nasal prongs hurt my nose.”
that steroids are an effective symptomatic treatment.
The patient had not been started on preventive
Rather, they are used in addition to standard abortive
treatment to reduce the number of attacks he was expe-
therapy for the attack.
riencing. Since he had a history of bipolar disorder
Steroids are typically given intravenously under
with previous episodes of mania requiring hospitaliza-
these circumstances, although one study suggested
tion, steroids were avoided but treatment with 240 mg
that oral treatment might also be effective. Doses in
of sustained release verapamil daily was started.
clinical trials have ranged from 10 to 24 mg IV, but
higher doses (ࣙ 15 mg of dexamethasone) appear to
be most effective. For this reason, and because there
are few adverse events associated with a one-time dose Was this patient receiving optimal
of dexamethasone, we recommend that a dose of at treatment for individual attacks of
least 15 mg should be used. The importance of dose
is underscored by a recent study that used just 10 mg
of IV dexamethasone to treat status migrainosus in Subcutaneous sumatriptan is approved by the FDA for
the ED, and did not find an effect compared with treatment of cluster headache attacks. It can be self-
placebo. administered with a reusable auto-injector. Individual
cartridges containing the medication are inserted into
the auto-injector and discarded after use. When the
drug was first introduced the auto-injector cartridges
Diagnosis were only available in a 6 mg “one size fits all” dose.
Status migrainosus. Since the daily dose limit for sumatriptan is 12 mg
a day, that meant patients could use just two injec-
tions in a 24-hour period. For patients with cluster
Tip headache, who can have more than two attacks in a
The addition of a high dose (ࣙ 15 mg IV) of dexam- 24-hour period, two injections a day are sometimes
ethasone to customary ED migraine treatment pre- not sufficient to treat all headaches. Additionally, the
vents headache recurrence with a number needed to 6 mg dose produced unpleasant side effects in some
treat of 9. patients.

Chapter 7: Errors in management of acute headache

For these reasons, cartridges that contain only 4 Table 7.6. Selected studies of oxygen for treatment of
individual attacks of cluster headache
mg of the drug have been introduced. This lower dose
of subcutaneous sumatriptan is useful in a number Flow rate and
of clinical settings, although some patients still need method of
the full 6 mg dose. In this patient’s case, however, the Citation administration Results
4 mg dose reportedly worked well, so his sumatriptan Kudrow L. Response of 6 liters/minute via Oxygen aborted
treatment does appear to be optimized. It should be cluster headache attacks face mask for more than 7/10
to oxygen inhalation. 15 minutes attacks in 82%
noted that parenteral sumatriptan can also be adminis- Headache. 1981;21:1–4. of patients
tered using a “needle-less” system that propels the drug Anthony M. Treatment 8 liters/minute via 100% oxygen
through the skin using a blast of air. This system can of attacks of cluster face mask for was effective in
only deliver a 6 mg dose, however, and in our experi- headache with oxygen 15 minutes relieving pain in
inhalation. Clin Exp all 12 patients
ence is more expensive than the original auto-injector. Neurol. 1981;18:195. studied
The patient’s description of “nasal prongs” when
Fogan L. Treatment of 6 liters/minute via 56% of patients
asked about his oxygen treatment, however, suggested cluster headache. A face mask for 15 had relief of
that oxygen is not being used correctly. Incorrect oxy- double-blind minutes headache in
gen administration – too low a flow rate, use of nasal comparison of oxygen v. over 80% of
air inhalation. Arch attacks
prongs instead of a face mask, or both – is probably the Neurol. 1985;42:362–3.
cause of apparent lack of benefit in this patient’s case.
Cohen AS, Burns B, 12 liters/minute 78% of patients
When administered by face mask at a high rate of flow, Goadsby PJ. High-flow via face mask for had complete
oxygen therapy is highly effective for treatment of indi- oxygen for treatment of 15 minutes or substantial
vidual attacks of cluster headache. This patient should cluster headache. JAMA. pain relief at
2009;302:2451–7. 15 minutes
be encouraged to repeat a trial of 100% oxygen at
12 liters/minute administered via face mask for
15 minutes at the onset of headaches. using a mask at a flow rate of 7 liters/minute for
15 minutes) with sublingual ergotamine. Oxygen
How does oxygen work to abort an attack of aborted more than seven out of ten attacks in 82% of
the patients. Ergotamine worked in 70% of patients but
cluster headache? the response time for oxygen was faster, on average
The way in which oxygen works to stop an attack of 6 minutes. Two small subsequent studies also reported
cluster headache is not known. A number of theo- good results using flow rates of 6 and 8 liters/minute.
ries have been advanced, including the possibility that A larger, carefully done study used 100% oxygen at a
oxygen inhibits firing of trigeminal afferent neurons, rate of 12 liters/minute for 15 minutes. At 15 minutes
affects the parasympathetic pathway, or decreases cere- after treatment 78% of patients were pain-free or had
bral blood flow through vasoconstrictive mechanisms. substantial pain relief, compared with only 20% who
Some experts have postulated that there is a hypersen- received high flow air placebo. These studies are sum-
sitivity to oxygen in cluster headache. marized in Table 7.6. To summarize, high flow 100%
oxygen administered at 12 liters/minute for 15 minutes
at onset of an attack with a nonrebreather face mask
Discussion should be prescribed.
The first clear description of cluster headache was pro-
vided by the English physician Wilfred Harris. He
accurately described many features of the disorder, Diagnosis
including the characteristic Horner’s syndrome. The Episodic cluster headache.
first systematic study of oxygen for cluster headache
was done by Dr. Lee Kudrow, himself a cluster
headache sufferer. He was reportedly prompted to Tip
investigate oxygen because of an account of its ben- Oxygen is an effective, evidence-based therapy for
efits by an optometrist with cluster headache who acute attacks of cluster headache, but high flow admin-
used it to abort his own attacks. Kudrow’s study com- istration through a nonrebreather mask is important to
pared self-administered 100% oxygen (administered avoid treatment failure.

Chapter 7: Errors in management of acute headache

Table 7.7. FDA use-in-pregnancy safety categories

A woman with migraine planning to
Category A: Controlled human studies show no risk
attempt pregnancy Category B: No evidence of risk in humans but there are no
controlled human studies
Case Category C: Risk to humans has not been ruled out
A 30-year-old woman with a history of episodic
Category D: Positive evidence of risk in humans
migraine with aura presented for a routine follow-up
visit to the headache clinic. At that visit, she told the Category X: Contraindicated in pregnancy
physician that she was having headaches about three
times a month and sumatriptan was reliably effective between triptan use in the first trimester and major
within an hour or so. She sometimes took metoclo- congenital malformations or other adverse preg-
pramide for nausea. She was planning to try to con- nancy outcomes. Triptan use in the second or third
ceive in the next six months, but wanted to know trimester was associated with a slightly increased
what options for treatment would be available. Her risk of atonic uterus and blood loss during labor. A
migraines were debilitating before she was started on registry for sumatriptan and naratriptan sponsored
sumatriptan, and she was concerned about missing by the manufacturer included 599 women, most of
work if she was unable to take the drug when she had whom had taken sumatriptan. Although this was
headaches. a small sample size, the number of birth defects in
this group was comparable to that in the general
What options are available for treating her population. Because of these studies, some headache
specialists feel comfortable prescribing sumatriptan
migraines during pregnancy? to patients whose headaches are not controlled by the
Management of headache during pregnancy ideally therapies discussed above. While this clinical scenario
starts even before pregnancy is attempted, when non- arises rarely in our practice, we do keep this on the
pharmacologic techniques such as biofeedback can be list of options for pregnant patients with otherwise
started. Lifestyle modifications such as regular meals, intractable headaches.
adequate hydration, regular sleep, and stress man-
agement can also be very helpful in decreasing the
frequency of headache. If pharmacologic treatment Discussion
is required, acetaminophen is first-line symptomatic The FDA categorizes safety of medications during
therapy and several antiemetics are commonly used in pregnancy into Categories A, B, C, D, and X (Table
pregnancy. Metoclopramide and ondansetron have an 7.7). Category A indicates the best evidence of safety,
FDA safety in pregnancy rating of B (more on FDA rat- and is limited to drugs for which there are controlled
ings below). If these treatments are ineffective, other human studies showing no risk. Category X indi-
commonly used options include butalbital-containing cates that a medication is contraindicated in preg-
medications and opiates, though these are both FDA nancy due to a clear risk of serious harm to the fetus.
Category C. Steroids such as prednisone or dexam- There are alternative systems for evaluating the safety
ethasone can also be used. Occipital nerve blocks are a of the medication during pregnancy, including the
low-risk procedural intervention. Conservative treat- TERIS risk rating and narrative reports from propri-
ments are usually initiated first, and treatments with etary databases such as REPROTox. The frequent dis-
greater risk are added only if necessary. agreement and inconsistencies among these systems
illustrates how difficult it can be to make determina-
tions based on the limited data available.
Can she take sumatriptan? Table 7.8 lists the FDA use-in-pregnancy rat-
Although controlled trials have not been performed, ings for commonly used symptomatic treatments of
the limited available information regarding safety migraine in pregnancy. Acetaminophen is perhaps
of sumatriptan in pregnancy is reassuring. In the the most commonly used due to its B rating, but all
population-based Norwegian Mother and Child of the antiemetics also have a long track record of
Cohort Study, 1535 women had taken triptans during safety. In fact, all of the antiemetics except ondansetron
pregnancy. There was no significant association have received a TERIS rating indicating that after

Chapter 7: Errors in management of acute headache

Table 7.8. FDA use-in-pregnancy ratings of commonly used pregnancy are often lacking and rely on postmarket-
symptomatic therapies for migraine
ing registries or observational data. In patients whose
Analgesics migraines are severe, lead to frequent protracted vom-
Acetaminophen B
Aspirin and NSAIDs C (D in the third trimester)
iting and dehydration, or cause significant emotional
Butalbital C distress, the benefits of using pharmacologic therapy
Caffeine C may be felt to outweigh the potential risks. This con-
Codeine C
Hydrocodone/acetaminophen C
versation should be documented in the chart.
(APAP) The natural history of migraine during pregnancy
Meperidine C is discussed in greater detail in Chapter 8. Although
Oxycodone C
All triptans C
migraine typically improves during pregnancy, it does
not always do so and often does not improve until after
Metoclopramide B
the first trimester. For this reason, all patients planning
Ondansetron B to attempt pregnancy should have a treatment plan for
Promethazine C any headaches that do occur.
Prochlorperazine C
Chlorpromazine C
Episodic migraine without aura, planned pregnancy.
exposure during gestation, teratogenic risk is unlikely
on the basis of evidence rated at least fair to good. As Tip
described above, there is so far no evidence suggesting Many migraine treatments are contraindicated in
teratogenicity associated with the use of sumatriptan. pregnancy, but several medications with a long track
Ergots should be avoided because they decrease uter- record of safety are available. Patients should be reas-
ine blood flow. NSAIDs are also typically avoided dur- sured that treatments are available if headaches persist
ing the first trimester, due to increased risk of spon- past the first trimester.
taneous abortion, and third trimetester, due to risk of
premature closure of the ductus arteriosus and renal
Treatment of headache during
Despite the availability of symptomatic med- lactation
ications generally regarded as safe during preg-
nancy, nonpharmacologic therapy is the mainstay of Case
acute headache treatment during pregnancy. Thermal The patient in the previous case had her baby and
biofeedback, relaxation training, and physical therapy returned for follow-up. Her migraines had improved
are all effective. It may be helpful to start training in during the second half of the pregnancy, but they had
these therapies prior to attempting pregnancy, so that recurred postpartum. She was having them about three
they are available when needed particularly during the days a week, and they seemed to be the same headaches
first trimester. In addition, the importance of hydra- that she had before she was pregnant. She thought
tion should not be underestimated. Sometimes, IV flu- the increased frequency might be related to disrupted
ids and resting in a dark quiet room are all that is sleep and the stress of returning to work. For treatment
needed to help a pregnant migraineur feel significantly of acute headaches she was still using acetaminophen,
better. which she had taken during during pregnancy, but this
In developing a treatment plan, physicians should was not consistently effective.
realize that the use of pharmacologic therapy to man-
age migraines during pregnancy is best based on How is medication safety during lactation
balancing the harms and benefits for an individual
patient. The FDA package inserts for most medications determined?
state that use of the medication in pregnancy is not The American Academy of Pediatrics (AAP) Com-
recommended unless “benefit outweighs risk.” Medi- mittee on Drugs has provided a non-exhaustive list
cations are not tested on pregnant women during clin- of medications, grouped into the following categories:
ical trials, so clear data about medication safety during usually compatible with breast-feeding; effects are

Chapter 7: Errors in management of acute headache

unknown but may be of concern; require temporary Table 7.9. Medication safety during lactation
cessation of breast-feeding; associated with significant Medications usually Medications Medications
effects on some infants and should be given with cau- compatible with associated for which
tion; contraindicated. This list was last updated in breast-feeding with significant safety during
2001. Another commonly used resource is Hale’s Med- (Hale rating is in effects/give lactation is
ications and Mothers’ Milk, which is updated every two parentheses where with caution unknown/may
years. These ratings are based on large-scale clinical available) during lactation be of concern
observations or controlled trials. Safety is rated L1–L5, Propranolol (L2) Aspirin Amitriptyline (L2)
with L1 indicating the safest drugs and L5 indicating Magnesium Atenolol Nortriptyline
Riboflavin Ergotamine SSRIs
those which are contraindicated during lactation. Valproate (L2) Lithium
Verapamil (L2)
What medications can be recommended to Acetaminophen
Ibuprofen (L1)
treat this patient’s headaches while she is Ketorolac
breast-feeding? Naproxen (L3)
Good options for symptomatic therapy during lacta- Simple opioids
Sumatriptan (L3)
tion include acetaminophen, NSAIDs, and sumatrip-
tan. Eletriptan is not listed in the AAP rating system,
but has a Hale rating of L2. Steroids, including dexa-
of headaches postpartum, and another study showing
methasone and prednisone, are usually compatible
with breast-feeding. None of the antiemetics have been that breast-feeding had a protective effect. At the very
least, it is safe to say that migraine is not a contraindi-
rated using the AAP rubric. Importantly, aspirin and
cation to breast-feeding.
ergots should be avoided.
In this case, the patient was started on propra- Because the mechanism of transmission of a med-
ication to the infant during lactation is different from
nolol for preventative therapy and had reduction of
the mechanism of transmission to a fetus during preg-
headaches to once per week. Minor headaches were
treated with ibuprofen and the more severe headaches nancy, the medication safety ratings of drugs may
differ depending upon whether they are used dur-
occurring about twice per month were treated with
sumatriptan. The excretion of maternally ingested ing pregnancy or during breast-feeding. Valproate, for
drugs into breast milk is determined by a number of example, has a class X safety rating during pregnancy
(positive evidence of risk in human or animal stud-
factors, including their lipophilicity. The amount of
maternally ingested sumatriptan that is excreted into ies) but is listed as “usually compatible” with breast-
breast milk is negligible; that fact in conjunction with feeding in the AAP rating system and has a Hale’s L2
the short half-life of the drug means that it is a good
choice for treatment of acute migraine in women who Likewise, some medications perceived to be safe
are breast-feeding. Some women choose to pump and during pregnancy are contraindicated during lacta-
tion, although this is less common. One important
discard breast milk for a period of time after ingesting
medications, but that is not medically necessary when example of this is nadolol, which is probably safe when
used in pregnancy, but is lipophilic and thus con-
sumatriptan is used.
centrated in breast milk. Case reports exist of several
infants who experienced heart block when breast-fed
Discussion by a mother taking this drug.
The postpartum period is a time when many women Several factors determine medication safety during
with migraine experience increased headache fre- lactation, including the amount excreted into breast
quency. In this case, the patient identified two com- milk, the concentration of drug in the mother, and the
mon triggers for increased frequency of migraine: dis- amount of milk ingested by the infant. Table 7.9 lists a
rupted sleep and increased psychologic stress. There selection of medications often used during lactation,
is conflicting information about the effect of lactation along with their safety ratings. In general, there are
itself on migraine frequency, with one study suggest- fewer contraindications to migraine treatments during
ing that breast-feeding did not change the frequency breast-feeding then during pregnancy.

Chapter 7: Errors in management of acute headache

If women remain concerned about exposing an triptans combined with selective serotonin reuptake
infant to a specific medication, there are strategies inhibitors or selective serotonin-norepinephrine
to reduce further the infant’s potential exposure. For reuptake inhibitors: American Headache Society
example, the mother can take the medication just after position paper. Headache. 2010;50(6):1089–99.
she has breast-fed her infant or just before the infant is Shapiro RE, Tepper SJ. The serotonin syndrome, triptans,
likely to have an extended sleep period. If the mother is and the potential for drug-drug interactions. Headache.
willing to express milk, she can pump and discard the
breast milk after taking the medication in question, the
so-called “pump and dump” method. Sulfa allergies and triptan use
Platt D, Griggs RC. Use of acetazolamide in sulfonamide-
allergic patients with neurologic channelopathies. Arch
Diagnosis Neurol. 2012;69(4):527–9.
Episodic migraine without aura in a breast-feeding Strom BL, Schinnar R, Apter AJ, et al. Absence of
woman. cross-reactivity between sulfonamide antibiotics and
sulfonamide nonantibiotics. N Engl J Med. 2003;349(17):
The safety profile of the same drug may differ depend-
ing upon whether it is used when a woman is preg- NSAID side effects
nant or when she is breast-feeding. A number of med- Garcı́a Rodrı́guez LA, Barreales Tolosa L. Risk of upper
ications can be safely used to treat headaches when gastrointestinal complications among users of
women are breast-feeding, so there is no need for a traditional NSAIDs and COXIBs in the general
population. Gastroenterology. 2007;132(2):498–506.
woman to avoid lactation in order to treat headaches.
Harirforoosh S, Jamali F. Renal adverse effects of
nonsteroidal anti-inflammatory drugs. Expert Opin
Further reading Drug Saf. 2009;8(6):669–81.
Optimizing triptan use Jacobsen RB, Phillips BB. Reducing clinically significant
gastrointestinal toxicity associated with nonsteroidal
Burstein R, Collins B, Jakubowski M. Defeating migraine
antiinflammatory drugs. Ann Pharmacother. 2004;38(9):
pain with triptans: a race against the development of
cutaneous allodynia. Ann Neurol. 2004;55:19–26.
Dodick DW. Triptan nonresponder studies: implications
for clinical practice. Headache. 2005;45:156–62. Extra-pyramidal side effects with neuroleptics
Tfelt-Hansen P. Maximum effect of triptans in migraine? A Erdur B, Tura P, Aydin B, et al. A trial of midazolam vs
comment. Cephalalgia. 2008;28:767–8. diphenhydramine in prophylaxis of metoclopramide-
induced akathisia. Am J Emerg Med. 2012;30(1):
Triptan sensations 84–91.
Dodick D, Lipton RB, Martin V, et al. Consensus statement: Kelley NE, Tepper DE. Rescue therapy for acute migraine,
cardiovascular safety profile of triptans (5-HT agonists) part 2: neuroleptics, antihistamines, and others.
in the acute treatment of migraine. Headache. 2004; Headache. 2012;52(2):292–306.
44(5):414–25. Lima AR, Soares-Weiser K, Bacaltchuk J, Barnes TR.
Papademetriou V. Cardiovascular risk assessment and Benzodiazepines for neuroleptic-induced acute
triptans. Headache. 2004;44(Suppl 1):S31–9. akathisia. Cochrane Database Syst Rev. 2002;(1):
Visser WH, Jaspers NM, de Vriend RH, Ferrari MD.
Chest symptoms after sumatriptan: a two-year clinical Parlak I, Erdur B, Parlak M, et al. Midazolam vs.
practice review in 735 consecutive migraine patients. diphenhydramine for the treatment of metoclopramide-
Cephalalgia. 1996;16(8):554–9. induced akathisia: a randomized controlled trial. Acad
Emerg Med. 2007;14(8):715–21.
Tura P, Erdur B, Aydin B, Turkcuer I, Parlak I. Slow infusion
Triptans, SSRIs, and serotonin syndrome metoclopramide does not affect the improvement rate of
Evans RW, Tepper SJ, Shapiro RE, Sun-Edelstein C, Tietjen nausea while reducing akathisia and sedation incidence.
GE. The FDA alert on serotonin syndrome with use of Emerg Med J. 2012;29(2):108–12.

Chapter 7: Errors in management of acute headache

Rescue treatments for migraine for patients treated in the emergency department? A
Kelley NE, Tepper DE. Rescue therapy for acute migraine, meta-analysis and systematic review of the literature.
part 1: triptans, dihydroergotamine, and magnesium. Acad Emerg Med. 2008;15(12):1223–33.
Headache. 2012;52(1):114–28. Wang SJ, Silberstein SD, Young WB. Droperidol treatment
Kelley NE, Tepper DE. Rescue therapy for acute migraine, of status migrainosus and refractory migraine.
part 2: neuroleptics, antihistamines, and others. Headache. 1997;37(6):377–82.
Headache. 2012;52(2):292–306.
Oxygen treatment for cluster headache
Kelley NE, Tepper DE. Rescue therapy for acute migraine,
part 3: opioids, NSAIDs, steroids, and post-discharge Haane DYP, Dirkx THT, Koehler PJ. The history of oxygen
medications. Headache. 2012;52(3):467–82. inhalation as a treatment for cluster headache.
Cephalalgia. 2012;32(12):932–9.
Whyte C, Tepper SJ, Evans RW. Expert opinion: Rescue me:
rescue medication for migraine. Headache. 2010;50(2): Acute treatment of headache during pregnancy
and lactation
Treatment of status migrainosus American Academy of Pediatrics Committee on Drugs.
Colman I, Friedman BW, Brown MD, et al. Parenteral Transfer of drugs and other chemicals into human milk.
dexamethasone for acute severe migraine headache: Pediatrics. 2001;108(3):776–89.
meta-analysis of randomised controlled trials for Cunnington M, Ephross S, Churchill P. The safety of
preventing recurrence. BMJ. 2008;336(7657):1359– sumatriptan and naratriptan in pregnancy: what have we
61. learned? Headache. 2009;49(10):1414–22.
Fisseler FW, Shih R, Szucs P, et al. Steroids for migraine Loder E. Migraine in pregnancy. Semin Neurol. 2007;
headaches: a randomized double-blind two-armed, 27(5):425–33.
placebo-controlled trial. J Emerg Med. 2011;40(4):463–8. Lucas S. Medication use in the treatment of migraine
Kelly AM, Kerr D, Clooney M. Impact of oral during pregnancy and lactation. Curr Pain Headache
dexamethasone versus placebo after ED treatment of Rep. 2009;13(5):392–8.
migraine with phenothiazines on the rate of recurrent Nezvalová-Henriksen K, Spigset O, Nordeng H. Triptan
headache: a randomised controlled trial. Emerg Med J. exposure during pregnancy and the risk of major
2008;25(1):26–9. congenital malformations and adverse pregnancy
Singh A, Alter HJ, Zaia B. Does the addition of outcomes: results from the Norwegian Mother and
dexamethasone to standard therapy for acute migraine Child Cohort Study. Headache. 2010;50(4):563–75.
headache decrease the incidence of recurrent headache Erratum in: Headache. 2012;52(8):1319–20.

Pitfalls in drug therapy to prevent

8 headaches

Decisions about preventive treatment for chronic the American Headache Society (AHS) and American
headache problems are focused on answering the ques- Academy of Neurology (ANN) categorize preventive
tions of when, which, how, and for how long. That medications according to the quantity and strength
is, when should preventive treatment be considered, of scientific evidence supporting their use, but apply-
which drug should be used, how should it be given, ing this in an individual patient’s case may not be
and for how long should therapy be continued? There simple.
is no single best answer to each of these questions, but
there are some principles that can help guide decisions.
These are listed in Table 8.1. Cases in this chapter illus- A man with frequent headaches who
trate the pitfalls that can be encountered in trying to does not want daily medicine
optimize preventive drug therapy in some commonly
encountered and challenging situations. Case
The major primary headache disorders of mi-
graine, tension-type headache, and cluster headache A 36-year-old man had a long history of episodic
occur in both episodic and chronic varieties. At migraine without aura. He had a good response to
either end of the frequency spectrum decisions about sumatriptan but his physician was worried that he was
whether to use preventive treatment are relatively using it too frequently. Two months after a follow-
simple, but when headache frequency is variable or up visit at which he had received a prescription for
intermediate it can be difficult to know whether to rec- 18 100 mg sumatriptan tablets with five refills, the
ommend preventive therapy. patient called to request more sumatriptan. He said he
Decisions about which drug to use are also com- was about to go on vacation, had used all of his recent
plicated. As several cases in this chapter illustrate, the prescription plus refills, and did not want to be with-
choice of a particular preventive treatment depends out it and risk a headache while abroad. The physician
heavily upon individual patient circumstances and approved a one-time refill of nine tablets, but asked
concomitant illnesses. A particularly thorny problem the patient to come in for an appointment when he
in clinical practice is how to approach patients who returned from vacation.
have “tried everything.” Is it ever appropriate to con- At that visit the physician told the patient that
clude that preventive treatment is ineffective? Treat- sumatriptan had not been studied for long-term daily
ment guidelines issued by authoritative groups such as use and that frequent use could paradoxically make
headaches worse and lead to medication overuse
headache. He recommended that the patient consider
Table 8.1. When to consider migraine prevention taking a daily medication to cut down on the frequency
r Three or more headache episodes per week of his headaches. The patient admitted that he was tak-
r Significant interference of headache with daily activity ing sumatriptan “at least” three or four days a week, but
r Acute medications ineffective, contraindicated, or overused
r Adverse effects from acute medications said that it was working well and he had no side effects
r Patient preference for prevention from it. Furthermore, he often had three or four days
r Associated comorbidities: e.g. depression, anxiety, each week without headache. He did not like the idea
hypertension, insomnia
r Special circumstances: elderly, pregnant, pediatric populations of taking medication every day to treat a problem that
was not daily.

Chapter 8: Pitfalls in preventive drug therapy

Table 8.2. Risk factors for the development of chronic

Was the physician right to urge the use of headache
preventive medication for migraine? Risk factor Comments
There are several reasons to believe that the physician Female sex Not modifiable
in this case was right in urging this patient to consider Baseline headache frequency Potentially modifiable with
preventive therapy. This patient’s pattern of frequent preventive medication that
headache and excessive use of symptomatic medica- reduces headache frequency.
Evidence is lacking to show
tion for headache was not a recent development. The that this has a long-term
physician had long been worried about how much favorable impact on the
medication the patient was taking. If anything, the natural history of headache
patient was probably underestimating his headache Overuse of medications to Modifiable
frequency, since he used 90 tablets of sumatriptan over treat acute headache
a 60-day period. In these circumstances most doctors Obesity Modifiable; preliminary
would have suspected medication overuse and prob- observational evidence
suggests possible benefit of
ably would have discussed preventive treatment with weight loss on headache
the patient earlier. In hindsight, this case also high- frequency in migraine
lights the importance of keeping track of headache fre- Stressful life events Modifiable through changes
quency and medication use through headache diaries, in environment and training
in stress reduction and
more frequent office visits, or even pill counts. coping techniques
It seems safe to assume that the patient’s true
Depression Modifiable through drug or
headache frequency corresponded to at least 6–14 other treatments
headache days per month. Patients in this interme-
Snoring Modifiable if snoring is related
diate frequency category are at high risk of devel- to upper airway problems or
oping even more frequent headaches over time. In produces sleep apnea
the Frequent Headache Epidemiology Study, such Excessive caffeine Modifiable
patients were followed for a year. The risk of develop- consumption
ing chronic headaches over that period was strongly Based on information from Scher AI, Stewart WF, Ricci JA, Lipton
related to baseline headache frequency. The risk of RB. Factors associated with the onset and remission of chronic
daily headache in a population-based study. Pain. 2003;106:81–
chronic headache increased exponentially at around 9.
one headache per week.
Based on this evidence, we discuss the use of pro-
phylactic treatment for headaches once a patient with
migraine or tension-type headache begins regularly headache frequency. These are listed in Table 8.2.
having one or more headaches a week. This does not Although headache sufferers with these characteris-
mean that everyone who has one headache a week tics were more likely to develop chronic headache, it
must be on preventive treatment. Rather, it means that remains unclear whether the relationship is causal. In
the use of preventive treatment should be considered. other words, it is not clear whether being obese or
If a patient’s headache frequency or medication use experiencing stressful life events caused headaches to
seems to be inexorably trending upwards, this is a nat- become chronic or whether some underlying factor
ural point at which to consider the institution of pre- was associated with both events. Because of this it is
ventive therapy. uncertain whether addressing the factors that can be
modified will alter the natural history of headache.
There are, however, other important medical rea-
Is daily medication the only thing that sons to address many of these risk factors, and emerg-
might reduce the risk of chronic migraine for ing evidence indicates that at least some of them
may be causally related to the development of chronic
this patient? headache. For example, several small studies have
The Frequent Headache Epidemiology Study identi- shown improvement in headache frequency in patients
fied a number of other risk factors for the devel- with chronic migraine who have lost weight following
opment of chronic headache in addition to baseline bariatric surgery.

Chapter 8: Pitfalls in preventive drug therapy

As outlined in Table 8.2, risk factor modification Tip

can include recommendations to lose weight, per-
Patients with an intermediate frequency of headache
form regular physical exercise, learn relaxation tech-
should be closely monitored for excessive medication
niques, maintain normal body weight, use treatments
use and an increase in the number of headaches. The
to reduce headache frequency, and treat snoring or
option of preventive therapy should be discussed once
sleep apnea if present.
headache frequency is approximately once a week.

Discussion Selecting preventive treatment for

Patients at the low end of the headache frequency spec- a patient who has never used it
trum have occasional headaches that are widely dis-
persed in time. They usually prefer and benefit from Case
treatment directed to individual attacks of headache. A 27-year-old woman had been followed for sev-
Most are reluctant to consider taking medication every eral years with a diagnosis of migraine with typical
day to cope with a problem that is intermittent. This is visual aura. She was otherwise healthy and on no
understandable, since even the best preventive drugs medications. Although triptans were effective for her
for migraine do not prevent all attacks, and most headaches, she did not tolerate them because of un-
are associated with unintended adverse events. These pleasant “triptan sensations” and fatigue. Fortunately,
range from minor nuisances to potentially serious or her individual headache attacks usually responded
permanent problems. The trade-offs – slightly fewer well to a combination of 10 mg of metoclopramide and
headache days a month in exchange for a daily side 1000 mg of aspirin.
effect – may not seem worth it to patients. At a recent visit, however, her diaries showed that
In contrast, patients with high frequency head- her headache frequency had gradually increased over
aches have daily or near-daily headache. In most cases, the last year from an average of three headaches a
they should strongly consider the use of preventive month to six or seven attacks. Some attacks lasted sev-
headache treatment. Even if an analgesic or triptan eral days and she was missing work because of them.
treatment for individual attacks works well, those She had recently gone to the emergency department
drugs are not intended to be used on a daily basis. Too- for a headache that lasted three days.
frequent use of medications intended to treat indi- Her neurologic and physical examination was nor-
vidual episodes of headache can produce serious side mal. She had tried to reduce headache triggers such
effects such as gastric ulceration with nonsteroidal as erratic sleep patterns, caffeine use, and stress but
anti-inflammatory drugs, or tolerance or addiction in had not been able to decrease her headache frequency.
the case of barbiturate- or opioid-containing medica- Because her doctor warned her about medication
tions. Additionally, most drugs used to treat individ- overuse headache, she had been careful not to use
ual attacks of headache are also suspected of produc- her abortive migraine treatment more often than two
ing medication overuse headache when used too often. days a week. Instead, she had “toughed things out” but
Decisions about whether preventive treatment is war- noted that “maybe that’s how I ended up in the ED.”
ranted can be difficult to make when patients report
an intermediate headache frequency. In such situations
it is useful for patients to keep an objective record
What does this patient’s clinical course
of headache frequency in the form of a headache suggest and what treatment would you
diary. That helps both patients and doctors identify the recommend?
point at which headache frequency is so consistently
This patient had experienced a steady, gradual increase
high that it no longer makes sense to treat headaches
in headache frequency, severity, and duration. She
was experiencing disability as a result of headaches
and is worried about losing her job. She had six
or seven attacks of migraine a month, some lasting
Diagnosis several days at a time. Her headache frequency was
High frequency episodic migraine. thus close to the cutoff of 15 days a month used to

Chapter 8: Pitfalls in preventive drug therapy

Table 8.3. Summary of the 2012 American Headache Society and American Academy of Neurology treatment guidelines for the
preventive treatment of episodic migraine
Level A: Established as effective
Should be offered to patients requiring migraine prophylaxis
Drug Dose examples Common side effects or problems
Divalproex/sodium valproate* 500–1000 mg/day Known cause of birth defects; weight gain, fatigue
Metoprolol 47.5–200 mg/day Fatigue, hypotension
Petasites (Butterbur) 50–75 mg bid Classified as a nutritional supplement and not subject to
the same level of FDA oversight or scrutiny
Propranolol* 120–240 mg/day Fatigue, hypotension
Timolol* 10–15 mg bid Fatigue, hypotension
Topiramate* 25–200 mg/day Paresthesias, cognitive impairment, acute angle-closure
glaucoma, renal stones
Level B: Probably effective
Should be considered for patients requiring migraine prophylaxis
Drug Dose examples Common side effects or problems
Amitriptyline 25–150 mg/day Fatigue, dry mouth, weight gain
Fenoprofen 200–600 mg tid Gastrointestinal side effects, platelet inhibition
Feverfew 50–300 mg bid; 2.08–18.75 mg tid for Classified as a nutritional supplement and not subject to
MIG-99 preparation the same level of FDA oversight or scrutiny
Histamine 1–10 ng subcutaneously twice a
Ibuprofen 200 mg bid Gastrointestinal side effects, platelet inhibition
Ketoprofen 50 mg tid Gastrointestinal side effects, platelet inhibition
Magnesium 600 mg trimagnesium dicitrate qd Diarrhea
Naproxen/naproxen sodium 500–1100 mg/day for naproxen; 550 Gastrointestinal side effects
mg bid for naproxen sodium
Riboflavin 400 mg/day Classified as a nutritional supplement (vitamin) and not
subject to the same level of FDA oversight or scrutiny
Venlafaxine 150 mg extended release/day Agitation
Level C: Possibly effective
May be considered for patients requiring migraine prophylaxis
Drug Dose examples Common side effects or problems
Candesartan 16 mg/day Hypotension
Carbamazepine 600 mg/day Fatigue, cognitive impairment, allergic reaction
Clonidine 0.75–0.15 mg/day; patch Hypotension
formulations also studied
Lisinopril 10–20 mg/day Nonproductive cough, hypotension
Nebivolol 5 mg/day Fatigue, hypotension
Pindolol 10 mg/day Fatigue, hypotension
Flurbiprofen 200 mg/day Gastrointestinal side effects, platelet inhibition
Mefanamic acid 500 mg tid Gastrointestinal side effects, platelet inhibition
Coenzyme Q10 100 mg tid Classified as a nutritional supplement and not subject to
the same level of FDA oversight or scrutiny
Cyproheptadine 4 mg/day Fatigue, dry mouth, weight gain
* FDA approved for migraine prevention.

Chapter 8: Pitfalls in preventive drug therapy

distinguish episodic from chronic migraine. This pat- Table 8.4. Weight gain as a side effect of migraine preventive
tern of gradually increasing headaches over time was
consistent with the development of chronic migraine Associated with Associated with
and is sometimes called “chronification.” Further- weight increase Weight neutral weight loss
more, this patient had attempted to address lifestyle Beta-blockers Lisinopril, Protriptyline
and other trigger factors that might have been con- candesartan
tributing to her headache. She also had none of the Divalproex sodium Riboflavin (vitamin Topiramate
potentially modifiable risk factors listed in Table 8.2. B2)
This is a situation in which the use of preventive treat- Most tricyclic Venlafaxine
ment for migraine should be strongly considered. Her
physician suggested that she consider the use of pre-
ventive treatment for migraine and prescribed dival- were other treatments that would help her headaches
proex sodium 250 mg twice daily, stating that this without producing weight gain.
drug was US Food and Drug Administration (FDA)
approved for migraine prevention and therefore a
good treatment choice. Which preventive medications are less likely
to produce weight gain or other undesirable
What are the choices for preventive side effects?
treatment of migraine, and how do you Patient preference for specific migraine drugs and
adherence to treatment regimens is strongly influ-
choose among them? enced by drug side effect profiles. Certain adverse
Table 8.3 summarizes medications that have shown events, such as weight gain or reproductive toxicity, are
evidence of benefit in preventing migraine attacks. The particularly undesirable for young women, the group
evidence is best for Level A drugs, less strong for Level of migraine patients most likely to need preventive
B drugs, and suggestive but least compelling for Level treatment for migraine. Unfortunately, weight gain is
C drugs. Not all drugs with a high level of evidence are a common side effect of many migraine preventive
FDA approved for migraine prevention, which might treatments.
be a factor in choosing a drug. Nondrug preventive The association between migraine and weight is
treatments for migraine and other headache disor- complex. There is no evidence that overweight or obe-
ders are covered in Chapter 9, but several (such as sity increases the risk of developing migraine in the
biofeedback-assisted relaxation) have good evidence first place. In people who already have migraine, how-
of benefit and could also be considered for this patient. ever, overweight and obesity are associated with an
The choice of preventive treatment is often based increased risk of migraine progression. Migraine is
on a patient’s comorbid conditions or other med- not, however, clearly associated with an elevated risk
ications. For example, in a patient with concomi- for becoming overweight or obese, at least in women.
tant hypertension, a beta-blocker such as propranolol Divalproex sodium is not only associated with
would be a good choice for migraine prevention. weight gain, but is also a known cause of birth defects.
The patient described in this case had no other med- It should be used with caution in women of reproduc-
ical problems that might benefit from a particular tive potential, especially when there are other alterna-
drug, but she is already using aspirin for treatment tives. It was probably not the optimal choice of preven-
of individual headaches. Thus, a nonsteroidal anti- tive treatment for this patient.
inflammatory drug such as ibuprofen, ketoprofen, Are there treatment choices that would reconcile
fenoprofen, or naproxen would not have been a desir- the patient’s desire to avoid weight gain with the physi-
able choice for preventive treatment. cian’s desire to use a Level A drug that is FDA approved
The patient in this case returned three months later for migraine prevention? Table 8.4 lists migraine drugs
for a follow-up visit. Her headache diary showed a that are associated with weight gain, weight loss, or
clear improvement in headache frequency, but she had are “weight neutral.” Only topiramate has Level A evi-
gained 8 pounds and was concerned about weight gain: dence supporting its use, is not associated with weight
“I’m hungry all the time.” She asked whether there gain, and is FDA approved for migraine treatment. It

Chapter 8: Pitfalls in preventive drug therapy

may increase the risk of facial clefts in infants exposed Do you agree that this patient is refractory
during pregnancy, so this patient should be cautioned
to use effective birth control measures. to commonly used migraine preventive
While it is certainly possible that this patient is
Discussion refractory to pharmacologic preventive treatment for
Specific patient factors should be taken into account migraine, the information she has provided is not suf-
in choosing preventive medication. Although it is gen- ficient to be certain. It is clear what drugs have been
erally preferable to start with drugs that have a high tried, but there is no information about the doses that
level of scientific evidence for their use, the side effect were used or the duration of the treatment trials. It is
profiles of the drugs and patient preference should be also not clear what the patient means by “helped a lit-
considered. tle bit.” For all of these reasons, it is not obvious that
this patient with chronic migraine is truly treatment-
refractory or that previous treatment trials have been
Diagnosis inadequate.
High frequency episodic migraine with typical visual Her new physician requested her previous treat-
aura. ment records and reconstructed her treatment history.
This showed that while she had tried many of the typi-
cally used treatments for migraine, the doses used were
Tip uniformly subtherapeutic and the treatments were
Even drugs that are FDA approved for migraine pre- continued for only a few weeks before being discon-
vention and rated Level A in treatment guidelines tinued. For example, the patient was treated for just
may be the wrong treatment choice for an individual three weeks with a dose of 20 mg of propranolol twice
patient. The optimal choice of preventive treatment a day before it was discontinued and another medica-
for migraine requires a comprehensive assessment of tion begun. The physician suggested to the patient that
comorbid medical conditions, patient side effect pref- the next step in her treatment was to retry some typ-
erences, and the strength of evidence for each drug. ical migraine preventive drugs, this time making sure
to aim for the target dose of medication and continue
each drug for two to three months. She also asked the
Preventive treatment for a patient patient to keep a careful record of her daily pain level
in order to quantify any response to treatment.
who has “tried everything”
Case How could the need for retrials of
A 38-year-old woman sought treatment for headaches
she had experienced for years. She had received a diag- medication have been prevented?
nosis of chronic migraine. Following a careful his- There are only a limited number of therapies with
tory, review of her previous testing results, and phys- good evidence of benefit for migraine prevention. It
ical examination her new physician concurred with is important to make sure each drug selected is given
that diagnosis. The patient had completed a written a careful therapeutic trial. Meticulous, systematic
headache history form required by the practice. This regimens of treatment ensure that drugs are used
showed that she has tried most typically used preven- appropriately and given every chance to work. This is
tive and abortive treatments for migraine. When asked important before concluding that one or all of them
whether any had been helpful, she said she couldn’t are ineffective.
recall exactly, but then remarked that some of them It is very discouraging for patients to be told that
“helped a little bit but obviously they weren’t that good they need to go back and retry medications that pre-
or I would still be on them.” She was interested in pur- viously seemed ineffective, yet this is necessary when
suing nondrug approaches to headache, and asked if those drugs have not been used correctly. Adequate
she was a candidate for the new “migraine surgery” she patient preparation for such trials is important. Before
had read about in the newspaper. embarking on preventive treatment, we explore patient

Chapter 8: Pitfalls in preventive drug therapy

Table 8.5. Principles of migraine prevention

r Choose preventive medications with comorbid conditions in
In a patient with apparent treatment-refractory head-
r Start with a low dose and increase slowly to target dose aches, it is important to verify that previous treatment
r Continue target dose for at least 2 to 3 months trials have been of adequate dose and duration.
r Monitor with calendar or diary: goal is a 30–50% reduction in
frequency or reduced intensity/use of abortive medication
r Consider combinations of preventive treatments in refractory
Consider tapering or discontinuing medication after 4 to 6
Another patient who has tried
months of good headache control everything
expectations about the benefits of preventive treat- Case
ment. Many have unrealistic ideas about how effective A 30-year-old female with a family history of migraine
these treatments are, and expect that they will quickly and a past history of menstrual-related migraine
eliminate all headaches. We educate patients that the sought treatment for a two- to three-year history of dis-
goal for preventive therapy is to reduce, not eliminate, abling chronic migraine. Multiple treatments through-
headaches. We also stress that the treatment may take out her 20s had failed to produce improvement in her
several months to become effective, and that the dose headaches, but she had found that with exercise and
of medication may need adjustment. We emphasize the maintenance of a low stress level she could control her
need to keep a headache calendar in order to identify symptoms.
the effects of treatment, since patient global recall of In her late 20s she began a stress-filled job in
drug effectiveness is unreliable. The success of preven- her family’s large antiques business and experienced
tive therapy can be enhanced by adherence to several an increase in migraine frequency to 20 days per
principles of treatment (Table 8.5). month. She was treated sequentially with amitripty-
line, divalproex sodium, topiramate, verapamil, pro-
Discussion pranolol, venlafaxine, sertraline, and indomethacin.
Treatment guidelines can aid in the choice of which These drugs failed to provide meaningful improve-
preventive medication to use. Even in the best of cir- ment in headache frequency or severity. Onabo-
cumstances most pharmacologic treatments to prevent tulinumtoxinA injections initially appeared to reduce
migraine reduce the frequency of migraine attacks the number of headache days per month to 15, but that
by about half. In some cases preventive therapy does improvement was not sustained.
not necessarily reduce headache frequency but instead The patient treated individual headaches with
makes individual headaches milder or improves the sumatriptan, which was partially effective, but her use
patient’s response to abortive treatment. These modest of that drug was limited by insurance restrictions on
benefits can be difficult to identify in a disorder like the number of tablets she could get each month. The
migraine which naturally waxes and wanes. patient continued aerobic exercise but her headaches
Then, too, most preventive drugs do not take effect remained poorly controlled and her previous physi-
immediately. It can be a month or two before treatment cian had told her that there was nothing more that
response is apparent. The dose of medication required could be done. He suggested that the patient reduce her
to suppress headaches also varies from patient to work hours or quit her job.
patient; many of the drugs are commonly started at
low doses in order to improve tolerability. If patients
discontinue use before the target dose is achieved, Has this patient tried everything, and
the drug may appear to be ineffective. Thus, two or should she quit her job?
three months of treatment at a drug’s target dose (or
The patient’s income was important to her family, so
the highest tolerated dose) should be required before
she hoped to avoid limiting or stopping work. Her
drawing conclusions about treatment benefits.
disappointing response to previous prophylactic ther-
apy is not unusual. Most preventive medications for
Diagnosis migraine have modest benefits at best, and the bench-
Chronic migraine. mark for considering a drug trial a success is a 50%

Chapter 8: Pitfalls in preventive drug therapy

reduction in headache frequency in 50% of subjects. Table 8.6. Some commonly used combinations of preventive
treatments for migraine
Some patients do not achieve even this level of success,
however. Thus, many patients in primary and specialty Combination Comments
care report that they have “tried everything” to pre- Tricyclic anti- Patients who do not tolerate a full
vent headaches and have failed to benefit from treat- depressant plus a dose of either drug alone may do
beta-blocker well on a combination of these two
ment. In some cases previous treatment trials were of drugs in low or moderate doses
inadequate duration or a subtherapeutic dose of med-
Topiramate plus Topiramate may counteract mild
ication was used, and retrials of monotherapy must be beta-blocker weight gain that can be seen with
pursued. Once the physician is satisfied that previous beta-blockers
attempts at prophylaxis have been adequate, however, Topiramate plus Topiramate may counteract the
it is reasonable to consider using combinations of pre- antidepressant sometimes substantial weight gain
ventive medications. that can be seen with tricyclic
Combination treatment or “rational polyphar-
Beta-blocker plus Different side effect profiles may
macy” is the standard of care for other intractable ill- divalproex sodium improve tolerability
nesses such as treatment-resistant depression or hyper-
Beta-blocker plus This combination of medications
tension, so it makes sense that it should also be used in other anti- generally does not produce
refractory headache problems. When chosen carefully, depressants, e.g. synergistic side effects
combination treatment strategies may target different, venlafaxine
complementary underlying causes of headache. As a Drug therapy plus There is substantial evidence that the
general rule, drugs from the same therapeutic class nondrug therapy combination of drug therapy (e.g.
amitriptyline) and relaxation training
should generally not be combined because of the risk is superior to either treatment alone
of synergistic side effects. Rarely, some combinations OnabotulinumtoxinA OnabotulinumtoxinA generally does
can have serious side effects. The combination of dival- injections with any not produce systemic side effects so
proex sodium and topiramate, for example, has been drug or nondrug it is a good choice for combination
therapy therapy in patients with chronic
reported to cause encephalopathy. migraine who have failed
Combination therapy may also be wise in patients monotherapy. Note: it is only
with multiple medical conditions such as depression approved for prevention of chronic,
not episodic, migraine
and headache. Finally, combination therapy may allow
the use of lower drug doses and reduce the chance of
troublesome side effects. For example, a patient whose
headaches are only controlled on a high dose of beta- There is also evidence about at least one combina-
blockers that produces fatigue might benefit from a tion of treatments that does not work. A recent trial
lower dose of the drug in combination with a small examined the benefits of adding long-acting propra-
dose of a tricyclic antidepressant. nolol 240 mg/day to topiramate 50–100 mg/day in
patients with chronic migraine (average of 15 or more
days of headache a month). There was no statistically
How firm is the evidence supporting the use significant difference between the two groups in treat-
of combinations of preventive treatments ment response: those in the combination group expe-
rienced an average drop of 4 headache days per month
for headache? compared with 4.5 days per month in the group that
One disadvantage of combination therapy is the lack received topiramate alone.
of evidence from clinical trials to inform treatment Since the superiority of some treatment combina-
choices. Table 8.6 summarizes some commonly used tions has been demonstrated in clinical trials, and since
combination treatments and the rationale for their use. such treatments are commonly used in clinical prac-
Given the wide possible variety of different drug com- tice, it is clearly important to conduct additional stud-
binations and the small number of studies, there is ies to examine this treatment strategy.
insufficient evidence to evaluate the benefit of many The successful therapeutic strategy for the patient
treatment combinations. However, there is evidence of in this case ultimately turned out to be a com-
added benefit from the combination of nonpharmaco- bination of four preventive drugs: amitriptyline,
logic therapies with some pharmacologic treatments. topiramate, venlafaxine, and onabotulinumtoxinA

Chapter 8: Pitfalls in preventive drug therapy

injections. Although the use of this many drugs in She reported that her headaches had again in-
combination is unusual, it is sometimes necessary in creased in frequency and were occurring twice a week.
patients with particularly refractory headaches. After They still met criteria for migraine and her neuro-
six months on this therapy the patient was able to logic examination was normal. Her blood pressure was
work full time without missing multiple days and was 142/92 mm Hg and she had changes consistent with
no longer overusing medication to treat individual rheumatoid arthritis on examination.
headache attacks.
Are the patient’s new medications or her
Discussion elevated blood pressure producing
The goal of combination therapy for prevention of
headaches is better pain relief or fewer side effects or
both. Ideally, the pain-reducing effects of the combi- The relationship of mild, chronic hypertension to
nation will be additive or even synergistic while side headache is still debated, but most experts believe that
effects will not be. Combining treatments can increase modest elevations in blood pressure are unlikely to
the risk of adverse events, but this problem can be min- produce headache. Sudden, abrupt elevations in blood
imized by choosing the drugs carefully and by careful pressure, however, can certainly provoke headache.
monitoring and adjustment of doses. Such “malignant hypertension” is caused by severely
elevated blood pressure and is associated with signs of
end-organ damage. Malignant hypertension can occur
Diagnosis with preeclampsia, cocaine intoxication, pheochro-
Chronic migraine refractory to monotherapeutic pre- mocytoma, following severe head injury, or a num-
ventive medications. ber of other conditions. Papilledema usually should
be present to make a diagnosis of malignant hyper-
Tip tension, and the blood pressure is usually above
Before concluding that a patient has failed preven- 220/120 mm Hg.
tive treatment for migraine, combinations of preven- In this case the patient’s blood pressure is well
tive agents should be tried. below that associated with urgent or emergent levels
of hypertension. She has a long-standing history of
Preventive treatment in a patient with headaches that occurred even when her blood pressure
was well controlled. It is not very likely that her blood
multiple medical conditions pressure is contributing to her current headache prob-
lem, but her blood pressure treatment might be con-
Case tributing to her headache problem. She was recently
A 58-year-old woman sought treatment for headaches switched from a thiazide diuretic to prazosin. Pra-
that started when she was a teenager. Until her late 20s, zosin is a postsynaptic alpha-adrenergic blocker that
headaches occurred once a month, usually a few days can cause substantial vasodilation, especially when
before the onset of menstrual bleeding. They met diag- it is first begun. Medications with strong vasodila-
nostic criteria for migraine. In her 30s the headaches tory effects commonly appear on lists of medica-
increased in frequency. She was started on propranolol tions thought to aggravate or precipitate migraine. The
and headaches decreased to once a month. The drug patient has also recently been started on trazodone,
was stopped a year later. In her early 40s the patient had another drug often identified as possibly aggravating
developed rheumatoid arthritis. Despite treatment pre-existing headache problems.
with nonsteroidal anti-inflammatory drugs, the con-
dition progressed and she was treated with hydroxy- How do the patient’s comorbid conditions
chloroquine and methotrexate. The patient also affect the choice of treatment for her
developed Raynaud’s phenomenon, insomnia, and
hypertension. Her blood pressure was well controlled migraines?
on a thiazide diuretic but she had recently been Most drugs used for headache prevention were
switched to prazosin and given trazadone to help with initially developed to treat other conditions, and
sleep. their beneficial effects on headache were discovered

Chapter 8: Pitfalls in preventive drug therapy

serendipitously. In some cases the multiple effects of vasodilating properties, and also for a decline in the
these treatments can be exploited to treat more than control of her hypertension. The medication was dis-
one condition with a single drug. For this patient, a continued and she was switched to verapamil, ulti-
calcium channel blocker such as verapamil could be mately reaching a dose of 120 mg/day. Trazodone was
considered as a way to treat three of the patient’s med- replaced with melatonin. Her headaches decreased to
ical problems: hypertension, migraine, and Raynaud’s her usual frequency of once a month, and her blood
phenomenon. Updated treatment guidelines from the pressure and Raynaud’s phenomenon were under good
AAN and the AHS reviewed clinical trial evidence for control.
the use of verapamil in the prevention of migraine, and
concluded it was insufficient to make a recommenda- Diagnosis
tion for or against the use of the drug. Clinical experi-
Migraine with superimposed headache attributed to
ence, however, suggests there is a place for this drug
in selected patients such as the one in this vignette,
who stands to benefit on several fronts if treatment is
successful. Tip
On the other hand, vasoconstrictive medications In patients with headache and other medical condi-
such as triptans, ergotamine, dihydroergotamine, or tions, it is important to determine whether the medical
isometheptene-containing compounds should prob- complaints are independent of the headache or related
ably be avoided in this patient, because they can in some way, since headache can be an early manifes-
elevate blood pressure. Hypertension should be well tation of many diseases.
controlled before the use of these drugs is consid-
ered. Although the beta-blocker propranolol was pre-
viously effective for this patient’s migraine, it is not
A woman on migraine prophylaxis who
a good treatment choice now that she has developed wants to become pregnant
Raynaud’s phenomenon.
A 24-year-old woman with a history of frequent dis-
Discussion abling migraines presented to a headache clinic to dis-
Headache that coexists with other medical complaints cuss the management of migraines during pregnancy.
can be challenging to manage. Headache can be aggra- Her headaches had previously been under good con-
vated or caused by a wide variety of conditions that trol on a combination of propranolol and riboflavin
include rheumatologic conditions such as systemic for preventive therapy and she used sumatriptan to
lupus erythematosus or giant cell arteritis. In this treat individual headache attacks. Her doctor had told
patient, however, the headaches preceded the onset of her that she should be off all medications for several
rheumatoid arthritis by many years, suggesting no link months before she tried to become pregnant. He also
between the two conditions. Similarly, her hyperten- suggested that her migraines might be a contraindica-
sion was not severe and was a long-standing condi- tion to pregnancy. The patient wanted to have a child
tion. Her headaches had fluctuated in severity inde- but was afraid that if she stopped her medications as
pendently of her blood pressure control, suggesting recommended her headaches would be much worse
that her mild hypertension was an unlikely explana- and that she would be without treatment.
tion for her current headache exacerbation.
Many medications are suspected of causing or
aggravating headaches. Patients with pre-existing
What is known about the effect of
headache disorders, such as the patient in this case, are pregnancy on migraine, and does migraine
probably most susceptible to this side effect. Vasodila- increase the risk of poor pregnancy
tors such as nifedipine, prazosin, and nitrates are prob-
ably the most common offenders. A careful review outcomes?
of this patient’s medications suggested that the recent For most women with migraine, the news about preg-
switch to prazosin might be responsible for both nancy is good. Studies consistently show that about
an increase in her headaches, perhaps related to its 70% of women with migraine experience some degree

Chapter 8: Pitfalls in preventive drug therapy

of improvement during pregnancy. This is most notice- Table 8.7. FDA use-in-pregnancy ratings for commonly used
migraine preventive drugs*
able during the second and third trimesters and is most
likely in women who do not have aura. Not all women FDA use-in-pregnancy
improve, however: some experience no change in their Agent rating
headaches and a minority find that their headaches are Amitriptyline C
worse. Some women with aura find that aura frequency Botulinum toxin C
and complexity may increase. Gabapentin C
Migraine without aura does not appear to increase
Nortriptyline Not assessed
the risk of pregnancy complications or poor outcomes.
Migraine with aura, however, is associated with an ele- Propranolol C
vated risk of a number of pregnancy complications, Topiramate D
including preeclampsia and peripartum stroke. Thus, Valproate (divalproex sodium) X for treatment of migraine
women who have migraine with aura should be moni- Verapamil C
tored carefully for the development of these problems. Steroids C
A particular clinical challenge is distinguishing wors-
* FDA use-in-pregnancy ratings: Category A: controlled human
ening of migraine from an early case of preeclampsia, studies show no risk; Category B: no evidence of risk in humans
where headache can be an early and only symptom of but there are no controlled human studies; Category C: risk to
the disorder. humans has not been ruled out; Category D: positive evidence
of risk in human or animal studies; Category X: contraindicated
in pregnancy.
What options are available for preventive
treatment of migraine during pregnancy? propranolol is therefore often considered the optimal
Most women with migraine are highly motivated to choice when migraine preventive treatment is needed
minimize the use of any medications during preg- during pregnancy. In contrast, the clinical trial and
nancy, and they commonly try to limit medication other scientific evidence of efficacy is not as strong for
use to individual attacks of headache. Appropriate many of the other “C” drugs, so it is more difficult to
abortive treatments for acute migraine are discussed in assess the balance of benefits and harms. We gener-
Chapter 7. If at all possible, it is desirable to avoid the ally avoid the use of preventive medications that do
use of preventive medication during the first trimester not have substantial evidence of benefit for migraine.
of pregnancy when most organogenesis is occurring. While it might be defensible to try such treatments
If headaches have not improved by the end of the first in nonpregnant patients, the stakes are higher during
trimester this is a natural time at which to consider pregnancy and we prefer to stick with Level A or Level
starting preventive treatment. B treatments (as rated in the 2012 AHS/AAN migraine
Table 8.7 lists commonly used migraine preven- prevention guidelines).
tive medications and their FDA use-in-pregnancy rat-
ings. For both topiramate and divalproex sodium there
is good quality evidence of potential fetal risks. Top-
What about natural treatments such as
iramate is associated with orofacial clefts and dival- herbs and vitamins or behavioral
proex with neural tube defects and cognitive deficits. treatments? Are these safe to use in
Ergot derivatives such as methergine, angiotensin-
converting enzyme inhibitors, and angiotensin recep- pregnancy?
tor blockers should also be avoided in women who Biofeedback-assisted relaxation therapy has good evi-
plan to become pregnant or who are not using effec- dence of benefit for the prevention of migraine. Phys-
tive birth control methods. The majority of commonly ical therapy is less well studied but likely to be safe.
used migraine preventives are Category C drugs. We generally encourage women with migraine who are
These ratings, however, do not always take clinical planning pregnancy to consider learning biofeedback
experience into account. For example, beta-blockers techniques prior to pregnancy so they are “on board”
are in Category C (except atenolol, which is pregnancy during the first trimester. In addition, the importance
Category D) yet have a long history of safe use dur- of staying well rested and well hydrated and min-
ing pregnancy. Because it has good evidence of efficacy, imizing daily stress should not be underestimated.

Chapter 8: Pitfalls in preventive drug therapy

Worsening headaches can sometimes be effectively is a common clinical challenge. Individually tailored
managed if patients cut back on home or work respon- management of migraine during pregnancy, empha-
sibilities. sizing nonpharmacologic options, is safe and effec-
Some preventive treatments for migraine are herbs tive for many patients. Most patients with migraine
or nutritional supplements. These are typically used cope well with pregnancy and are able to manage
in very large, supraphysiologic doses. These drugs are their headaches. Poor pregnancy outcomes are not
not subject to the same level of FDA scrutiny as pre- common, so there is no good reason to discourage
scription pharmaceuticals, and since they are pur- patients in good health from undertaking pregnancy.
chased without a prescription patients do not interact Given the epidemiology of migraine, the management
with pharmacists. A worrying study done a number of of headache in the setting of pregnancy is a common
years ago had researchers posing as pregnant women occurrence.
inquire at health food stores about “safe” treatments Women with migraine should be encouraged to
for headache and nausea. A surprisingly large propor- plan their pregnancies and to come for a precon-
tion of women were advised by store workers to use ception visit prior to attempting pregnancy. This is
products that are not recommended for use during a useful opportunity to discuss a pregnancy treat-
pregnancy or that are known to be unsafe in preg- ment plan for headaches and to educate patients about
nancy. Patients may assume that these “natural” treat- the effects of pregnancy on migraine. It is also good
ments are likely to be safer than prescription pharma- practice to recommend that all women of childbear-
ceuticals, but in fact the safety of these substances in ing potential take a multivitamin with folate, which
pregnancy is not known and we recommend they be has been shown to reduce the risk of neural tube
avoided for this reason. defects.
In developing a migraine treatment plan for preg-
How long do patients need to be off of nancy, physicians and patients should weigh the poten-
tial benefits of treatment against known harms, tak-
preventive medications before becoming ing an individual patient’s situation into consideration.
pregnant? This is not always easy, since medications are not tested
Patients often ask how long before attempting preg- in pregnant women and data about pregnancy effects
nancy they need to discontinue their migraine med- are thus sparse and imperfect, often based on post-
ications, if medications are going to be discontinued. marketing registry surveillance studies or other obser-
There is no evidence that a prolonged waiting period vational data. Certainly for women whose migraines
is required. The commonly used migraine preventive are severe and lead to frequent protracted vomiting
drugs do not accumulate in the body so once they and dehydration, or cause significant emotional dis-
have been stopped no meaningful fetal exposure will tress, the benefits of using pharmacologic therapy will
occur. Because the maternal–fetal blood supply is not usually be judged to outweigh the potential risks.
connected until a week or so after fertilization, we Conversations about the harms and benefits of treat-
sometimes suggest that patients who are tracking ovu- ments during pregnancy should be documented in the
lation can take their medication until ovulation has patient’s chart.
The patient in this case decided to start Diagnosis
biofeedback-assisted relaxation training and gradually
reduce the doses of her preventive medications before Episodic migraine without aura; preconception
attempting pregnancy. If headaches were still trouble-
some at the end of her first trimester of pregnancy, she
planned to restart propranolol and use acetaminophen Tip
to treat individual headaches.
Women with migraine should not be discouraged from
attempting pregnancy and long medication-free inter-
Discussion vals are not needed prior to pregnancy. Most women
Migraine is highly prevalent during the reproductive with migraine will be able to manage headaches during
years, so the management of headache in pregnancy pregnancy and will have good pregnancy outcomes.

Chapter 8: Pitfalls in preventive drug therapy

A middle-aged woman with chronic take her topiramate regularly and “might have missed
some doses.” In the hospital she received the medica-
migraine and paresthesias tion as prescribed. After discharge, because she paid
so much attention to taking her warfarin regularly,
Case her adherence to the prescribed regimen of topira-
A 39-year-old woman returned for a follow-up visit mate remained excellent. This effectively increased the
for migraine. She had a lengthy history of chronic amount of topiramate the patient was getting and
migraine as well as asthma, Type II diabetes melli- resulted in a side effect that had not occurred at lower
tus, hypothyroidism, Factor V Leiden deficiency, and doses.
a strong family history of venous thromboembolism. Paresthesias are the most common adverse
Her headaches had always been difficult to control. event with topiramate treatment. In clinical trials
During her early 30s she had frequently missed work for migraine, 47% of subjects randomized to the
because of them. Eventually headaches improved on 100 mg/day dose of topiramate developed paresthesias
a combination of topiramate (50 mg twice daily) and compared with just 5% of those receiving placebo, and
propranolol (60 mg a day). the risk of paresthesias was dose-related. In contrast,
The patient did well for several years on this combi- concentration difficulties and memory impairment,
nation treatment. Two months before her current visit which are more widely recognized as topiramate side
she underwent left shoulder manipulation and steroid effects, occurred in just 5% and 6% of topiramate
injections in her elbow under anesthesia. She awoke subjects, respectively, compared with 1% and 2% of
from the surgery with shortness of breath and severe subjects who received placebo.
pain under her left breast. She was eventually diag- Most side effects of topiramate occur during titra-
nosed with an embolus in the left main pulmonary tion. One study concluded that if adverse events have
artery and was admitted to the hospital for anticoag- not appeared after six weeks of treatment at a stable
ulation and monitoring. dose of the drug, they are unlikely to occur. In this case,
In the hospital, propranolol was discontinued however, the patient’s dose of medication was not sta-
because of low blood pressure. The patient’s hospi- ble when she developed paresthesias. It had effectively
tal stay was prolonged by the development of bilat- been increased because of improved adherence to the
eral upper and lower extremity paresthesias. Topira- prescribed dosing regimen. Because this explanation
mate was dismissed as a possible cause because the for her problem was so quickly discounted, the patient
patient had been on the same dose of the drug for years almost underwent risky discontinuation of anticoagu-
without any problems. A neurology consultant iden- lation in order to have an invasive procedure.
tified no clear explanation for the paresthesias. Nerve
conduction studies were performed and her doctors
considered discontinuing her warfarin in order to per-
Can topiramate-associated paresthesias be
form electromyography (EMG) but ultimately decided prevented or treated?
against it. The patient asks you whether she should Starting with a low dose of topiramate, such as 25 mg
have an EMG. once daily, and working up slowly to the target
dose (usually 100 mg/day) may help minimize adverse
What was the most likely cause of this events and allow patients time to develop tolerance to
them. Paresthesias often improve over time. If they do
patient’s paresthesias? not, we find that many patients are willing to tolerate
This patient did have diabetes, a condition that them in exchange for improved control of migraines,
can produce paresthesias and neuropathy, but it is as long as they understand they are not dangerous and
unlikely that diabetic neuropathy and paresthesias will remit when the drug is stopped.
would present so suddenly. Despite the fact that this Anecdotal evidence suggests that potassium sup-
patient had been on topiramate for years without plementation may help reduce paresthesias. One
troubling side effects, topiramate was the most likely expert suggests the use of 20–40 mEq of potassium
cause for her bilateral paresthesias. When this was chloride per day. Reducing the dose of the drug can
reviewed with the patient, she recalled that prior to also be helpful, and patients sometimes find that
her recent illness she had not always remembered to although a higher dose was needed to gain control

Chapter 8: Pitfalls in preventive drug therapy

of headaches, a lower dose works well once control is Will preventive treatment work in a
established. In some cases, however, a tolerable balance
between benefits and side effects cannot be found and patient who is overusing abortive
patients discontinue it. treatments?
A meta-analysis compared adverse drug events in
trials of topiramate for migraine with adverse events Case
in trials using the drug for epilepsy. The researchers A 32-year-old man with chronic migraine was using
found that the adverse event profile of topiramate dif- over-the-counter analgesics on a daily basis. He
fered in the two conditions. Migraineurs were more reported that his headaches had gradually increased
likely than subjects with epilepsy to experience a wide over the years, and that he had responded by using
range of side effects and more likely to drop out of his prescribed and over-the-counter medications more
trials because of adverse events. For the 100 mg/day frequently. He was taking as many ibuprofen tablets
dose of topiramate, the risk ratio for paresthesias in each day as allowed by the package instructions, along
migraineurs compared to epilepsy subjects was 2.7 with several tablets a day of a combination analgesic
(99% confidence interval: 1.80–3.97). containing acetaminophen, caffeine, and aspirin. He
obtained temporary relief with these treatments but
after an hour or so experienced the gradual return of a
Discussion dull headache. He also typically awoke in the morning
This case reminds us that adherence to medications with a dull generalized headache.
can never be assumed. The patient in this case was a The patient’s neurologist advised him that he
nurse-practitioner and was certainly aware of the need should be on preventive treatment for his headaches
to take medications as prescribed. Even motivated but that preventive therapy for migraine would not
patients, however, find it difficult to take prescribed work while he was overusing short-acting medicines.
medications regularly, and adherence decreases as the He suggested abrupt discontinuation of these medica-
number of daily doses increases. tions, followed by a medication-free period of a month.
Topiramate affects the activity of some types of glu- The doctor told the patient he probably had medica-
atamate receptors, gamma-aminobutyric acid (GABA) tion overuse headache and that it was important to see
receptors, and voltage-gated sodium and calcium if headaches improved when the overused medications
channels. It is also a carbonic anhydrase inhibitor had been removed. The doctor told the patient that he
and it is this pharmacologic effect that produces the had done everything he could and that “the ball is in
reversible paresthesias seen with topiramate and other your court.” The patient was afraid that he would be
carbonic anhydrase inhibitors such as acetazolamide unable to work if he discontinued his headache med-
and zonisamide. Some researchers have suggested that ications. His doctor then said that he could offer no
the development of paresthesias is a favorable sign and further advice and the patient left care.
indicates an enhanced likelihood that migraine pro-
phylaxis will be successful.
Does this patient have medication overuse
Diagnosis It can be difficult to distinguish chronic migraine
Chronic migraine; paresthesias due to topiramate. from medication overuse headache, but it seems
likely that medication overuse is contributing to this
patient’s pattern of chronic headaches. Medication
Tip overuse occurs when a susceptible person uses exces-
When patients develop apparently new side effects sive amounts of symptomatic medication. Susceptibil-
while on a previously tolerated medication, it is ity to medication overuse headache probably varies,
worth exploring whether their compliance with rec- and there are many patients in whom the develop-
ommended dosing has changed, or whether other ment of daily headache clearly precedes the medica-
medications have been added or discontinued that tion use. With medication overuse headache, headache
might affect the side effect profile. severity and frequency and medication use increase in

Chapter 8: Pitfalls in preventive drug therapy

lock-step, so that patients may feel they are trapped tical. Additionally, abrupt discontinuation of certain
in a never-ending cycle of more medication and more medications can produce a discontinuation syndrome
headaches. This patient’s headaches have worsened in that is unpleasant or even dangerous: for example, too-
close proximity to his increased use of medication. rapid elimination of barbiturate medications can pro-
The concept of medication overuse headache is voke seizures while abrupt opioid withdrawal may pro-
controversial, and diagnostic criteria change fre- duce a flu-like syndrome. Before insisting upon “cold
quently. Several clinical features, however, are useful turkey” and rapid withdrawal of symptomatic medica-
clues to the possibility that medication is making a tions, we prefer to “meet patients half-way” by institut-
headache worse. First, the headache has either begun ing preventive therapy first followed by slow reduction
or worsened substantially in the setting of daily or in the overused medication.
near-daily use of analgesic medications such as trip-
tans, ergot derivatives, opioids, or combination anal-
gesics. (Nonsteroidal anti-inflammatory drugs are not
universally believed to produce medication overuse Despite the fact that preventive medication may be
headache.) Second, the headache usually improves and helpful even in patients who are overusing symp-
sometimes even resolves when the offending medica- tomatic medications, it is still important to address
tion is stopped. This latter criterion is particularly con- the problem of medication overuse. For some patients
troversial, however, since some experts suspect that preventive treatment will not be fully effective until
medication overuse can occasionally trigger an irre- the problem of overuse has been addressed. The pro-
versible headache increase. We also find it useful to cess of medication reduction can sometimes be expe-
inquire about whether headaches are present upon dited with the use of a burst and taper of steroids (e.g.
awakening or waken a patient from sleep, which was methylprednisolone) or (if an infusion center is avail-
the case with this patient. Overnight is the longest able) with repetitive intravenous dihydroergotamine
period of time many patients are without their med- given for several days during medication withdrawal.
ication, and their rest is often disturbed by withdrawal
headache. Diagnosis
Medication overuse headache in the setting of chronic
Is it always necessary to remove a
medication that is suspected of producing Tip
medication overuse headache before Abrupt discontinuation of overused symptomatic
medication is not always necessary to ensure response
instituting preventive treatment of to preventive treatment.
Clinical experience suggests that many patients who A woman on migraine treatment
are overusing symptomatic medication do obtain
benefit from migraine preventive medications. Once who has a fall
headache frequency decreases, those patients may be
able to decrease their use of symptomatic medication Case
and experience substantial additional improvement in A 56-year-old woman was seen for evaluation of
headaches. This clinical impression has been bolstered headaches that had started in her early teens. The
by several recent clinical trials in which even patients headaches were unilateral, throbbing, and of severe
who were overusing analgesic medications responded intensity, with associated nausea but no photo or
to onabotulinumtoxinA or topiramate treatment for phonophobia. She did not experience aura. Headaches
chronic migraine. typically lasted six to eight hours. They had occurred
While it is often desirable for a patient with refrac- roughly twice a month until about a year ago, when
tory migraine to decrease the use of symptomatic med- headache frequency increased in the setting of a new,
ication, the specifics of the clinical situation, as with more demanding job. She was experiencing two to
this patient, may make abrupt discontinuation imprac- three headaches a week. Investigation, including an

Chapter 8: Pitfalls in preventive drug therapy

MRI scan of the head, had been normal. A variety of reached. When the reasons for this “start low, go slow”
migraine preventive treatments had been tried, includ- approach are explained, many patients are grateful that
ing propranolol, divalproex sodium, and topiramate, the physician is attempting to prevent the develop-
but the patient had not tolerated them. ment of side effects (which are a significant concern
She reported that she was very sensitive to drug to patients who may be otherwise healthy). We have
side effects. On propranolol she was fatigued, on top- learned the hard way, however, that any patient being
iramate and divalproex she had memory problems, started on antihypertensive treatment for migraine
and on amitriptyline she gained weight. Because the should be cautioned about the possibility of orthosta-
patient was very worried about sedation, cognitive sis and presyncopal symptoms.
impairment, and fatigue, she was given a prescription Rapid movement from lying to sitting or standing
for lisinopril, a drug that is typically not associated is a common activity, but also a considerable physio-
with those adverse events. logic stress. Syncope can occur with immediate, tran-
Three weeks after beginning treatment with 10 mg sient drops in blood pressure or from delayed ortho-
of lisinopril, the patient awoke in the middle of the static hypotension, which can occur as late as 15 min-
night and got out of bed to use the bathroom. She utes after a change in posture.
stumbled and fell, sustaining a mild head trauma. In The risk of orthostatic drops in blood pressure
the emergency department her blood pressure was appears to increase with age, so older patients should
104/68 mm Hg. She was told that orthostatic hypoten- be particularly cautioned to take care when getting out
sion from the lisinopril might have caused her fall and of bed or rising quickly from a sitting position.
was advised to stop the drug.

Was this patient’s fall caused by her Although antihypertensives are among the best tol-
migraine treatment? erated migraine preventative medications, there are
Many things, including tripping over furniture or a several side effects that can limit use. As may have
rug, missing her footing in the dark, or even an acute occurred in this case, medication-related hypoten-
neurologic event, could have caused this patient’s fall. sion and bradycardia may contribute to symptomatic
Still, it is hard to ignore the temporal association orthostasis. Syncope or near syncope due to ortho-
between the fall and the initiation of an antihyper- static hypotension can affect patients of any age. The
tensive medication for migraine. The patient took her side effects of orthostasis and bradycardia should be of
once-daily dose of lisinopril at bedtime, so its maxi- particular concern in older patients, who are at greater
mal effect on blood pressure probably occurred dur- risk for falls and for negative sequelae of falls.
ing the night. A syncopal or presyncopal event related Although syncope or presyncope is a side effect
to orthostatic hypotension is therefore high on the list with great potential for harm, antihypertensives can
of possible explanations for her fall. The patient in this also produce other adverse events. Exercise intoler-
case did not recall any presyncopal symptoms because ance may be seen with both beta-blockers and cal-
she was groggy and sleepy. cium channel antagonists, although more commonly
with the former. Beta-blockers have been suspected
of aggravating depression, although this is controver-
Could this have been prevented? sial. In any case, patients are likely to recognize this
Although the patient was started on a relatively low if cautioned about the possibility ahead of time. Beta-
dose of lisinopril, starting at even smaller doses might blockers also may worsen active asthma or Raynaud’s
be reasonable when there is a possibility of side effects. disease. Calcium channel antagonists, on the other
Some patients, however, will forget to increase the dose hand, may produce constipation or pedal edema, as
as instructed or will become impatient with a per- well as generalized fatigue.
ceived lack of effect at lower doses. They may prema- Although the phenomenon has not been systemat-
turely discontinue therapy. ically studied, we have found that many patients with
In our experience, though, most patients are able migraine are unusually susceptible to medication side
to follow a simple titration schedule starting at a low effects. In most patients, it is prudent to start pre-
dose and increasing each week until the target dose is ventive medications at quite low doses and increase

Chapter 8: Pitfalls in preventive drug therapy

slowly as tolerated. For example, lisinopril could be about starting a preventive medication “to get me over
started at 5 mg nightly and increased to 10 mg after the hump” until she could adjust to the demands of
a week. Propranolol could be started at 20 mg nightly, her new routine. She was healthy except for border-
and the dose then increased by 20 mg every week or line increased intraocular pressures, and her blood
two if no side effects develop. This strategy of slow up- pressure was 108/74 mm Hg. Her physician consid-
titration is also recommended for patients with a his- ered prescribing topiramate but knew that it might
tory of childhood asthma. Verapamil can be started raise intraocular pressures and precipitate glaucoma,
at 40 mg nightly. Patients can also be asked to moni- so instead he prescribed amitriptyline. The patient’s
tor their blood pressure and heart rate between visits, headache decreased in frequency, but two months
although they should be cautioned that this is to assess later the physician received a call from the emer-
for side effects and not for drug benefit. There is no gency department. The patient had presented several
clear correlation between heart rate or blood pressure hours ago with acute bilateral loss of vision and eye
and headache benefits. erythema. Ophthalmologic evaluation showed acute
Although some patients are unusually sensitive to angle-closure glaucoma.
the side effects of antihypertensives, this reaction is
somewhat idiosyncratic. In fact, most patients with What happened?
low baseline blood pressure do very well when placed
Topiramate is the migraine preventive most commonly
on blood pressure medicines for the treatment of
associated with increased intraocular pressure. Topi-
migraine. Antihypertensives are frequently well toler-
ramate may precipitate secondary acute angle-closure
ated by patients concerned about possible sedation or
glaucoma due to cilio-choroidal effusion, which dis-
cognitive side effects with other migraine preventives,
places the lens and ciliary body. This can sometimes
and are an important tool in the migraine armamen-
be bilateral. It is less commonly realized that tricyclic
tarium. The key is to start low and go slow, advice com-
antidepressants or antipsychotic medications (such as
monly given when starting any medication for older
promethazine, which is sometimes used for headache-
patients, but worth remembering for younger patients
associated vomiting) can also produce angle-closure
as well.
The anticholinergic actions of tricyclic antidepres-
Diagnosis sants and some antipsychotic medications produce
Migraine; orthostatic syncope possibly due to lisino- mydriasis, which can precipitate angle closure in sus-
pril. ceptible patients. Thus, all of these drugs should be
avoided in patients known to have narrow angles. In
Tip contrast, patients with open-angle glaucoma are usu-
ally able to take these medications without adverse
Orthostatic hypotension is a possible complication
events. Many patients are not aware of the difference
of migraine prophylaxis with antihypertensive drugs.
between the two types of glaucoma, so consultation
Susceptibility to this side effect may increase with age,
with the patient’s optometrist or ophthalmologist is
so starting with low doses and warning patients about
useful if either drug is considered for patients who
falls are important.
report a history of intraocular pressure abnormalities.

An older woman with blurry vision on Discussion

amitriptyline Many medications used to prevent headaches can
occasionally be associated with ocular adverse events.
Case Lithium can produce eye irritation and keratoconjunc-
A 66-year-old woman with a long-standing history tivitis, probably because of effects on sodium trans-
of migraine without aura reported an increase in port. In addition to producing angle-closure glaucoma
headache frequency which she attributed to family in predisposed patients, tricyclic antidepressants (and
stress from her daughter’s divorce. She had begun to even selective serotonin reuptake inhibitors [SSRIs])
care for her young grandchild, and described this as cause mydriasis and accommodation problems that
“simultaneously delightful and stressful.” She asked may result in blurred vision. This is often transient.

Chapter 8: Pitfalls in preventive drug therapy

Ocular dystonias, eye movement disturbances and Several lines of evidence suggest topiramate may
changes in color or contrast discrimination have rarely produce bone loss. Bone loss is a known complica-
been associated with the use of topiramate or SSRIs. tion of many antiepileptic drugs. The risk has been
best studied with older antiepileptic drugs such as
divalproex sodium. Less evidence has accumulated
Diagnosis about the risk of bone loss with newer antiepilep-
Acute angle-closure glaucoma from amitritpyline. tic agents such as topiramate. However, a recent pilot
study evaluated biochemical and radiologic tests of
Tip bone metabolism in women with migraine taking topi-
ramate, and found osteopenia in 53% of them. The risk
Potentially serious ocular complications are well rec-
seemed to be associated with duration of use but not
ognized with topiramate therapy but can also occur
with other commonly used migraine preventive drugs,
A risk of bone loss with long-term topiramate treat-
particularly tricyclic antidepressants.
ment is plausible because it is a carbonic anhydrase
inhibitor. The mild metabolic acidosis it produces may
A woman with migraine worried about have an effect on bone metabolism and turnover. This
is concerning because migraine is a condition of long
the long-term safety of topiramate duration and most patients requiring long-term pro-
phylaxis for it are women.
A 33-year-old woman presented to establish care with
a local neurologist after recently moving to the area. Discussion
She had a long-standing history of migraine without
The package insert for topiramate recommends
aura and had been started on topiramate five years ago
obtaining baseline and periodic serum bicarbonate
by her previous headache specialist. Trials of several
levels in patients using topiramate, based on a correla-
other preventive treatments had not been successful,
tion between chronic metabolic acidosis and fatigue. It
but with topiramate her headaches decreased from a
is unclear, however, whether serum bicarbonate levels
frequency of three times a week to three times a month.
can be used to indicate a risk for bone loss. A number
Her breakthrough headaches were effectively treated
of strategies have been proposed to minimize or
with a triptan. She did not want to change her treat-
prevent bone loss in patients treated with antiepileptic
ment regimen, but she wanted to know if it was safe to
drugs. At present, however, evidence is insufficient to
take topiramate long term.
make strong recommendations for these proposals,
which include:
What is known about the long-term risks of r Measurement of bone mineral density at the start
chronic topiramate treatment? of treatment and periodically on an individualized
Both cognitive side effects and paresthesias are com- r Monitoring of vitamin D (25-OHD) levels so that
mon side effects of topiramate. They often lead to treat-
vitamin D supplementation can be adjusted on an
ment discontinuation, but they are entirely reversible.
individualized basis.
Even the rare but feared complication of acute angle- r Recommendations for adequate calcium intake
closure glaucoma usually resolves if the drug is dis-
and vitamin D supplementation with 400–
continued promptly. The incidence of nephrolithiasis
2000 IU/day as prophylaxis.
in patients taking topiramate was approximately 1%
per year in clinical trials; thus, the cumulative risk Many patients have questions about the long-term
increases with duration of therapy but there is no evi- safety of migraine preventive treatment. This is under-
dence that the risk of kidney stones persists after the standable because most people with migraine are oth-
drug is stopped. In contrast, bone loss is painless and erwise healthy. Migraine itself confers little long-term
insidious, and the pharmacologic profile of topiramate health risk, unlike other conditions such as epilepsy
does suggest that this may be an irreversible complica- for which long-term preventive treatments might be
tion of long-term topiramate treatment. used. Thus the balance of benefits and harms of

Chapter 8: Pitfalls in preventive drug therapy

long-term treatment with drugs such as topiramate depends upon the dose of topiramate: estrogen levels
will be weighed differently depending upon whether decreased by an average of 21% when used with a daily
they are used for migraine or epilepsy. dose of 400 mg of topiramate, and 30% with a dose of
800 mg/day. This interaction is substantial enough to
interfere with the contraceptive efficacy of the newer
Diagnosis low estrogen dose combination contraceptives that are
Chronic migraine. now in common use.
When used for migraine prophylaxis, however,
topiramate is rarely given in doses above 150 mg/day;
Tip a more typical dose is 100 mg/day. One study showed
Most of the troublesome side effects of topiramate that healthy volunteers taking topiramate at doses of
are reversible when treatment is stopped, but patients 50–200 mg/day did not experience statistically signifi-
should be cautioned about the possibility of long-term cant decreases in ethinyl estradiol concentrations. On
treatment-related bone loss. the basis of these data, which are described in the
package insert, it is very unlikely that topiramate pre-
scribed at doses less than 200 mg/day will decrease
A woman worried about whether her the efficacy of combination estrogen–progestin con-
contraceptive will work when she is on traceptives. For patients who might be imperfectly
compliant with oral contraceptive regimens, however,
topiramate one might consider the use of a combination hor-
monal contraceptive with a slightly higher dose of
Case ethinyl estradiol. In this case, the patient had a neg-
A 21-year-old woman with chronic migraine called ative pregnancy test and a normal menstrual period
her physician to report that she had missed her last two weeks later. The patient decided to continue topi-
menstrual period. She was taking an oral combina- ramate, but changed to a combination hormonal con-
tion hormonal contraceptive for contraception and traceptive with 30 ␮g of ethinyl estradiol “for peace of
menorrhagia. Six months ago she had started topi- mind.”
ramate. When she went to the pharmacy to fill the
prescription, the pharmacist told her there was an
interaction between topiramate and oral contracep- Diagnosis
tives that might increase her risk of becoming preg- Chronic migraine.
nant. At the time, her physician had reassured both
her and the pharmacist that this interaction was not Tip
clinically important with typical anti-migraine doses At doses less than 200 mg/day, topiramate does not
of topiramate. However, in view of her missed period interfere with contraceptive efficacy.
the patient was concerned that perhaps the pharma-
cist had been right, and that she might now be preg-
nant. Although the physician had used these two med- An older man treated for tension-type
ications in many patients over the years, she was also headache who has a seizure
What is the interaction between topiramate A 66-year-old man with history of hypertension,
hyperlipidemia, and peptic ulcer disease was referred
and combination hormonal contraceptives? to the headache clinic to discuss treatment for his
Topiramate decreases serum concentrations of ethinyl migraines. Headaches had become more frequent over
estradiol, the active estrogen component in many com- the last year, and he attributed this worsening to emo-
bined hormonal contraceptives. Studies show that the tional stress and worry about his other health prob-
blood concentration of ethinyl estradiol is reduced by lems. His doctor had started him on amitriptyline
an average of 18% in women who are also taking top- 25 mg at bedtime, and his headache frequency had
iramate 200 mg daily. The decline in estrogen levels decreased from three times a week to once a week. He

Chapter 8: Pitfalls in preventive drug therapy

Table 8.8. Some headache treatments that may lower the

seizure threshold
Certain medications are more frequently associated
Category Drug Example
with drug-induced seizures than others. In a recent
Antidepressants Tricyclic antidepressants
SSRIs and selective serotonin–
retrospective analysis, the drugs most frequently asso-
norepinephrine reuptake ciated with seizure induction were bupropion (23%),
inhibitors (SNRIs) diphenhydramine (8.3%), tricyclic antidepressants
Local anesthetics (if systemic Lidocaine (7.7%), tramadol (7.5%), amphetamines (6.9%), isoni-
absorption occurs) Bupivacaine azid (5.9%), and venlafaxine (5.9%). Since a large num-
ber of the events analyzed in this study were described
Analgesics Fentanyl as suicide attempts, it is probably that the risk is dose
Pentazocine related as well as drug related. Overall, about 18% of
Propoxyphene seizure cases involved the use of more than one medi-
Tramadol cation known to lower the seizure threshold.
Neuroleptics Clozapine
Butyrophenones Diagnosis
Other Anticholinergics Probable medication-induced seizure in a patient
Anticholinesterases taking multiple medications with epileptogenic side
Baclofen effects.
Hyperbaric oxygen
also reported improved sleep. At this visit he reported The use of multiple epileptogenic medications should
continued worry about the migraines he still had, how- be avoided when possible, especially in older patients.
ever, because he had not found any effective treatment
for the infrequent but very severe events. Triptans were A fatigued patient on two blood
contraindicated because he had coronary atheroscle-
rosis, and he could not take anti-inflammatory drugs
pressure medicines
such as aspirin because of a previous problem with
gastric ulcers. His physician thus prescribed tramadol.
Three weeks later the patient was admitted to the hos- A 47-year-old woman was being treated for intractable
pital for evaluation after suffering a witnessed, gener- chronic migraine. She had previously experienced a
alized tonic–clonic seizure. small reduction in headache frequency with propra-
nolol and had remained on a dose of 120 mg a day.
Multiple other trials of preventive medications had
What happened? failed to produce improvement, so verapamil was
Medications that lower the seizure threshold are added at a dose of 40 mg orally three times a day. At her
thought to cause about 6% of new-onset seizures. follow-up visit she reported profound fatigue, which
Many medications associated with an increased risk she attributed to her medications. She was having trou-
of seizures are used to treat migraine and some are ble getting out of bed and getting to work. On exami-
listed in Table 8.8. The epileptogenic effects of med- nation, her blood pressure was 95/62 mm Hg and her
ication are not often taken into consideration when pulse was 48 beats per minute.
prescribing treatment for patients with no prior his-
tory of or risk factors for seizure. In clinical prac-
tice, drug-induced seizures are rare in patients who are
What was the likely cause of her fatigue?
using a single medication that lowers seizure thresh- Nonselective beta-adrenergic blockers, including pro-
old. In contrast, the risk of inadvertent seizures is pranolol, inhibit sympathetically mediated increases
higher in patients who are prescribed a combination of in heart rate and cardiac contractility. Calcium channel
such medications. Older patients may be particularly antagonists such as verapamil reduce atrioventricular
vulnerable. nodal conduction and cause peripheral vasodilation.

Chapter 8: Pitfalls in preventive drug therapy

Used together, these medications may produce a syn- Table 8.9. Combinations of preventive headache medications
that should be avoided
ergistic decrease in cardiac conduction, heart rate, and
contractility. This can lead to symptoms such as fatigue Combination Comment
that are caused by bradycardia and reduced cardiac Any combination of An exception could be made for
output. Peripheral edema also may develop as a result antihypertensives patients with treatment-resistant
of vasodilation.
Caution should be exercised when prescribing Two or more Gabapentin might be an exception to
antiepileptic drugs this rule because it is generally well
two migraine preventive medications with overlapping tolerated both alone and when used
mechanisms of action. Another example of migraine with other drugs
preventive medications where a deleterious overlap Topiramate and A recent randomized trial
in effect might be expected in combination would be propranolol demonstrated no benefit for the
the use of beta adrenergic blockers with angiotensin- combination compared with
topiramate treatment alone
converting enzyme inhibitors or angiotensin recep-
Two or more In addition to synergistic side effects of
tor blockers (such as lisinopril or candesartan, respec- antidepressants agitation or somnolence, there is a risk
tively). This combination would also be likely to of serotonin syndrome with this
lower blood pressure significantly and perhaps pro- combination. Psychiatrists occasionally
use a full dose of an SSRI or SNRI with a
duce fatigue or syncope. small dose of a tricyclic antidepressant
Combinations of antiepileptics used for migraine at night to help with sleep
prevention may also produce synergistic adverse
events. Concurrent use of sodium valproate and top-
iramate, for example, may increase the risk of hyper- Diagnosis
ammonemia. The concurrent use of valproate and Profound fatigue related to drug-induced hypotension
lamotrigine may increase the risk of development of and bradycardia.
Stevens–Johnson syndrome. In contrast, gabapentin
can often be combined with other antiepileptic drugs Tip
without causing intolerable side effects. Its useful- Avoid prescribing drugs with overlapping side effect
ness for migraine prevention, however, is not set- profiles.
tled. Recent guidelines for preventive migraine ther-
apy issued by the AAN and the AHS concluded that How about Botox for episodic
current evidence is insufficient to recommend for or
against use of gabapentin for migraine prophylaxis. migraine?
Nonetheless, as with other drugs in that category, the
drug is still in widespread use for migraine prophylaxis Case
based on clinical impressions of benefit. A 42-year-old woman had a 20-year history of
migraine without aura, with headaches consistent with
migraine occurring about eight days a month for the
Discussion last three years. On the advice of a friend she began
The use of medications in combination has been sug- treatment for her headaches with intramuscular injec-
gested in the management of refractory migraine and tions of 155 units of onobotulinumtoxinA every three
in most instances is considered safe and well toler- months. After one year of therapy she had not noted
ated. The potential complications of such combina- significant improvement. She presented for a second
tions, though, should be borne in mind. Anticipating opinion at a headache center; she suspected that the
drug side effects is important. Many potential syner- lack of improvement was due to an inadequate dose
gies can be suspected based on what is known about of botulinum toxin. She requested that the headache
a drug’s mechanisms of action. Since most drugs have center take over the treatment and requested a letter
multiple actions, however, this is not always the case. to her health insurance company in support of con-
Table 8.9 lists combinations of preventive medications tinued therapy using 300 units of the drug every three
that should be used with caution due to the potential months. She was not overusing analgesic medications
for overlapping side effects. and her past medical history was otherwise notable

Chapter 8: Pitfalls in preventive drug therapy

only for a mild anxiety disorder. Her neurologic exam- haps combined with lifestyle management including
ination was normal and her body mass index (BMI) exercise and weight control, along with attention to
was 33. any comorbid conditions such as anxiety that may be
aggravating her headaches.
Is onabotulinumtoxinA the best therapeutic
recommendation for this patient’s pattern of Discussion
headache? The FDA approved onabotulinumtoxinA for treatment
Although the use of onabotulinumtoxinA is supportd of chronic migraine in 2010. International Classifi-
by an FDA indication for chronic migraine, it is not cation of Headache Disorders (ICHD)-3 beta crite-
approved for treatment of episodic migraine. A large ria for chronic migraine require the presence of more
number of randomized clinical trials have failed to than 15 headache days a month, with a headache
show benefit in episodic migraine and AAN treatment day defined as one with headache for more than four
guidelines currently label onabotulinumtoxinA ther- hours. Headaches must have features of migraine, or be
apy as “probably ineffective” for treatment of episodic relieved by the use of a migraine-specific drug (triptans
migraine. The likely explanation for the lack of bene- or ergots) on at least eight days a month.
fit of this therapy in the patient described in this case The approved treatment protocol for onabo-
is that this is simply the wrong choice of treatment for tulinumtoxinA for chronic migraine is the administra-
her condition. tion of 155 units intramuscularly in 31 fixed sites. The
It may have seemed logical to use onabotulinum- drug is diluted with saline, and individual injections of
toxinA therapy; one argument might have been that 5 units are made in the muscles depicted in Figure 8.1.
this was a safe therapy for use in a motivated patient Subsequent analysis further suggests that patients
with high frequency migraine who might have been with concomitant medication overuse derive a simi-
in the process of transformation to a chronic pattern. lar benefit from onabotulinumtoxinA therapy to those
Why not, it could be asked, just skip all the systemically without, and that in many patients the clinical effect
active preventive medication trials with their possible increased over subsequent injection cycles. Reported
associated side effects and jump all the way to onabo- adverse events overall leading to study discontinuance
tulinumtoxinA treatment? There are several problems were low (3.8%) in the clinical trial treatment groups.
with this approach. The first, as noted, is that there is Neurotoxins appear to be taken up in peripheral
high quality evidence that the drug is not effective for motor neurons, where they disrupt the protein com-
episodic migraine. Why this is so is not entirely clear, plex that facilitates vesicle fusion and release from
but may be tied to its mechanism of action and the the axonal endplates. Neurotoxins also may affect sen-
processes that come into play when migraine trans- sory neurons by inhibiting release of proinflamma-
forms from an episodic to a chronic form. In part, tory mediators at several sites within the neuron. More
the mechanism of action of onbotulinumtoxinA in specifically, onabotulinumtoxinA could suppress neu-
chronic migraine may relate to peripheral inhibition rogenic inflammation near the scalp or facial injec-
of neurotransmitter and neuropeptide (e.g. substance tion site by preventing the release of the neuropep-
P, calcitonin gene-related peptide [CGRP], glutamate) tides CGRP and substance P from free nerve endings
release which is related to inhibition of central sensiti- that provide sensory innervation to the skin and mus-
zation. In episodic migraine patients, whose disorder cles. In addition, neurotoxins may exert central effects
is unlikely to involve sustained central sensitization, by blocking the release of CGRP and glutamate from
onabotulinumtoxinA may not work. nociceptive nerve fibers in the spinal cord, leading to
a suppression of stimulation of second-order neurons
What therapy should be offered to that are theoretically associated with the maintenance
of central sensitization and pain.
the patient? The exact mechanism of onabotulinumtoxinA in
This patient should be advised that onabotulinum- chronic migraine is unknown, and perhaps multiple
toxinA therapy has been shown to be ineffective for mechanisms are involved, as above, including: inhi-
her condition. Instead, a combination of abortive and bition of neurotransmitter release from both motor
preventive medications should be considered, per- neurons and sensory nociceptive neurons, leading to

Chapter 8: Pitfalls in preventive drug therapy

Figure 8.1 Paradigm for injection of onabotulinumtoxinA for chronic migraine.

a reduction in both peripheral and central sensitiza- the treatment of episodic migraine headache. Its use
tion; direct antinociceptive effects through blockage should be restricted to patients with well-documented
of CGRP and glutamate release; and lastly possibly chronic migraine.
a direct muscle relaxant effect that may also play
a role in headache management in some patients.
The most likely mechanism of action is that the A young woman with headache after
preventive effect of onabotulinumtoxinA in chronic
migraine is due to its ability to inhibit motor neu-
head trauma
ron overactivity and sensory neuron hyperexcitabil-
ity, leading to suppression of peripheral and central
sensitization. A 19-year-old woman presented to the headache clinic
One barrier to use of onabotulinumtoxinA is with her mother to discuss persistent headaches after
cost, although this may be partially offset by sav- head trauma. The initial head injury had occurred one
ings through reduced usage of expensive abortive year ago while she was playing soccer and hit the shoul-
agents such as triptans and by fewer emergency der of another player with her head. There was no loss
department visits by those who respond to the treat- of consciousness but she immediately developed dizzi-
ment. Preapproval from the patient’s health insurer ness, nausea, holocephalic headache, and balance dif-
is often required, a process that adds work and can ficulties. She was taken off the field and evaluated by
significantly delay the initiation of therapy. Onabo- her team trainer, who diagnosed her with a concus-
tulinumtoxinA is not interchangeable with other sion. The concussive symptoms improved over the next
botulinum neurotoxins and at the time of this writ- three months without medication and she was able to
ing is the only FDA-approved neurotoxin for chronic return to play.
migraine. Several months later, during a spring-break trip,
she sustained a second head trauma while off-roading
on an all-terrain vehicle (ATV). She was not wear-
Diagnosis ing a helmet and her ATV overturned, resulting in
Episodic migraine without aura. head trauma with loss of consciousness for around a
minute. Her symptoms, including headache, imme-
diately returned and have been persistent since. The
Tip headaches are present daily, usually better in the morn-
Though useful in chronic migraine, onabotulinumtox- ing and worse in the evening. Activity, concentration,
inA (Botox) is not indicated for and is not useful in trying to read, sustained visual focus, loud noise, and

Chapter 8: Pitfalls in preventive drug therapy

bright lights all worsen the headache. The pain may In this case, the patient had headaches with
reach 8/10 and can be pulsing or throbbing, though migrainous features and the care plan was modeled
still holocephalic. She denies any aura. She has been after treatment of migraine. She was started on topira-
tried on amitriptyline without any benefit. mate with partial improvement, and had nearly com-
plete resolution after the addition of low-dose propra-
nolol. A trial of sumatriptan was also successful for
What factors predisposed this patient to the management of individual severe headaches.
development of post-traumatic headache?
In a recent multicenter study done to determine fac- Discussion
tors associated with headache after traumatic brain
The incidence of TBI is increasing in many popu-
injury (TBI), the most significant risk factor for post-
lations, particularly among soldiers returning from
traumatic headache (PTH) was a pre-injury diag-
combat and young athletes. Increased recognition of
nosis of migraine or other headaches. Penetrating-
TBI in athletes may also be contributing to the rise
type TBI and female sex were also associated. In
in diagnosis. Headache is the most common symp-
a pediatric study, adolescents were more likely to
tom accompanying concussion, is often the first post-
develop post-concussive headache than younger chil-
concussive symptom to develop, and is typically the
dren. Other factors associated with development of
last to resolve. PTH does not have any specific clini-
chronic PTH include mild (rather than moderate or
cal criteria, and the diagnosis is made based on a tem-
severe) head trauma and low educational and socio-
poral relationship to head trauma. Currently, the diag-
economic status. While not formally studied as a risk
nosis is made if the headache develops within seven
factor for headache, repeat head trauma often causes
days of head or neck trauma, although many providers
more severe post-concussive symptoms.
who work with TBI patients describe development of
Upon further questioning, the patient denied a his-
headache with a longer delay. This is a subject of some
tory of headache prior to the injury. She was very prone
debate in the headache community.
to motion sickness both as a child and through her
A headache meeting the clinical criteria for
adolescent years, however, and she had a strong family
migraine is the most commonly described phenotype,
history of migraine. Both of these factors suggest an
but headaches resembling tension-type headache,
underlying predisposition to headache. This, in com-
cluster headache, occipital neuralgia, and cervico-
bination with her age, sex, and multiple head trau-
genic headache have also been described after head
mas, may have contributed to the risk of developing
trauma. The headache may also be nonspecific, not
fully resembling any primary headache disorder. Diag-
nosis can be complicated in younger patients as the
What treatment options are available for first presentations of many primary headache disor-
ders typically occur at this age. In our experience,
this patient? mild head trauma may be a precipitating event for
There is very little high quality evidence on which the emergence of an underlying primary headache
to base treatment of PTH. Prophylactic therapies for disorder.
PTH have not been compared in randomized con- There are no data to suggest when to start pro-
trolled trials, but open-label studies suggest benefit phylactic treatment for PTH. Some providers feel that
from amitriptyline, topiramate, valproate, and propra- starting treatment early, after a few weeks of headache,
nolol. Case reports have also described improvement may return the patient to function more quickly. This
after botulinum toxin treatments. We approach PTH approach may delay return to play/work, however, as
as though it were the underlying headache disorder it the current guidelines for returning to full activity
most resembles, treating migrainous headaches with require that the patient is asymptomatic off medica-
migraine preventatives, cluster-like headaches with tions. Other providers wait for several months of per-
verapamil, etc. This approach extends to symptomatic sistent headache prior to starting preventive treatment.
treatment, and in our experience patients who have If the patient is still having bothersome headache sev-
migrainous PTH exacerbations may respond well to eral months after the injury, preventive therapy is
triptan therapy. indicated.

Chapter 8: Pitfalls in preventive drug therapy

As this case illustrates, patients with repeated head Silberstein SD, Holland S, Feitag F, et al. Evidence-based
trauma may experience more severe post-concussive guideline update: pharmacologic treatment for episodic
symptoms with each successive head trauma. Patients migraine prevention in adults: report of the Quality
should be educated about the risks of successive head Standards Subcommittee of the American Academy of
Neurology and the American Headache Society.
traumas and advised to wear helmets. After a second Neurology. 2012;78:1337–45.
concussion, a discussion about whether to return to
playing contact sports may be warranted. Refractory migraine
Peterlin BL, Calhoun AH, Siegel S, Mathew NT. Rational
Diagnosis combination therapy in refractory migraine. Headache.
Chronic post-traumatic headache after mild head
Silberstein SD, Dodick DW, Lindblad AS, et al.
injury. Randomized, placebo-controlled trial of propranolol
added to topiramate in chronic migraine. Neurology.
Post-traumatic headache is more likely to occur after Combination treatment of migraine
mild head injuries and with repeat head injuries. Treat- Domingues RB, Silva AL, Domingues SA, Aquino CC,
ment is aimed at the headache type it most resembles. Kuster GW. A double-blind randomized controlled trial
of low doses of propranolol, nortriptyline, and the
Further reading combination of propranolol and nortriptyline for the
preventive treatment of migraine. Arq Neuropsiquiatr.
When to start prevention in patients with frequent 2009;67(4):973–7.
headaches Holroyd KA, Cottrell CK, O’Donnell FJ, et al. Effect of
Buchanan TM, Ramadan NM. Prophylactic preventive (beta blocker) treatment, behavioural
pharmacotherapy for migraine headaches. Semin Neurol. migraine management, or their combination on
2006;26(2):188–98. outcomes of optimised acute treatment in frequent
migraine: randomised controlled trial. BMJ. 2010;341:
Weight gain and migraine
Blumenfeld AM, Bloudek LM, Becker WJ, et al. Patterns of Paresthesias on topiramate
use and reasons for discontinuation of prophylactic Lee ST, Chu K, Park JE, et al. Paresthesia as a favorable
medications for episodic migraine and chronic predictor of migraine prophylaxis using topiramate. Eur
migraine: results from the second international burden J Neurol. 2007;14(6):654–8.
of migraine study (IBMS-II). Headache. 2013;53(4):
Luykx J, Mason M, Ferrari MD, Carpay J. Are migraineurs
at increased risk of adverse drug responses? A
Vo M, Ainalem A, Qiu C, et al. Body mass index and adult meta-analytic comparison of adverse drug reactions in
weight gain among reproductive age women with epilepsy and migraine. Clin Pharmacol Ther. 2009;85(3):
migraine. Headache. 2011;51(4):559–69. 283–8.
Winter AC, Wang L, Buring JE, Sesso HD, Kurth T. Silberstein SD. Control of topiramate-induced paresthesias
Migraine, weight gain and the risk of becoming with supplemental potassium. Headache. 2002;42(1):85.
overweight and obese: a prospective cohort study.
Cephalalgia. 2012;32(13):963–71. Medication overuse headache
Young WB, Rozen T. Preventive treatment of migraine: Hagen K, Albretsen C, Vilming ST, et al. Management of
effect on weight. Cephalalgia. 2005;25(1):1–11. medication overuse headache: 1-year randomized
multicentre open-label trial. Cephalalgia. 2009;29(2):
Principles and guidelines for migraine prevention 221–32.
Loder E, Burch R, Rizzoli P. The 2012 AHS/AAN guidelines Tepper SJ. Medication-overuse headache. Continuum
for prevention of episodic migraine: a summary and (Minneap Minn). 2012;18(4):807–22.
comparison with other recent clinical practice
guidelines. Headache. 2012;52(6):930–45. Orthostatic hypotension
Rizzoli, P. Acute and preventive treatment of migraine. Mussi C, Ungar A, Salvioli G, et al. Orthostatic hypotension
Continuum (Minneap Minn). 2012;18(4):764–82. as cause of syncope in patients older than 65 years

Chapter 8: Pitfalls in preventive drug therapy

admitted to emergency departments for transient loss of Synergistic side effects and drug-induced headache
consciousness. J Gerontol A Biol Sci Med Sci. Brouwers L, Iskar M, Zeller G, van Noort V, Bork P.
2009;64(7):801–6. Network neighbors of drug targets contribute to
O’Mahony D, Foote C. Prospective evaluation of drug side-effect similarity. PLoS One. 2011;6(7):
unexplained syncope, dizziness, and falls among e22187.
community-dwelling elderly adults. J Gerontol A Biol Sci Chakor RT, Bharote HS. Topiramate-valproate-induced
Med Sci. 1998;53(6):M435–40. encephalopathy in migraine. Headache. 2012;52(8):
Angle-closure glaucoma from amitriptyline
Chen HC, Tsai SJ. Trazodone-induced severe headache.
Boentert M, Aretz H, Ludemann P. Acute myopia and Psychiatry Clin Neurosci. 2011;65(7):681–2.
angle-closure glaucoma induced by topiramate.
Neurology. 2003;61:1306. Garcia-Serna R, Mestres J. Anticipating drug side effects by
comparative pharmacology. J Expert Opin Drug Metab
Bouassida W. Drug-induced acute angle closure glaucoma. Toxicol. 2010;6(10):1253–63.
Curr Opin Ophthalmol. 2007;18:129–33
OnabotulinumtoxinA for chronic migraine
Long-term side effects of topiramate
Blumenfeld A, Evans RW. OnabotulinumtoxinA for
Ali II, Herial NA, Orris M, Horrigan T, Tietjen GE. chronic migraine. Headache. 2012;52:142–8.
Migraine prophylaxis with topiramate and bone health
in women. Headache. 2011;51:613–16. Blumenfeld A, Silberstein SD, Dodick DW, et al. Method of
injection of onabotulinumtoxinA for chronic migraine: a
Láinez MJ, Freitag FG, Pfeil J, et al. Time course of adverse safe, well-tolerated, and effective treatment paradigm
events most commonly associated with topiramate for based on the PREEMPT clinical program. Headache.
migraine prevention. Eur J Neurol. 2007;14(8):900–6. 2010;50:1406–18.
Mikati MA, Ataya N, El-Hajj Fuleihan G. Re: Epilepsy- Durham PL, Cady R. Insights into the mechanism of
associated bone mineral density loss should be onabotulinumtoxinA in chronic migraine. Headache.
prevented. Neurology. 2009;72(10):943; author reply 2011;51:1573–7.
Evans RW. A rational approach to the management of
Pack AM, Morrell MJ. Adverse effects of antiepileptic drugs chronic migraine. Headache. 2013;53:168–76.
on bone structure: epidemiology, mechanisms and
therapeutic implications. CNS Drugs. 2001;15(8): Post-traumatic headaches
Erickson JC, Neely ET, Theeler BJ. Posttraumatic headache.
Topiramate and oral contraceptive efficacy Continuum (Minneap Minn). 2010;16(6 Traumatic Brain
Reddy DS. Clinical pharmacokinetic interactions between
antiepileptic drugs and hormonal contraceptives. Expert Lucas S. Headache management in concussion and mild
Rev Clin Pharmacol. 2010;3(2):183–92. traumatic brain injury. PM R. 2011;3(10 Suppl 2):
Drug-induced seizures Walker WC, Marwitz JH, Wilk AR, et al. Prediction of
headache severity (density and functional impact) after
Ruffmann C, Bogliun G, Beghi E. Epileptogenic drugs: a
traumatic brain injury: a longitudinal multicenter study.
systematic review. Expert Rev Neurother. 2006;6(4):
Cephalalgia. 2013;33(12):998–1008.

Pitfalls in nonpharmacologic treatment

9 of headache

Although medications are the mainstay of headache treatments for headache. Surgical interventions some-
treatment, nonpharmacologic treatments are often times considered for migraine treatment are also dis-
pursued as alternatives by both patients and practi- cussed here. We will start, however, by considering the
tioners. “Nonpharmacologic treatment” is a term often role of trigger avoidance in headache management –
used to refer to a heterogeneous group of therapies. perhaps the most basic nonpharmacologic interven-
These range from adjunctive integrative therapies such tion of all.
as vitamins, supplements, mind–body practices, and
behavioral therapies to potentially invasive procedures A young woman with dietary triggers
including surgery or nerve blocks.
Interest in nonpharmacologic treatments of all for headache
kinds is increasing in both popular culture and among
providers. As more evidence about these treatments Case
emerges they are becoming more mainstream and A 23-year-old woman presented for a second opinion
patients feel more comfortable asking for or about regarding her headaches. She had had headaches that
them. Many of these treatments have useful roles met criteria for migraine without aura about twice a
in headache management. They can be helpful for week since she was a young teenager. Initially these
patients who do not tolerate medications or are averse were well managed with nonprescription analgesics
to the idea of taking medications, and can also be such as acetaminophen or ibuprofen, but over the last
useful adjuncts for those already using pharmacologic several years these had stopped working. A previous
therapy. physician had started her on a triptan but also told her
While these therapies may seem benign, they that her headaches were likely due to environmental
are not without risk. In fact, the risks of the non- factors, specifically her diet. Since that visit about one
medication treatments are often overlooked because year ago, she had eliminated multiple foods, includ-
these treatments are perceived as safer and less likely ing lactose, gluten, soy, tree nuts, alcohol, and choco-
to cause adverse events than traditional medical ther- late. Although none of these changes had substantially
apy. In many cases, however, there is a lack of good reduced the frequency of her headaches, the patient
data about safety, and nonpharmacologic medications was afraid to reintroduce any of these foods lest her
that are classified as dietary supplements are exempted headaches worsen. She was very distressed in the office
from strong regulatory oversight by the US Food and and stated that she is losing weight because she did not
Drug Administration (FDA). These things should lead know what she could eat, and that her dietary restric-
to considerable caution on the part of doctors in rec- tions prevented her from going out with her friends.
ommending some of these interventions to patients. On exam she was very thin but the neurologic exam
Most of all, nonpharmacologic interventions should was normal.
not be assumed to be risk free, but instead should be
evaluated in a similar fashion to any other interven-
tion. In general, however, these treatments are well
What is the relationship between dietary
tolerated. factors and migraine?
In this chapter, we will discuss the indications and The possible effect of specific foods or even entire diets
pitfalls of some commonly used nonpharmacologic on frequency of migraine is a topic of much interest to

Chapter 9: Pitfalls in nonpharmacologic treatment

both headache practitioners and patients. Up to 46% wine with disguised vodka, 9 of 11 patients who drank
of migraine patients identified dietary triggers in one red wine developed a headache while none of the 8
study, and another found that 75% of patients with who drank vodka did. As limited as these studies may
chronic headache reported they were aware of possi- seem, there is even less evidence for any influence of
ble connections between food intake and headache. gluten or lactose. No studies examining the effect of
There are several possible mechanisms by which a gluten- or lactose-free diet on headache have been
a certain food could be associated with migraine: performed.
direct toxic or metabolic effect, vasodilation, and aller- In this case, the patient could be counseled about
gic response are all plausible biologic explanations, the lack of clear evidence for a relationship between
while conditioned taste aversion, expectation and self- diet and migraine. It may be helpful, however, to
fulfilling prophecy, or confounding factors (such as provide anecdotal support for the idea that certain
when stress triggers both consumption of a certain foods seem to reliably trigger headaches in some
food and headache, thus causing the appearance of the patients. She need not be hostage to her diet, however.
food triggering the headache) are behavioral explana- She could try gradually reintroducing foods one at a
tions. The role of anchoring cannot be overstated, as time, keeping a headache diary to assess the effect on
patients will almost always have eaten something in the headaches.
24 hours prior to having a migraine, thus making diet
an easy target when searching for a cause.
Fasting is a well-established trigger for migraines. Discussion
This is often seen in patients who fast for religious There is clearly tremendous interest in dietary precip-
reasons, including the observance of Yom Kippur or itants of headache and migraine, both among patients
Ramadan. Patients who habitually skip meals may find and among physicians. A “migraine diet” of some kind
that eating more regular meals may help to reduce is a common component of many popular lay pro-
headache frequency. grams for treatment of migraine. The rise of public
In this case, it seems unlikely that any individual awareness about food intolerance and allergies, along
dietary trigger is responsible for her headaches. This with widespread acceptance of elimination diets also
case illustrates the potential harm in placing undue contributes to interest in this approach to management
emphasis on dietary triggers in migraine. of medical conditions. A possible dietary effect on
migraine is biologically plausible, especially for sub-
stances that have direct or indirect effects on vascu-
What is the quality of evidence about lar tone or levels of biogenic amines that might be
dietary triggers for migraine and what is the involved in migraine. The evidence, however, is scant,
making an evidence-based recommendation about
best advice for this patient? dietary triggers in migraine challenging.
Given the subjective experience of patients and the It should also be noted that recommending the
emphasis given to dietary triggers in popular literature, elimination of multiple food groups from the diet
controlled studies are surprisingly inconclusive and is not entirely benign. A lack of specific nutrients
by and large do not support any universal statements may result, or, as in this patient, an overall lack of
about the relationship between diet and migraine. adequate caloric intake. Because the list of potential
Alcohol, chocolate, and cheese are the dietary trig- dietary migraine triggers is vast, large food groups
gers most commonly reported by migraineurs. In become suspect to the patient. Over time this can
one placebo-controlled dietary challenge study, sub- encourage secondary anxiety around dietary choices,
jects who identified chocolate as the dietary trig- which in and of itself may contribute to worsened
ger were fed bars containing chocolate or a closely headache. Placing undue emphasis on dietary trig-
matched placebo, and headaches in fact were more gers of migraine may also have the unintended effect
likely to occur in the group fed chocolate. No study of making patients feel that they are entirely respon-
has found that cheese or its hypothesized triggering sible for the frequency of their headaches. The bal-
agent, tyramine, reliably causes migraine. Red wine, ance between supporting a healthy lifestyle by avoid-
however, does have some evidence to support activ- ing known triggers and placing too much emphasis on
ity as a migraine trigger. In a study comparing red environmental factors is a fine one.

Chapter 9: Pitfalls in nonpharmacologic treatment

If a patient expresses interest in an elimination diet, Is this patient’s description of headache

it may be helpful to give a systematic framework within
which to conduct such trials. They may be encour- triggers likely to be accurate?
aged to keep a diary for a time, identify a list of pos- Most patients with migraine (between 75% and 95%)
sible triggering foods, and then track headache fre- report having headache triggers, or precipitants that
quency with the diary while eliminating each food bring on a headache. The research on triggers is, how-
for several weeks at a time. If no improvement in ever, surprisingly thin. Researching triggers is diffi-
headaches is seen, the food can be reintroduced. This cult as it is very difficult to isolate single triggers in
approach may help to identify triggers while prevent- real life. Studies attempting to verify specific triggers
ing cumulative loss of whole food groups from the such as changes in barometric pressure have been
diet. negative. Further complicating the question of trig-
gers is the phenomenon of “anchoring,” which occurs
when patients (or doctors!) connect an outcome with a
recent event. Those two conditions become connected
Diagnosis in the patient’s mind and evidence to support this con-
Episodic migraine without aura with possible dietary nection is unconsciously sought. Therefore, someone
triggers. may have a headache after eating tomatoes and from
then on pay more attention to headaches that occur
after eating tomatoes.
Tip Despite the lack of evidence, however, some events
Evidence is lacking to support a clear role for many are clearly linked to the development of headaches in
foods commonly assumed to be triggers of migraine. some patients. In this case, the patient noted a clear
Physicians should be cautious about recommending correlation between drinking coffee and developing a
highly restrictive dietary regimens for migraine, and headache within the next hour. When a trigger is reli-
avoid placing undue emphasis on dietary triggers. ably reproducible, as in this example, the trigger is
likely to be accurate.

A young woman with frequent Is there a role for trigger avoidance in

headaches attempting pregnancy this case?
Just as there is poor evidence for individual triggers for
Case headache, there are also very few data about the role
A 29-year-old woman presented for evaluation of of trigger avoidance in the management of headache.
migraine headaches since age 19, which had occurred Despite this lack of evidence, however, we have seen
about once a month until six months ago when they repeatedly in our clinic that patients often improve
increased in frequency. The migraines seemed to be when some effort at trigger management is under-
provoked by stress, let down from stress, lack of taken. Some triggers can be addressed with behavioral
sleep, skipping meals, and caffeine. She worked as changes, such as sleep hygiene (see Table 9.1), eat-
a medical assistant and had started school to train ing regular meals, and avoidance of situations likely to
as a physician assistant around the time that the involve strong sensory input (the perfume counter at a
headaches worsened. Her quality and duration of department store, for example). Some triggers, how-
sleep had declined. At the time of evaluation the ever, cannot be avoided, particularly including hor-
headaches were occurring about twice a week, and monal variation. For some patients, avoidance of some
she had started to miss work occasionally because of life stress is simply not feasible. Trigger management
them. alone is therefore unlikely to be an adequate preventa-
She had been treating her headaches only with tive strategy in cases of severe or frequent headache,
acetaminophen because she was planning to attempt but in less frequent or milder headache disorders it
pregnancy in the next several months. Her exam was may be sufficient. In a setting where preventive med-
normal and she had a negative MRI and blood work ication therapy is not desired, these and other behav-
ordered by her primary care physician. ioral interventions are at least a good place to start. In

Chapter 9: Pitfalls in nonpharmacologic treatment

Table 9.1. Sleep hygiene principles life stress. This increased stress could paradoxically
Maintain a regular sleep schedule, including a regular bedtime increase headaches, as the most commonly reported
and waking time headache trigger is typically “stress/tension.”
Do not sleep more than is necessary to feel rested. Do not stay
in bed long after waking up
Do not “force” sleep. If unable to sleep after 20 minutes, get out
of bed and do something calming Episodic migraine.
Avoid caffeine after 2 p.m. and alcohol in the evenings
Avoid smoking, especially at night Tip
Do not go to sleep on an empty stomach or just after eating Patient recall of triggers is not always reliable. Trigger
Keep a daily exercise schedule, but avoid exercise within 4 hours
management alone is often not sufficient treatment for
of bedtime frequent or severe headache disorders.
Let go of worries before bed. It may help to take a warm bath
before bed
A woman with frequent headaches and
Avoid electronics including TV before bed or in the bedroom
multiple allergies
this case, quality and duration of sleep was addressed Case
with sleep hygiene and her headaches decreased in fre-
A 39-year-old woman presented for evaluation of
quency to about once a week. These were relatively well
headaches occurring about three times a week. The
managed with acetaminophen.
headaches often met criteria for tension-type headache
but several times a month she would have more
Discussion severe migrainous headaches. The headaches had been
We distinguish between migraine triggers, which increasing in frequency over the last year, which she
increase the probability of a migraine attack in the associates with increased stress related to her mother-
short term (usually ⬍ 48 hours) and migraine aggra- in-law becoming ill and moving into their home for an
vating factors, which are associated with a longer-term increased level of care. She also had two young chil-
(usually weeks to months) increase in the severity or dren. She reported a large number of medication aller-
frequency of attacks. The most commonly reported gies and sensitivities including several previously tried
triggers, in addition to stress/tension, are menstru- headache preventatives. At the visit, she stated that she
ation in women, sensory input (bright lights, loud didn’t want to pursue further preventive trials because
sounds, strong odors), lack of sleep, and skipping “I do very badly with medications and they don’t work
meals. It also seems likely that triggers may be additive, for me anyway.”
in that one trigger alone may not be sufficient to cause
headache but multiple simultaneous triggers (such as What nonpharmacologic options are
exposure to a strong odor while sleep deprived) may
provoke a headache. available to this patient?
The idea that avoidance of triggers is an effective The term “nonpharmacologic” may refer to several
method of migraine prevention is an old one, and we types of therapeutic interventions, including behav-
see many patients who have been told that if they ioral strategies, complementary modalities, and herbs
could only identify and eliminate their headache trig- and supplements. These last are discussed later in
gers, their headaches would resolve. This idea seems this chapter. Of the treatments that do not involve
to follow good old-fashioned common sense: if some- ingesting a substance of some kind, behavioral treat-
thing gives you a headache, do not do it. There are cur- ments, which include biofeedback, cognitive behav-
rently no data to suggest that this is an effective strat- ioral therapy, and relaxation training have been the
egy for migraine prevention, however. Placing undue most studied. Complementary treatments (also called
emphasis on trigger avoidance may even be detrimen- complementary and alternative, alternative, or integra-
tal to the patient’s care. It has been suggested that try- tive therapies) include acupuncture, chiropractic care,
ing to manage migraines through trigger avoidance craniosacral therapy, and massage. Table 9.2 lists some
can lead to a restriction of activities and increased nonpharmacologic interventions.

Chapter 9: Pitfalls in nonpharmacologic treatment

Table 9.2. Nonpharmacolgic therapies no difference in improvement seen when comparing

Behavioral: groups who received true vs. sham acupuncture. Com-
– Relaxation training pared with pharmacologic treatments, acupuncture
– Biofeedback had slightly better outcomes and fewer side effects.
– Cognitive behavioral therapy
There are very little data available regarding the effi-
Complementary and alternative: cacy of craniosacral therapy.
– Craniosacral therapy
– Acupuncture
Systematic reviews show spinal manipulation pro-
– Chiropractic care duces some improvement of migraine headache com-
– Massage pared with drug treatment, but the level of evidence
was rated as modest. There have been only a hand-
ful of very small studies of the efficacy of massage for
The primary limitation for access for many patients headache, though these did show some benefit.
is cost, as these therapies are rarely covered by insur- While it is difficult to say which treatment will
ance. Despite the greater volume of data for behav- be the most successful in a given individual, thermal
ioral therapies, there are typically fewer practitioners biofeedback combined with relaxation training and/or
trained in modalities such as biofeedback or relaxation cognitive behavioral therapy may be the most helpful
training than there are practitioners of acupuncture, interventions.
chiropractic care, or massage. Access to these behav-
ioral treatments may also therefore be limited by access
to providers. The effort required to learn a behavioral Discussion
treatment may also be a barrier to use. Behavioral treatments for management of headache
For this patient, who is motivated and willing to have been in use for at least 40 years. The three types
pay for treatments not covered by her insurance, non- most often used and studied are cognitive behavioral
pharmacologic therapies could be recommended on therapy, relaxation training, and biofeedback. Cog-
the basis of efficacy. nitive behavioral therapy is conducted by a licensed
therapist and is short term, goal directed, and predi-
How successful are these interventions likely cated on the idea that changing dysfunctional thoughts
can change emotions and on reducing behaviors or
to be? environmental events that reinforce negative thoughts.
Behavioral treatments used for treatment of migraine As applied to headache, it is often used as a tool for
are the most studied nonpharmacologic treatments. A identifying and managing sources of stress. Empha-
comprehensive review of meta-analyses and evidence- sis is placed on empowering an individual to choose
based reviews included studies of all three types of their thoughts and behaviors (internal locus of con-
behavioral treatments. This review found that all three trol). A therapist with the goal of inducing global
forms of behavioral therapy are associated with a 30– relaxation or the “relaxation response” does relax-
60% reduction in headache activity. Likewise, the US ation training. This may be done as a guided exer-
Headache Consortium in 2000 gave a Grade A recom- cise or may be enhanced by biofeedback. Biofeedback
mendation to relaxation training, thermal biofeedback allows the patient to control physiologic parameters
combined with relaxation training, electromyography through awareness of those parameters. Parameters
(EMG) biofeedback, and cognitive behavioral therapy, used include muscle tone (measured by EMG), tem-
indicating that all of them “may be considered as treat- perature, and heart rate. This may also help patients to
ment options for prevention of migraine.” learn how to relax certain muscles, lower their heart
Acupuncture has also been shown to be effective rate, or induce relaxation.
for prevention of migraine. A 2009 Cochrane review The rationale for various complementary and alter-
found that compared with no preventative treatment native therapies varies. It is sometimes difficult to
or routine care, acupuncture reduced headache fre- generalize about them as there is a lot of hetero-
quency, overall headache days, and headache scores geneity within specific modalities. Acupuncture was
at three to four months. In the one trial with long- developed in ancient China and involves placing nee-
term follow-up, this effect seemed to dissipate by dles into the skin at certain prespecified points. It is
nine months after cessation of treatment. There was one of the most commonly used complementary and

Chapter 9: Pitfalls in nonpharmacologic treatment

Table 9.3. Who will benefit from behavioral treatments? behavioral therapy. With this, her headaches improved
Patients who to about once a week. Although she would have pre-
– Have poor or inadequate response to treatment with ferred to have fewer headaches, she felt this headache
medication frequency was manageable.
– Have contraindications to pharmacologic treatments for
medical reasons
– Do not tolerate pharmacologic treatments well
– Express a preference for nonpharmacologic treatments
– Are prone to overusing acute medications or have medication Mixed tension-type and migraine headaches.
overuse headache
– Have a high burden of stress or would benefit from improved
stress coping strategies
– Are pregnant or planning to become pregnant
In patients who cannot or do not want to use pharma-
cologic treatment, nonpharmacologic treatments may
alternative treatments, and is frequently used for treat- be helpful preventive strategies.
ment of headache. Of the roughly 4% of the US popu-
lation who reported ever being treated with acupunc-
ture, 10% of them had been treated for headache.
A woman interested in natural
Craniosacral therapy is a gentle hands-on modality treatments for migraine
that purports to make changes to pulsations of the
intracranial fluid by manipulation of the cranium and Case
sacrum. A 33-year-old woman presented to the office with a
Chiropractic care and massage are usually focused history of migraines with aura since age 16. They had
on addressing musculoskeletal abnormalities which occurred about twice a month until earlier in the year,
are hypothesized to cause headache. Although the when frequency gradually increased. She was not able
literature has not formally addressed this issue, we to identify any precipitating factors. At the time of the
have anecdotally seen many patients whose headaches visit she was having headaches at least once and some-
worsen after these treatments. Massage or manipu- times twice a week. The headaches were unchanged in
lation of the cervical area particularly may provoke character from her previous migraines. Her neurologic
headache. Other patients, particularly those whose examation was normal and no further workup was felt
headaches seem strongly related to cervical and shoul- necessary. Preventive medications were offered, but
der tension, may derive some benefit from massage. she said “I don’t want to go on a medication every day.
Evidence about the relationship between vertebral Isn’t there something natural I can take?”
artery dissection (VAD) and chiropractic cervical
manipulation is contradictory, but at this time we What vitamins or supplements might be
advise our patients to avoid high velocity rotational
manipulation of the neck. recommended?
Not all patients are equally likely to benefit from The three most commonly studied vitamins or supple-
a nonpharmacologic intervention. A review of behav- ments for the prevention of migraine are magnesium,
ioral therapies found that between 40% and 70% vitamin B2 (riboflavin), and coenzyme Q10 (CoQ10).
of patients did not respond to these therapies, and Several placebo-controlled trials have examined the
that medication overuse headache, chronic daily or efficacy of magnesium with conflicting but largely pos-
unremitting headache, and cluster or post-traumatic itive results. The doses tested were 360 to 600 mg/day.
headaches are all more likely to be refractory to these Magnesium is generally well tolerated, with gastro-
treatments. The US Headache Consortium defined a intestinal symptoms being the most common and diar-
group of patients who are likely to benefit from non- rhea being the most frequently described limiting fac-
pharmacologic therapies, as shown in Table 9.3. tor. This may depend to some extent on the magnesium
The patient in this case met several of the US salt used. Magnesium should not be used in patients
Headache Consortium criteria for patients who would with renal failure.
benefit from behavioral therapies. She was recom- Riboflavin has also been tested in controlled trials
mended to undergo training for biofeedback, which and generally has very few side effects other than turn-
she was able to do. She was also referred for cognitive ing urine bright yellow. Riboflavin, like magnesium,

Chapter 9: Pitfalls in nonpharmacologic treatment

Table 9.4. Vitamins, supplements, and herbs used in the prevention of migraine
Vitamins and
supplements Doses Side effects Notes
Magnesium 400–600 mg/day Gastrointestinal upset, diarrhea Oxide and chelates may be better tolerated. May
require several months for efficacy. Useful for
migraine with aura and in pregnancy
Vitamin B2 400 mg/day Turns urine fluorescent yellow Not studied in pregnancy. May require up to 3
(riboflavin) months for efficacy
Coenzyme Q10 150 mg/day Insomnia if taken at night Less evidence for efficacy; expensive
Feverfew 150 mg daily Mouth ulcerations, gastrointestinal upset Not safe in pregnancy
Butterbur 75 mg bid Burping Not safe in pregnancy. Quality control of preparation
(Petasites) essential given toxicity of whole plant extracts

likely needs to be taken for several months for bene- Discussion

fit to be seen. CoQ10 has little data to support its use
Vitamins, supplements, and herbal treatments com-
in the adult population of migraineurs, although there
monly used for migraine are summarized in Table 9.4.
is some evidence in the pediatric literature. It seems
These treatments may be useful for patients who would
well tolerated aside from causing possible insomnia if
like to avoid “medications” on a daily basis. Because
taken at night.
the FDA has limited regulatory oversight of herbal
Magnesium and riboflavin may be good options to
treatments and supplements, there can be a great
recommend to the patient described in the vignette.
deal of variation in the strength of the active ingre-
Magnesium may be particularly effective in patients
dients. The amount of active ingredient in feverfew
with aura, yet another reason to recommend it to this
preparations, for example, has been shown to vary by
400% among different formulations. Different magne-
sium salts may also have differing effects, as some are
Would herbal treatments be helpful? absorbed more completely than others. For this rea-
Feverfew and butterbur (genus Petasites) are two son, it can be difficult to know exactly what substances
herbs that have been studied for migraine prevention. a patient is taking when they purchase and use one of
In the recent American Headache Society/American these treatments.
Academy of Neurology (AHS/AAN) guidelines, but- Many of the trials of magnesium for prevention
terbur received a Category A rating, suggesting that of migraine have shown some benefit on migraine,
it “should be offered” to patients. Feverfew received with one negative trial and one equivocal trial. Cer-
a Category B rating (“should be considered”) in the tain magnesium salts are not well absorbed and may
AHS/AAN guidelines. The dose of butterbur is typ- not be effective, though no study has compared effi-
ically 75 mg given orally twice daily and feverfew is cacy between the different magnesium salts. In our
typically dosed at 150 mg orally daily. There are safety practice, we typically recommend 400–500 mg daily
concerns about the possible presence of alkaloids in and caution that it may take several months for effects
butterbur and neither of these herbs should be consid- to be seen. We find magnesium chelate and oxide to
ered safe during pregnancy. This may limit recommen- be generally well tolerated. The clearest indication for
dations for their use in this woman of childbearing age. magnesium is probably in patients who have migraine
Both of these treatments are generally well tol- with aura, or in patients in whom aura is the predom-
erated. One sometimes bothersome effect of butter- inant symptom. Magnesium is also considered a good
bur is eructation (burping). Feverfew can cause mouth choice during pregnancy. Patients who have low serum
soreness and ulcerations, stomach upset, and abdom- or red blood cell magnesium may also be particularly
inal pain. Because of their benign tolerability profile, responsive.
patients often seek these herbal treatments. Lastly, for Riboflavin (vitamin B2) and CoQ10 are mitochon-
treatment of nausea in this patient, ginger tea or ginger drial cofactors. CoQ10 has been somewhat better stud-
supplements might be recommended. ied in the pediatric literature. One trial compared

Chapter 9: Pitfalls in nonpharmacologic treatment

a combination of riboflavin, magnesium, and fever- The patient noted that his pain was left-sided and radi-
few (marketed as Migralief) with riboflavin alone and ated to the back and front of his head on the left. It was
with placebo. Both the riboflavin and the combination constant, not throbbing, and rated on average 4–5 on
group saw improvement in migraine days and other a 0–10 pain scale. He had no other associated symp-
endpoints. In our clinic, we often recommend mag- toms such as nausea, photo or phonophobia. Initially
nesium and riboflavin 400 mg daily given together. the pain was intermittent and cleared after the patient
Riboflavin has not been studied in pregnancy and left work and was able to change his posture and relax.
we do not recommend its use in pregnancy. Because For the last few months, however, it had increased in
CoQ10 is a costly supplement with limited evidence frequency and was often present after he left work. He
of benefit, we do not recommend it to patients in our was using over-the-counter ibuprofen with some relief.
clinic. Adverse events are, however, low and there have The patient’s physical and neurologic examinations
been no reported safety concerns, so we do not dis- were normal with the exception of decreased neck
courage its use if patients are interested. mobility, tenderness, and tightness and increased tone
Although herbal treatments are available without of the left neck and shoulder muscles. Neck X-rays
a prescription, there are safety concerns. In addition showed minimal degenerative changes of the spine.
to being associated with significant fetal risk, Peta- He was diagnosed with migraine and given rizatriptan
sites requires another safety caveat: the rhizomes and to use when the pain was severe. He returned several
stalks of the plant contain hepatotoxic and carcino- weeks later to say that the rizatriptan was not helpful
genic pyrrolizidine alkaloids. Concentrations of these and his pain was unchanged.
alkaloids are lowest in the leaves. Because of this, Pet-
asites should be obtained from a reputable manufac-
turer who can guarantee lack of toxicity. Feverfew has Was a diagnosis of migraine warranted?
not been studied in pregnancy and should be avoided This patient did have unilateral head pain of moder-
given its herbal medicine use to induce labor and pro- ate intensity, but he had no other migrainous symp-
mote uterine contractions. toms. Furthermore, his pain did not respond to a rea-
sonable dose of a specific anti-migraine drug. Based on
Diagnosis this patient’s history, his pain was most likely originat-
ing in structures in the neck. His problem appeared
Episodic migraine with aura.
shortly after the installation of a new computer sys-
tem at work that required him to sustain a nonphys-
Tip iologic head posture frequently and for long periods
Although available without a prescription, the long- of time. A cervical origin for his pain was suggested
term safety of herbal treatments and supraphysiologic by examination findings of reduced neck mobility and
doses of vitamins has not been well studied. They may increased muscle tenderness and tone in the area of
be helpful for patients who are averse to taking pre- the affected muscles. His X-ray findings showed only
scription medications, however. minimal arthritic changes that were unlikely to be the
cause of his pain.
A middle-aged man with pain in The existence of “cervicogenic headache” is contro-
versial, and there is disagreement about how a cervi-
the neck cal cause of headache should be diagnosed, with sev-
eral sets of diagnostic criteria in use. Most experts sug-
Case gest that a diagnosis of cervicogenic headache can be
A 38-year-old data entry technician was seen with made when pain is plausibly related to a neck source
complaints of refractory neck and head pain. He and experienced in the head or neck. This requires
reported that his pain began about three years ago demonstration by imaging or other methods such as
when his work station was equipped with a new mon- clinical examination of a disorder in the spine or adja-
itor and computer system. The placement of the mon- cent soft tissues that is known to be a cause of head
itor required him to turn his head to the left to see it, pain. The connection between this lesion and the pain
while the data he entered into the computer were on should be demonstrated by pain relief with diagnostic
cards and pieces of paper that were placed to his right. blockade of the nerve supply of the suspected causative

Chapter 9: Pitfalls in nonpharmacologic treatment

structure, as well as clinical signs that implicate the up visit the patient was no longer bothered by substan-
neck as a pain source. tial neck pain. Some physicians would refer patients
There are plausible biologic explanations for the like this who do not respond to conservative measures
radiation of pain from neck structures into the for local anesthetic blocks of the C2 nerve root, but in
head. The upper three cervical spinal nerves synapse our experience this invasive procedure is rarely neces-
with second-order neurons in the trigeminocervical sary. We do, however, sometimes use greater occipital
nucleus of the upper spinal cord, which also receives nerve blockade in these patients.
input from first-order trigeminal neurons. “Cross-
talk” between these converging neurons in the trig-
eminocervical complex may result in neck problems
producing head pain, and vice versa. Cervicogenic headache is a cause of unilateral head
and neck pain, usually without any side shift. The pain
is moderate to severe but otherwise lacks typical asso-
What treatment should be suggested for ciated features of other unilateral headache disorders.
this patient? Specifically, it is not associated with nausea, vomiting,
The authors of a recent systematic review of man- or photo or phonophobia, which can help distinguish
ual therapies for cervicogenic headache located only it from migraine. It is not associated with autonomic
seven studies. These studies tested a variety of manual features, which can help distinguish it from cluster
techniques for patients with head pain originating in headache. It is usually provoked by neck movement
the neck, including physical therapy, spinal manipula- or posture. A plausible cervical source of pain, lack of
tion, and jaw mobilization techniques. Unfortunately, response to triptan therapy, and the absence of typical
the quality of included studies was low, with only one migraine symptoms all point to a diagnosis of cervico-
including a “no treatment” control group. The authors genic headache.
concluded that although there was a suggestion that Although cervicogenic headache is a controversial
physical therapy and manipulative therapy might be diagnosis with no clear agreement on its features, it
helpful for cervicogenic headache, further research likely originates from anatomic structures in the cer-
was needed to substantiate any benefits. A recent ran- vical spine. Depending upon the criteria used to diag-
domized controlled trial of exercise and manipulative nose cervicogenic headache, its prevalence in the gen-
therapy for cervicogenic headache showed benefits for eral population ranges from 1% to 4.6%.
both of the individual treatments compared with a
control group. The combination of the two treatments, Diagnosis
however, was not clearly superior to either treatment Cervicogenic headache.
In the absence of consensus about the best method
of treatment, our practice is to begin with symp- Tip
tomatic treatment such as muscle relaxants and refer Nonpharmacologic approaches such as physical ther-
patients for physical therapy. We encourage the use apy, postural training, and attention to environmental
of time-limited active physical therapy interventions triggers are first-line treatment measures for cervico-
such as postural training and exercises to improve genic headache.
cervical muscle mobility and strength. Passive tech-
niques such as massage or ultrasound may offer tem- A man with frequent headaches
porary pain relief to some patients, but they can foster
the development of dependence on continued physi- interested in chiropractic treatment
cal therapy that is not in the patient’s best long-term
interests. Case
In this case, the occupational environment was A 52-year-old man came to the office for evaluation
likely contributing to the patient’s problem. The phys- of headaches which had been gradually worsening
ical therapist visited the patient’s workplace and over the last four years. They were bilateral, poste-
suggested rearrangement of his work station and alter- rior, aching or throbbing, moderate in severity, with-
ation of the flow and processes of his work. At a follow- out associated features, and lasting up to several hours

Chapter 9: Pitfalls in nonpharmacologic treatment

at a time. He thought they seemed worst when he series of patients treated with spinal manipulation,
woke up in the morning, and he had attributed this minor adverse effects were reported by 30–60% of
to sleeping in a bad position and thus straining his patients and included neck pain, stiffness, headache,
neck. He had tried several different pillows without any and fatigue. In retrospective and case–control studies,
improvement. Imaging showed mild arthritic changes more dangerous outcomes included disc herniation,
in the cervical spine. He was given a diagnosis of cer- vertebral fracture, dural tear, and, perhaps of most
vicogenic vs. tension-type headache and was offered concern, VAD. (Carotid artery dissection is reported
medication, but he wanted to try nonpharmacologic much less frequently than VAD, likely because the
treatments instead. He particularly wanted to know carotid artery is not tethered in the cervical region and
if chiropractic care would be an option as one of his therefore moves more freely.)
friends had good success with chiropractic treatments Several case–control studies have shown a relation-
for back pain. ship between spinal manipulation, whether by chiro-
practors or other practitioners (orthopedic surgeons,
Is spinal manipulation likely to help shiatsu practitioners), and VAD. One case–control
study showed that patients with VADs were more likely
his headaches? to report previously having chiropractic upper spinal
A recent systematic review of spinal manipulation for manipulation than those with carotid artery dissection
treatment of migraine included three randomized con- (6% vs. 30%). Dissection of both vertebral arteries was
trolled trials, all of which had significant method- also linked to spinal manipulation. In another case–
ologic limitations. The study with the best methodol- control study, the odds of developing a VAD were five
ogy showed no benefit for spinal manipulative ther- times higher within one week after a chiropractor visit
apy, while one study showed benefit but was of poorer in patients under the age of 45. These cases with VAD
quality. An attempt to develop evidence-based guide- were also five times more likely to have seen a chiro-
lines performed by a group of chiropractors and not practor three or more times in the last month for a cer-
limited to randomized controlled trials found mod- vical diagnosis.
erate or below level evidence. They concluded that Other studies have suggested a possible alternative
spinal manipulation could be recommended for treat- interpretation for these data, however. A large case–
ment of episodic or chronic migraine and for cervico- control and case-crossover study of the population
genic headache, and that no recommendations could of Ontario evaluated risk for VAD or vertebrobasilar
be made for treatment of tension-type headache. The stroke as a function of visits to the chiropractor or to
authors state that “adverse events were not addressed the primary care physician. In this study, cases under
in most clinical trials; and if they were, there were none the age of 45 were three times more likely to have vis-
or they were minor.” These two studies illustrate some ited their primary care physician and three times more
of the differences between recommendations made by likely to have visited a chiropractor in the month prior
physicians and by chiropractors. In brief, there are no to the stroke. There was no association for cases over
good quality studies suggesting benefit from spinal the age of 45.
manipulation, and studies showing benefit are limited The authors of this study and others note that
by methodologic weaknesses. In clinical practice, there VAD often presents with headache and neck pain.
are anecdotal reports of patients who have improve- Because of this, patients may present to a chiroprac-
ment in their headaches after spinal manipulation. tor or to their primary care physician for these symp-
This patient was told that anecdotally some patients toms before they are finally diagnosed. It may there-
report improvement in their headaches with chiro- fore be difficult to assess whether the increased risk
practic treatments but that the studies have not sup- noted by the case–control studies is simply a reflec-
ported it as a treatment. tion of usage of chiropractic care for a pre-existing con-
dition, rather than the chiropractic care causing that
What should he be told about risks condition. As one chiropractor summed up the results
of this study: “You are no more likely to have a dis-
associated with spinal manipulation? section walking out of my office than out of your pri-
Spinal manipulation has been associated with risks mary care physician’s office.