You are on page 1of 24


Neurological disorders in pregnancy

Dominic Heaney

Introduction neurological principles to situations where

published data are not available.
Neurological disorders are a significant cause This chapter outlines neurological assess-
of morbidity and mortality in pregnancy. In the ment and investigation in pregnancy before
UK between 2003 and 2005, women with neu- considering the most common and important
rological conditions accounted for 37 out of a neurological presentations multiple scle-
total 87 indirect maternal deaths (those not rosis, stroke, epilepsy and pain. These con-
arising from pregnancy or birth), with stroke ditions arise from different pathologies, but
(n = 24) and epilepsy (n =11)1 accounting their management allows the basic principles
for the majority of deaths from neurological of management of other neurological condi-
causes2. tions to be described. Overall women with pre-
Neurological assessment can distinguish existing neurological conditions, or women
between neurological symptoms that are com- who present with neurological symptoms dur-
mon in pregnancy (such as dizziness, pain and ing pregnancy rely on good communication
urinary frequency) and those that might indi- between their obstetrician, neurologist, physi-
cate a more significant disorder. Neurological cian, neonatologist and anesthetist to achieve
conditions can arise from organic disorders, optimal obstetric and neonatal outcomes.
which are usually associated with clear-cut
abnormalities with neurological investiga-
tions, but may also result from less well NEUROLOGICAL ASSESSMENT
understood functional problems, typically (CLINICAL EXAMINATION
causing pain, fatigue or even alteration in lev- AND INVESTIGATIONS)
els of consciousness. All neurological condi-
tions may present both medical and obstetric Neurological assessment in pregnancy is
challenges to clinicians unfamiliar with their undertaken in the usual way with clini-
management within the context of pregnancy. cal history and examination guiding use of
Fortunately, the natural history of key neu- investigations.
rological conditions has been observed during Even when a number of neurological symp-
pregnancy, providing a useful basis on which toms may be attributed to pregnancy, as when
to counsel women and organize services. considering neurological symptoms in the
Less information is available about the safety non-pregnant woman, much can be estab-
of neurological treatments or investigations lished by careful emphasis on the onset, pro-
during pregnancy or obstetric and neonatal gression and associated features.
outcomes for these women. Therefore, many Neurological examination should consider
decisions rely on applying key obstetric and higher mental function, cranial nerves and


neurological assessment of the trunk and Uterine shields are routinely employed and
limbs. In later pregnancy, small mechani- reduce fetal exposure further. In some cases
cal changes in gait and ability to participate the advantages that CT imaging of the head
in tests of hip power may be noted, but, in offers it can usually be obtained promptly
general, changes in the key markers of tone, and interpreted with ease particularly when
power, co-ordination, sensation or reflexes cerebral hemorrhage is suspected may out-
should not be attributed to pregnancy itself. weigh any putative risk of exposure to ionizing
Similarly, inspection should not reveal any radiation to mother and fetus.
changes in muscle bulk or symmetry. Nevertheless, in most cases magnetic reso-
Neurological investigation may cause more nance imaging (MRI) offers greater resolution
concern in pregnancy, although in most cases and extent than CT and therefore is likely to
anxieties are unfounded and tests may be be more clinically valuable. It does not involve
undertaken as in the non-pregnant woman. ionizing radiation and thus avoids any issues
For example, neurophysiologic assessment related to fetal radiation exposure in preg-
such as electroencephalography (EEG), nerve nancy. However, although there is no evidence
conduction studies and electromyography of fetal harm in humans after 20 years of wide-
(EMG) may be uncomfortable but are non- spread use of this technology many guide-
invasive and safe. lines recommend that, when possible, MRI be
Cerebrospinal fluid is obtained by the stan- delayed until the end of the first trimester and
dard method of lumbar puncture, although particularly when high-field (3 Telsa) magnets
access may be difficult in later pregnancy and are proposed. However, even in the first tri-
should be undertaken by suitably skilled prac- mester, the advantages may on occasion out-
titioners. Spinal fluid can show evidence of a weigh any theoretical risks.
wide range of central nervous system disor- It should be noted that in contrast to CT,
ders including infection, inflammation, demy- which may be acquired in a few minutes, MR
elination or, with cytological analysis, presence images are significantly more time-consuming
of neoplasia. Lumbar puncture is relatively to obtain. Depending on the technology being
invasive and standard relative and absolute used, MR scanning of more than two areas of
contraindications should be considered. In the neuroaxis (for example, head and cervi-
particular, consideration of any coagulation cal spine) may take in excess of 60 minutes.
disorders should be made, and lumbar punc- This may not be appropriate when an individ-
ture should not be undertaken in suspected uals condition is unstable, or the woman is
raised intracranial pressure, as brain stem her- confused.
niation may occur. If there is any doubt, clot- In later pregnancy, even with shorter dura-
ting studies and neuroimaging should precede tion MRI, women may have difficulty lying
this investigation. supine as compression of the inferior vena cava
Neuroimaging is often an integral part of by the gravid uterus impairs venous return,
patient assessment. The ionizing radiation leading to hypotension and ultimately fetal
associated with computerized tomography hypoxia. MR radiographers will be familiar
(CT) neuroimaging leads to obvious concern with methods by which this can be overcome,
on the parts of both clinicians and patients such as use of a lateral wedge or tilt. The size
when imaging the pregnant woman. Neverthe- of a heavily pregnant woman may occasionally
less, CT of the head involves a fetal radiation preclude the use of some CT or MRI scanners
dose exposure of less than 0.0005 rad and the and claustrophobia may be more of an issue
impact of fetal exposure is likely to be small3. than usual.

Neurological disorders in pregnancy

Iodinated contrast agents can be used in MULTIPLE SCLEROSIS

pregnancy and lactation but the intravenous
contrast agents used in MR (e.g. gadolinium) Definition and incidence
should not4.
Certain clinical situations may indicate the Multiple sclerosis (MS) is a disease of the cen-
use of other ionizing radiological techniques tral nervous system (brain and spinal cord)
such as digital subtraction angiography typi- that is characterized by both neuroinflam-
cally used to visualize the intracranial cerebral mation and neurodegeneration. Incidence is
vessels or spinal vessels in suspected vascular 3.6 cases per 100,000 person years (95% CI
malformations such as aneurysms or arterio- 3.04.2) in women and is higher in northern
venous malformations. The expected radiation latitudes, although this trend seems to be
dose depends on the procedure. For example, reducing5. The disease can have many clinical
this may be low with head imaging, but higher manifestations (with sensory loss in the limbs,
for spinal procedures. Clinicians should bal- visual loss, subacute motor loss, double vision
ance the risks of not fully investigating a and gait disturbance being most common) and
patient against potential harm to the fetus. a highly variable pace of progression (relaps-
ing remitting, secondary progressive, primary
progressive and progressive relapsing). The
NEUROLOGICAL CONDITIONS: median survival from onset of symptoms is
EPIDEMIOLOGY AND DEMOGRAPHY 38 years. MS is usually diagnosed between 20
and 50 years old, and so women with MS will
In many cases, women with neurological dis- therefore become pregnant relatively early in
orders are aware of their condition before con- the course of their illness and usually have
ception. But neurological illness may change correspondingly little associated disability.
or even present for the first time during preg- Nevertheless the issues of symptom manage-
nancy, presenting a very wide range of clinical ment and counseling are relevant at all stages
and obstetric challenges to those involved in throughout pregnancy.
their care.
Nevertheless, it is self-evident that the
majority of neurological conditions seen by Preconception
obstetric services reflect those disorders that
typically present among women of childbear- Preconceptional counseling offers the oppor-
ing age. Thus obstetric-neurology clinics or tunity to discuss the potential effect that preg-
other forms of neurology-obstetric liaison ser- nancy may have on MS, the effect that MS-
vices will be dominated by the four disparate related neurological impairment may have on
conditions of multiple sclerosis, stroke, epi- pregnancy and delivery, and the use of disease
lepsy and pain syndromes typically headache. modifying or symptomatic treatments in both
Consequently, this chapter focuses on these pregnancy and postpartum.
four disorders. Nevertheless, the descriptions Many people with relapsing and remitting
of the neurological impairments that arise MS are treated with disease modifying drugs
in these conditions and their management (DMDs). These include beta-interferons and
can be applied to other rarer conditions. It is a synthetic polypeptide, glatiramer acetate.
very unusual for women with neurodegenera- These treatments are administered by sub-
tive conditions such as Parkinsons disease or cutaneous or intramuscular injection at least
dementia to be pregnant, and these disorders twice weekly and have been shown to have
are mentioned only in passing. a modest but significant effect on reducing


relapse rate (approximately 30% per year) and transfer) of drug compared with oral treat-
to some extent long-term disability. ment, there are practical concerns regarding
The safety of these treatments in pregnancy kinking and blockage of treatment catheters.
has not been established. The US Food and Nevertheless, case reports have reported suc-
Drug Administration (FDA) have assigned cessful obstetric outcome8.
glatiramer acetate to pregnancy safety cate-
gory B (i.e. appears to be safe for pregnancy in
animal studies but not adequately studied in Genetics of multiple sclerosis
pregnant humans)6. It has not been shown to
be a teratogen in animal studies and its large The inheritance of MS is poorly understood
molecular weight (470011,000) suggests that
and is likely to involve a complex interaction
if its crosses the placenta, it does not do so by
between genetic and environmental factors.
simple diffusion. The interferons are US FDA
Nevertheless, in general, the child of a mother
pregnancy safety category C (i.e. human stud-
with MS has an approximately 23% chance
ies are lacking and animal studies are either
of developing MS, compared with 0.1% preva-
positive for fetal risk or lacking as well). In
high doses, interferons appear to be abortifa- lence in the general population9, although the
cients, but not teratogens. mother should be counseled that should MS
Current advice is that in most cases women occur, symptoms are unlikely to present until
should stop treatment with DMDs if they the third decade or later. The risk is increased
are planning to become pregnant or find further if both parents are affected.
themselves unexpectedly pregnant. Although
stopping treatment may expose the pregnant
woman to increased risk of relapse, in absolute Effect of pregnancy on relapse rate
terms the risk is relatively low (an extra 0.2
relapse/year)7. This risk is further modified by Women with MS who are considering preg-
the beneficial effect of pregnancy itself (see nancy will seek to establish whether preg-
below). It should be noted that some experts nancy will affect the number of relapses they
would offer women with MS the option of suffer and the overall course of their condi-
continuing agents for which there are limited tion. In the past, the relatively high number
reassuring pregnancy data, such as glatiramer of relapses observed postpartum led to the
acetate. false conclusion that pregnancy might lead
Women with MS may also be taking other to a long-term poorer outcome when com-
drugs, such as antimuscarinics for bladder dis- pared with non-pregnant women who have
orders (e.g. oxybutinin), antispasmodics (e.g.
not been pregnant. More reassuring evidence
baclofen or diazepam) and antidepressants
has been obtained from a large, prospective
(e.g. tricyclic antidepressants). Antimuscarin-
study (PRegnancy In MS/PRIMS), which has
ics, benzodiazepines or tricyclic antidepres-
continued to monitor enroled mothers many
sants have low teratogenic potential.
Rarely, patients with MS or other conditions years after enrolment into the study10. This
that cause spasticity have implantable devices study showed that, although risk of relapse
delivering an antispasmodic agent, typically postpartum was increased by a factor of two
baclofen, intrathecally. These pumps are for approximately 3 months postpartum, this
sited extraperitoneally within the abdominal was equally balanced by the observation of
wall. Although the local infusion of baclofen significantly fewer relapses during pregnancy.
intrathecally presents a significant theoreti- Thus, overall no difference is observed in the
cal advantage in terms of significantly reduced long-term number of relapses and disability of
serum levels (and therefore transplacental women who become pregnant.

Neurological disorders in pregnancy

Relapse management Non-pharmacological advice to improve

quality of sleep (sleep hygiene) should be
Although the risk of relapse is reduced dur- offered. These include examination of an indi-
ing pregnancy, this protective effect is less viduals sleep routine and the sleeping envi-
pronounced during the first and second tri- ronment. Women should be encouraged to
mesters. Should relapse occur, management avoid psychologically stimulating activities in
is the same as for non-pregnant women. Mild the evenings and intake of pharmacological
relapses require no treatment, but are likely stimulants such as caffeine should be mini-
to warrant assessment by occupational or mized. Drug treatments such as amantadine
physiotherapists as levels of disability may or modafanil, which are sometimes used to
be increased for several weeks. High-dose reduce MS-related fatigue, cannot be recom-
corticosteroids (for example a total of 1 g of mended, as there is no evidence to support
methyl-prednisolone administered intrave- their safety in pregnancy.
nously or orally for 35 days) may be used A sense of restlessness in the lower limbs
to speed up remission. Steroids present well is common in pregnant women, but women
known risks including reduced bone den- with MS may also have a degree of spasticity
sity, infection, mood alteration and adverse manifest as painful or irritating spasm. Neu-
gastrointestinal effects. Nevertheless, where rological examination can be useful to demon-
standard precautions and pretreatment assess- strate spasticity and localize relevant muscle
ment are applied, steroids may avoid the need groups. Physiotherapy advice and possibly use
for hospitalization or reduce considerably the of benzodiazepines, which are regarded to be
length of stay. relatively safe in pregnancy may be warranted.
Occasionally, relapses are severe and pro- Urinary symptoms should be carefully evalu-
gressive. In the non-pregnant woman these ated as impaired bladder emptying in women
may be managed with more aggressive treat- with MS predisposes to infection. Baclofen to
ment, such as mitoxantrone (a chemothera- treat bladder spasm is a reasonable option in
peutic agent) or natalizamub (a monoclonal pregnancy. Women with pre-existing urinary
antibody, which reduces white cell traffic into problems may be concerned that vaginal deliv-
the CNS). These treatments are either clearly ery will exacerbate urinary symptoms postpar-
teratogenic and embryotoxic or have no evi- tum, and this may be a matter of understand-
dence to support their safety in pregnancy. able concern to the individual. In such cases,
Their use should consider the advantages and individualized advice should be given.
risks to both mother and fetus on a case-by- Immobility in pregnancy increases the risk
case basis, taking into account the mothers of thromboembolism, so there should be a low
condition and the gestation. Intravenous threshold for the use of thromboprophylac-
immunoglobulin (IVIG)11 or possibly hemo- tic measures, such as graduated compression
dialysis may pose less risk to the fetus, but stockings or low molecular weight heparin in
there is less evidence to support their efficacy women with impaired mobility due to MS.
in reducing MS progression.

Third trimester and delivery

Management of other symptoms
In most cases, the third trimester and the
Women with MS may suffer from a range of peripartum period are not different between
symptoms during pregnancy including fatigue, women with MS and the non-MS population.
restless lower limbs and urinary symptoms. Specifically, obstetric and neonatal outcomes


do not differ with similar rates of induction, treatment should be counseled that should
instrumentation, cesarean section and infant they suffer relapse, most centers require good
mortality. recovery before DMD is restarted and so fur-
Nevertheless, women with MS may have ther delay may occur.
specific neurological impairments that affect A small number of centers have advocated
interpretation of symptoms during pregnancy. prophylactic use of IVIG postpartum to pre-
For example, women with plaques at T11 or vent relapses14. Activity of disease in the year
lower will have impaired bladder and bowel prepregnancy and in the first trimester to
function but normal sensation of uterine con- some extent predicts the risk of relapse in the
tractions and pain. Lesions between T6 and 3 months postpartum. Nevertheless, a recent
T10 will impair perception of uterine contrac- authoritative multivariate model predicts that
tions and where significant lesions are pres- using this information to make a decision
ent above T6, other signs of labor will have to about treatment would lead to 50% of women
be considered, such as worsening lower limb being treated unnecessarily13. IVIG is expen-
spasticity. sive and is derived from pooled human serum,
Urinary retention can occur during labor or which presents theoretical risks of infection.
postpartum, particularly when epidural anal- Thus, the current consensus is not to treat
gesia is used; consideration should be given to prophylactically.
the use of an in-dwelling urinary catheter from
early in the labor until after delivery (when the
woman is mobile and the effects of any epi- STROKE
dural have worn off) in women with impaired
bladder function. Difficult (and in particular Definition and incidence
instrumental) vaginal delivery may exacer-
bate urinary incontinence, and so should be Stroke is an acute neurological impairment
avoided if this is a pre-existing problem. that follows interruption of blood supply to a
In the past there have been theoretical con- specific part of the brain. Blood supply may be
cerns expressed regarding the safety of epidu- interrupted by thrombosis (either arterial or
ral anesthesia, and exacerbation of MS. Evi- venous) or embolism, or in a smaller propor-
dence from the PRIMS study, however, has tion of women by hemorrhage. As previously
been reassuring, demonstrating no significant described, it is the main cause of neurologi-
difference in outcomes between women with cal mortality in pregnancy, and represents a
and without epidural anesthesia12. significant proportion of all indirect maternal
Stroke is an uncommon but serious compli-
Postpartum cation of pregnancy. The incidence of stroke
in non-pregnant women aged 1544 years has
Women who are more mildly affected by MS are been reported to be as low as 10.7 per 100,000
more likely to choose to breastfeed. Although woman-years15. Multicenter or long-term
there are small studies suggesting otherwise, observational studies are therefore required to
the general consensus is that breastfeeding establish the incidence in pregnancy. Estimates
does not protect against (or cause) a postpar- using such methods produce widely differing
tum relapse13. Women who have been taking rates between 4.3 and 210 strokes per 100,000
DMDs before pregnancy balance the benefits deliveries depending on inclusion criteria with
of breastfeeding against the risk of relapse most studies suggesting an increased risk of
without treatment. Women who defer DMD stroke associated with pregnancy1618. Most

Neurological disorders in pregnancy

(up to 90%) strokes in these studies occurred Migraine with aura (see later) also produces
peripartum and up to a few weeks after the excess risk for stroke, but this condition is
birth. common and stroke in pregnancy is rare, so
caution should be used when counseling
women about this risk factor.
Risk factors for stroke Previous stroke during pregnancy presents
a particular dilemma for women considering
Physiological risk factors further pregnancy. Unfortunately, few data are
available, although in a follow-up study 13 of
Progressive physiological changes occur- 489 (2.7%) women aged 1540 who had suf-
ring throughout pregnancy predispose to fered a stroke had a recurrent event, but only
stroke including increasing hypercoagulabil- two of these occurred during pregnancy20. Full
ity, venous stasis and vascular wall changes. ascertainment of vascular risk factors, includ-
Pushing in the active phase of the second ing CT or MR angiography prior to pregnancy,
stage of labor involves episodes of significantly
is appropriate to best inform the individual of
increased intrathoracic pressure (Valsalva) and
her likely risk of pregnancy related recurrent
elevation of cerebral perfusion pressure, which
may lead to changes in cerebral blood flow
particularly where cerebral autoregulation or
anatomy is disordered19.
Clinical presentation and
management of stroke

Obstetric risk factors

Presentation and investigation
The main obstetric factor associated with an
Stroke presents as in the non-pregnant woman
increased risk of stroke is pre-eclampsia and
and clinical features may suggest either infarc-
eclampsia, in particular uncontrolled systolic
tion or hemorrhage, but neuroimaging is
hypertension. This is still the major cause of
required to confirm the diagnosis. The pos-
death due to pre-eclampsia in the UK1. Age
sibility of stroke should be considered in any
more than 35 years, black ethnicity, greater
woman who presents with any of the symp-
parity and multiple gestation are all risk fac-
tors for stroke, although quantifying this risk toms listed in Table 1. While in an imperfect
is not possible from available data. screen with a specificity of 88% and a sensi-
tivity varying from 66 to 100%, the Cincin-
nati Prehospital Stroke Scale may be a use-
Co-morbidity risk factors ful screening scale to help guide whether an
obstetric patient presenting with headache or
Women who become pregnant may have co- other softer neurologic complaints warrants
morbidity that increases the risk of vascular prompt complete neurologic assessment and
events including stroke; such factors include neuroimaging. This screening scale is summa-
obesity (BMI >30 kg/m2), diabetes, pre-exist- rized in Table 2.
ing hypertension, renal and heart disease, vas- If any of the three elements in the scale or
culopathies such as sickle cell disease, vascu- any other neurologic findings are newly abnor-
litis and collagen or atherosclerotic disease. mal, the possibility of acute stroke is high and
Alcohol, tobacco and cocaine use may cause a the patient should have urgent imaging and
vasculopathy or hypertension. evaluation by a neurologist.


Table 1 Symptoms that warrant consideration of stroke

Sudden weakness or numbness of face, arm or leg, especially if on one side of the body

Sudden confusion

Trouble speaking or understanding

Sudden trouble seeing in one or both eyes without a prior history of migraines

Sudden trouble walking

Sudden loss of balance or coordination not readily attributable to pregnancy

Sudden severe headache with no known cause

Table 2 Cincinnati Stroke Scale

Three elements

1. Facial droop
a. Have the patient smile and assess for facial droop
i. Normal: both sides of face move equally
ii. Abnormal: one side of face does not move

2. Arm drift
a. Have the patient hold both arms out and up with palms facing upwards
i. Normal: both arms move equally
ii. Abnormal: one arm drifts compared with the other

3. Speech
a. Have the patient repeat a sentence
i. Normal: patient uses correct words with no slurring
ii. Abnormal: slurred or inappropriate words or mute

As stated above, most pregnancy related cere- onset of symptoms. Therefore all patients
bral infarction occurs around the time of deliv- with symptoms suggestive of stroke require
ery and early puerperium16,21,22 at a time when prompt neuroimaging to determine whether
the mother is often bed bound, still hyperco- they have had an ischemic stroke that may
agulable and may just have had pelvic surgery, benefit from the use of thrombolytic therapy.
i.e. cesarean section or instrumental vaginal Table 3 lists the recommended guidelines for
delivery. Widespread adoption of postpartum timing of interventions for patients presenting
thromboprophylaxis use in the UK has been with acute ischemic stroke. Table 4 reviews the
associated with a decrease in the incidence of other assessments recommended for patients
cerebral infarction due to emboli, but not that presenting with possible acute stroke. Patients
due to uncontrolled systolic hypertension1. should be positioned with the head of the
Stroke is a medical emergency. Patients with bed lowered between 0 and 15 degrees and, if
acute arterial ischemic stroke from embolism blood pressure is greater than 180/105mmHg,
or thrombosis can have their long-term out- it should be treated acutely with intravenous
come greatly improved by the use of throm- labetalol. Aspirin should not be given while
bolytic therapy within 180 minutes of the investigating an acute stroke. Table 5 reviews

Neurological disorders in pregnancy

Table 3 Patients presenting with symptoms of acute stroke to an emergency room should

Be seen by a provider within 10 min with early notification of local stroke team of possible stroke patient

Have a neurologic assessment and head CT performed within 25 min of presentation

Have the head CT scan read with determination of whether they are candidates for fibrinolytic therapy
within 45 min of presentation

Receive fibrinolytic therapy within 60 min of presentation to the A&E and no longer than 180 min since
the time of onset of symptoms

Table 4 Additional investigations for patients presenting with possible acute stroke

Assess airway (can the patient protect her own airway or does she require intubation), breathing (what is
her respiratory rate and oxygenation) and circulation (are her pulse and blood pressure normal)

Obtain secure intravenous access

Obtain a fingerstick glucose measurement

Obtain a full blood count, urea and electrolytes, liver function tests serum glucose, serum troponin

Consider urine toxicology screen and blood alcohol level

Arterial blood gas if oxygen saturation abnormal

Obtain an ECG

the contraindications for thrombolytic therapy. It is not recommended to use thrombolyt-

The use of thrombolysis should be considered ics for acute ischemic stroke in the setting or
for pregnant and postpartum women with probable or confirmed pre-eclampsia.
severe acute cerebral non-hemorrhagic infarc- Specific syndromes are considered below,
but in most cases pregnant women who have
tion if it can administered within 180 minutes
cerebral infarction should be managed within
of onset of the neurologic deficit23. Thrombol-
a multidisciplinary stroke unit. Low dose aspi-
ysis is well tolerated by the fetus in pregnancy rin is the mainstay of treatment for acute isch-
and does not seem to increase the risk of pla- emic stroke. Aspirin and the other antiplatelet
cental abruption, so should not be withheld agents (aspirin with dipyridamole, or clopi-
if the maternal condition is life-threatening. dogrel) are also the most effective preventive
Postpartum, thrombolytic therapy carries a treatment of stroke. Unfractionated or low-
risk of uterine or pelvic hemorrhage. How- molecular weight heparin is not recommended
ever, this risk decreases with increasing time for acute stroke or stroke prevention except in
after delivery and, in most cases, the benefits the case of stroke from cardioembolism, arte-
rial dissection or large artery intraluminal
of thrombolysis will outweigh the risks. Local
thrombus. Warfarin is teratogenic and usually
hemorrhage can usually be dealt with by local
avoided in pregnancy.
hemostatic measures (such as uterine balloon) The differential diagnosis of acute stroke
or surgery. Close liaison between neurologist, in pregnancy is broad and includes migraine,
obstetrician and physician is essential in such transient ischemic attacks, head trauma,
cases. brain tumor, Todds palsy (a neurologic deficit


Table 5 Contraindications and cautions to thrombolytic therapy for acute stroke


Intracranial bleed on CT

Presentation suggests SAH

Multilobar infarction on CT

History of intracranial hemorrhage

Uncontrolled hypertension (>185/110 mmHg when treatment fibrinolytics to be given)

Known AVM/neoplasm

Witnessed seizure at onset of stroke

Active bleeding/acute bleeding diathesis (platelets <100, PTT elevated, INR >1.7)

Within 3 months of intracranial or intraspinal surgery/serious head trauma or previous stroke

Arterial puncture at a non-compressible site in the past 7 days


Consider whether benefits of thrombolytic therapy outweigh risks

Minor or clearing stroke

Within 14 days of major surgery or trauma

Within 21 days of GI/GU hemorrhage

Within 3 months of acute MI

Post MI pericarditis

Glucose <50 or >400 mg/dl

SAH, subarachnoid hemorrhage; AVM, arteriovenous malformation; PPT, partial prothrombin time; INR,
international normalized ratio; GI, gastrointestinal; GU, gastric ulcer; MI, myocardial infarction

following a seizure), systemic infection, func- zig-zag lines) or sensory (pins and needles)
tional deficits (conversion disorders) and in the perioral region. Less commonly, they
toxic metabolic disturbances (e.g. hypoglyce- are sensory in the upper limbs or difficulties
mia, acute renal failure, hepatic insufficiency, with speech (typically disarticulation with
drug intoxication). Perhaps the most chal- word finding difficulty or use of wrong words
lenging and common differential diagnosis for but no difficulty with comprehension). Neuro-
stroke in the obstetric population is migrain- logic symptoms other than this should not be
ous aura. Migrainous auras are typically brief casually attributed to migrainous aura. Visual
and more likely to be positive (the alteration or sensory symptoms should be one sided,
of a sensory perception) rather than negative gradually progress and last between 5 and
(the absence of a perception), e.g. wavy lines 60 minutes. If more than one aura symptom
in vision versus no vision or pins and needles is present, symptoms should occur in succes-
versus numbness. Migrainous auras are most sion rather than simultaneously. Importantly,
commonly visual (typically scotoma and/or migraine and migrainous aura is by definition

Neurological disorders in pregnancy

a recurring problem and the diagnosis cannot decrease intracranial pressure. Blood pressure
be made on first presentation of symptoms. is typically brought to a level of 160/90mmHg
If there is doubt about whether a patients (a mean arterial pressure of 110 mmHg) and
symptoms represent stroke/transient ischemic not much lower as some degree of hyperten-
attack or migrainous aura, an evaluation by a sion may be needed to maintain cerebral per-
neurologist and neuroimaging is advisable.
fusion and prevent ischemia. The presence of
an intracerebral bleed will complicate options
for obstetric anesthesia and an obstetric anes-
Specific stroke syndromes
and their management thetist should be involved early in these cases.

Pre-eclampsia and eclampsia

Reversible cerebral vasoconstriction syndrome
Presentation Pre-eclampsia is a multisys-
tem disorder affecting 35% of pregnancies24. Presentation Reversible cerebral vasocon-
Although only a tiny proportion of those striction syndrome (RCVS) is an underrecog-
affected by pre-eclampsia suffer from stroke, nized and often misdiagnosed syndrome char-
up to 45% of women who have pregnancy- acterized by a sudden-onset, severe headache
related stroke have pre-eclampsia or eclamp- seen in association with a neurologic deficit.
sia16,25. Uncontrolled systolic hypertension It is caused by reversible vascular narrowing
and endothelial dysfunction may lead to hem- involving the circle of Willis and its imme-
orrhage or infarction26. Disordered cerebral diate branches. RCVS can present in con-
autoregulation may also play a role, especially
junction with hypertensive encephalopathy,
pre-eclampsia and reversible posterior leuko-
encephalopathy, physical exertion or bathing
Investigation and management Investiga- and it can occur in isolation27. Women may
tion and management of pre-eclampsia will
have had an uncomplicated pregnancy and
be familiar to the obstetrician and is reviewed
present a few days after delivery with head-
elsewhere. Most cases of stroke in the setting
ache, cerebral irritation and neurological defi-
of pre-eclampsia are due to arterial hemor-
rhage (when thrombolytic therapy is contrain- cit. Investigations demonstrate infarction and/
dicated) but cases of acute arterial thrombo- or hemorrhage.
sis also occur. While pre-eclamptic stroke is The differential diagnosis includes subarach-
most likely in the setting of severe hyperten- noid hemorrhage, migraine, arterial dissec-
sion (>180/110mmHg), it can occur at blood tion, vasculitis or infection.
pressures much lower than this and the acute
change in blood pressure may be as important Investigation and treatment CT, MR or cath-
a factor as the absolute pressure. The only
eter angiography may demonstrate multifocal
definitive treatment for pre-eclampsia is deliv-
segmental narrowing of the cerebral vessels,
ery of the fetus and placenta. Prompt neuro-
which resolves within 46 weeks. Spinal fluid
imaging should occur in pre-eclamptic women
with sudden onset (thunderclap) headache should be normal, and this distinguishes this
and/or any persistent neurologic deficit. Neu- syndrome from subarachnoid hemorrhage
rosurgical consultation should be urgently (SAH).
sought if intracerebral blood is found on CT Treatment is supportive, although vasodila-
or MRI to guide the need for interventions to tors and steroids have been used.


Intracranial hemorrhage better than CT at identifying a SAH in the days

following the acute event.
Presentation Most intracranial hemorrhage Ruptured aneurysmal SAH may be compli-
occurring during an otherwise normal preg- cated by rebleeding with an associated mor-
nancy is the result of aneurysmal SAH and tality rate of 5070%29, so monitoring and
arteriovenous malformation (AVM). Intracra- management of such patients should take high
nial arterial dissection is a much rarer etiol- priority. Four per cent of patients will rebleed
ogy. Hypertension, smoking, alcohol and fam- within 24 hours of the initial bleed, and up
ily history are all risk factors. The incidence to 20% within the first month. Vasospasm,
of SAH from aneurysmal rupture is 311 per cerebral infarction, hydrocephalus, increased
100,000 pregnancies16, but 50% of all aneurys- intracranial pressure, seizures and hyponatre-
mal rupture in women below 40 occurs in the mia are other possible complications. Medical
context of pregnancy28. Cavernoma and other treatment usually involves intravenous fluids,
venous anomalies are a very infrequent cause
bed rest, compression stockings, analgesia,
of hemorrhagic stroke.
laxatives and nimodipine 60 mg 4-hourly30.
Presentation of intracranial hemorrhage
Medical management of SAH should be under-
is the same as in the non-pregnant woman.
taken at or in close liaison with a neurosurgi-
Symptoms are dominated by the sudden onset
cal center.
of headache, often described as the worst
Once the diagnosis is established, the eti-
headache of my life; this presentation should
ology for the SAH must be determined with
always prompt consideration of the diagnosis
cerebral angiography, CT angiography or MR
of SAH. Meningeal irritation (due to blood
spreading through the cerebrospinal fluid), angiography. While cerebral angiography
altered consciousness, collapse or vomiting at remains the most sensitive test, it is rapidly
onset, and the absence of lateralizing neuro- being replaced by CT angiography due to the
logic findings are features that are characteris- ease of testing and steadily improving technol-
tic of SAH but not universal. ogy. All of these tests can be safely performed
in pregnant or postpartum women when nec-
essary (see earlier discussion).
Investigation and management CT scan is very
Definitive treatment usually involves endo-
sensitive for SAH in the first 12 hours after the
vascular coiling or surgical clipping and the
event, but is less sensitive with smaller bleeds
timing of these interventions will be decided
and as the days go by after the initial event.
by the neurosurgeon. In most cases, treatment
Lumbar puncture is recommended in patients
with a history suggestive of SAH who have a of the mother is the primary concern, although
normal CT scan, especially if more than a day near to or during labor, in some cases the baby
has passed since the onset of their symptoms. may be delivered first28,3133. Outside preg-
The presence of xanthochromia on cerebrospi- nancy, coiling is felt to produce better over-
nal fluid is highly suggestive of a SAH but will all outcomes than clipping34. In pregnancy,
not be present until 26 hours after the acute the risks of periprocedure use of radiation,
event. postprocedure anticoagulation and postcoil-
CT offers advantages compared with MR in ing rupture in remaining aneurysm tissue are
ease of obtaining a study and in the past was generally outweighed by the benefit of effec-
viewed as better than MRI at identifying early tive treatment. At the time of SAH, women
hemorrhage. However, the use of FLAIR and with aneurysms may temporarily lack capac-
T2 sequences with MRI may be as good or bet- ity to consider these issues, but in any case,
ter than CT at identifying an early SAH, and is detailed discussion between the obstetrician,

Neurological disorders in pregnancy

the neurosurgical team and the family should Unruptured arteriovenous malformation
take place whenever possible.
Women who have had a previous aneurysm Presentation Arteriovenous malformations
completely obliterated by clipping or coiling (AVM) are less common than arterial aneu-
may consider vaginal delivery35. Use of epidu- rysms but present similarly with acute SAH.
Unruptured AVMs present a lower bleeding
ral anesthesia is advised. Some recommend
risk than aneurysms, and the overall risk of
avoidance of spinal anesthesia in women in
primary hemorrhage occurring during preg-
whom the aneurysm is not totally obliterated36,
nancy is 3.5%, which is similar to the normal
based on the hypothesis that the decrease population38. Individual case reports suggest
in intracranial pressure caused by dural tap that pregnancy is not associated with signifi-
could cause an increase in transmural pres- cant changes to AVM39, although the obstetri-
sure across the arterial wall, thus facilitating cian or neurologist should emphasize the pau-
rupture of a potential vascular malformation; city of data to guide decisions in this area.
however, anesthetic input is required as this
fall in pressure is likely to be preventable. Treatment AVMs are treated with combina-
tions of surgery, endovascular embolization
and stereotactic radiosurgery. The decision
Treatment of unruptured aneurysm about treatment is guided by a number of fac-
tors including the site and complexity of the
In general, management of unruptured aneu- lesion.
rysms should be the same as in the non-preg- In most cases, AVMs are managed outside
nant state, and guided by ISUIA (the Inter- pregnancy. Although there is concern that
national Study on Unruptured Intracranial untreated or partially treated AVMs may be
Aneurysms)37. Although rupture of aneurysm at risk of hemorrhage from the hemodynamic
changes of labor, the observed risk of hemor-
is associated with significant mortality, treat-
rhage is recognized to be low particularly
ment of aneurysms also carries risk. ISUIA
when epidural analgesia is used and pushing
data suggest that the risk of treating certain
in the second stage is limited, with early resort
low-risk aneurysms (small (<7 mm), asymp- to instrumental delivery40,41.
tomatic, stable, anterior artery aneurysms)
may be greater than the risk of conservative
watching and waiting. In common with many Cerebral venous thrombosis
trials, pregnancy has not been specifically con-
sidered. After discussion with the woman, it Presentation Cerebral venous thrombosis
may be felt appropriate to treat such aneu- (CVT) may account for approximately 20%
rysms prior to conception or during pregnancy. of strokes during pregnancy22 and should be
considered in any pregnant woman complain-
While there are no data to guide management
ing of headache and drowsiness particularly
of women with untreated aneurysms in labor
if focal neurological signs or seizures are evi-
and at delivery, most clinicians would recom-
dent. Its occurrence is now increasingly recog-
mend early good pain control, ensuring blood nized with the more widespread use of MRI;
pressure remains less than 140/90 mmHg and the incidence in pregnancy is estimated at 11.6
limiting the active phase of the second stage of per 100,000 deliveries in the US. Thrombosis
labor. An untreated aneurysm is not, however, of cerebral veins or dural sinuses causes injury
viewed as an indication for cesarean delivery. to tissue through increased venous pressures


and (in the case of dural sinus thrombosis) as acetylsalicylic acid (ASA). Patients with a
decreased CSF reabsorption and increased PFO (regardless of whether they have had a
intracranial pressure. The presentation is stroke) should therefore receive anticoagula-
highly variable and may include headache with tion with warfarin or heparin only if they have
or without vomiting, focal deficits, seizures another indication for anticoagulation. Deci-
and/or mental status changes. Headache is sions about the management of patients with
the most common presentation with gradual recurrent stroke and PFO, or those with a PFO
onset and often localized. with a single stroke but multiple risk factors
for recurrence (thrombophilia, atrial septal
Investigation and management The presence aneurysms) should be made in collaboration
of papilloedema is not a sensitive or specific with a cardiologist, hematologist and (when
sign of CVT and diagnosis relies on neuroim- pregnant or considering pregnancy) a high risk
aging, which will demonstrate venous distri- obstetrician and obstetric physician. Options
bution infarction with possible hemorrhage. to be discussed include full anticoagulation or
MRI in combination with MR venography is surgical closure of the PFO prior to pregnancy
the best test for diagnosing CVT. CT scans can but there is currently little evidence to guide
be normal in up to 30% of cases. management in these situations.
Even when hemorrhage is evident on imag-
ing, CVT is treated by anticoagulation for 612
months. During pregnancy, therapeutic doses EPILEPSY
of low molecular weight heparin may be used,
although evidence for its benefit is lacking. All women with epilepsy are likely to have
Data from non-pregnant patients suggest that considered the impact their condition and its
80% have complete recovery and that the rate treatment might have on their ability to have
of recurrence is well below 10%. and bring up children. For example, a postal
survey of 12,000 female members of Epilepsy
Action (a UK patient charity) obtained 2000
Paradoxical embolism
responses. The most important issues high-
lighted by women 1944 years were risk of
Patent foramen ovale (PFO) is an interatrial
epilepsy/medication affecting the unborn child
communication present in approximately 27%
(87%), effect of pregnancy on seizure control
of adults, but in up to 50% of young patients
(49%), and risk of child developing epilepsy
presenting with stroke42. This abnormal com-
munication may allow right-to-left shunting of
venous emboli directly into the arterial circu- Anecdotally, these considerations are also
lation, or provide a focus of thrombus forma- affected by broader social and economic factors.
tion. While casecontrol studies consistently For example, women may weigh the effects of
show a relationship between stroke and PFO, pregnancy and family on their employment in a
prospective data show that a PFO is not asso- way that differs from women who do not have
ciated with an increased risk of first or recur- epilepsy. As has been described previously,
rent stroke. While there is still controversy women with epilepsy are over-represented in
over this issue, most experts would currently areas of socioeconomic deprivation, and this
recommend that patients with a single prior may affect choice of partner. Furthermore,
stroke, a PFO and no other thrombotic risks women may be concerned about their partner
should receive only the usual stroke preven- and relationship feeling they may need to
tion treatments, i.e. antiplatelet agents such rely more on them than other women.

Neurological disorders in pregnancy

The history of advice given to women with has been associated with polycystic ovary
epilepsy about their ability to have children syndrome in a number of species, includ-
has reflected a great uncertainty and some ing humans45, although the strength of
prejudice against women with this condition. this association has been contested (see
Anecdotally, older women with epilepsy report below).
being told in no uncertain terms not to con- 4. Fertility may be reduced because of non-
ceive by their doctors. epilepsy related factors. Women may be
Over the past 40 years, women have not overweight, smoke or consume too much
been so actively dissuaded from having chil- alcohol all of which may reduce their
dren. Nevertheless, ideally, a woman should ability to conceive.
be informed of the full matrix of consequences
arising from the interactions of (1) the epilepsy
Thus, female fertility may also be affected by
syndrome; (2) seizures; (3) co-morbidities or
the epileptic syndrome, severity of seizures
other relevant health issues, such as smoking
and other co-morbidities independently of the
or alcohol consumption; and (4) anti-epileptic
effects AEDs4650.
drugs (AEDs), and these should be discussed
Epilepsy is more prevalent among women
with respect to the obstetric and fetal outcome
of low socioeconomic status (SES), and this
of pregnancy43.
may also have an impact on observed fertil-
ity rates, although this aspect of care has not
been directly investigated. Traditionally, fer-
tility rates among low SES women are higher
than among high SES, although pregnancy
Overall, fertility rates among women with epi-
outcomes are better in the latter group.
lepsy are slightly lower than the general popu-
lation. A number of epilepsy and non-epilepsy
related explanations for this observation have
Effect of pregnancy on epilepsy
been offered.
1. Women with epilepsy may find it difficult Anecdotally, most women express concern
to establish or maintain relationships with about how becoming pregnant might affect
men and therefore not be in a position to seizure control. Even women who are seizure-
consider planned pregnancy. A number of free while taking an AED are concerned about
studies demonstrate that such women are the medical implications (injuries, mortality,
less likely to marry or remain in relation- psychiatric effects) or the social consequences
ships than women without epilepsy44. (for example loss of driving licence) of a break-
through seizure.
2. Seizures may have an adverse effect on the
The risk that seizure control may deteriorate
menstrual cycle. In one study, over 35% of
has been considered in a number of studies,
women with partial seizures of temporal
recently summarized in a comprehensive lit-
origin had anovulatory cycles when stud-
erature review51. In this review, the percentage
ied over three cycles, compared to 8% of
of patients with unchanged seizure frequency
controls. The authors considered that
in these studies ranged from 54 to 80%. The
seizures might have a direct effect on the
highest rate of unchanged seizure frequency
hypothalamicpituitary axis, independent
was the 80% reported in AED-compliant
of drug effects.
patients, documented by serum levels. The
3. Certain AEDs have been highlighted rate of seizure decrease ranged from 3 to 24%.
to have particular effects on the female The rate of seizure increase ranged from 14 to
reproductive system. Valproate (VPA) 32%. Unfortunately, interpretation of these


studies is limited as none included a control Effect of seizures occurring

group of women with epilepsy who were not during pregnancy
pregnant: epilepsy is well known to ebb and
flow with seizure frequency spontaneously Women with epilepsy are familiar with the
changing over periods of months or years. effect that seizures may have on their own
There are several reasons why women with person. They are typically aware but less cer-
epilepsy may suffer increased risks of seizures tain about the effects seizures may have on the
during pregnancy. Specifically, AEDs which developing fetus.
have previously controlled epilepsy, may There has been a longstanding awareness
become less effective for a number of reasons: that tonicclonic seizures may cause abnor-
1. Women may take their treatment less malities in fetal heart rate56,57. Higher mis-
regularly or stop altogether because of carriage rates have been observed, but have
first trimester nausea, or fear of the poten- also been associated with there being family
tial risks from AEDs to the fetus52. As is history of epilepsy, particularly either parent
always the case with non-adherence, this having epilepsy58. The published data include
may not be reported at the time to the case reports where pregnant women have suf-
physician. fered seizures while the fetal heart rate has
been monitored, and fetal distress was docu-
2. Particularly in the later stages of preg-
mented54. Nevertheless, population studies to
nancy, women may be more sleep deprived,
inform clinicians and patients about the rela-
which can trigger seizures in susceptible
tionship between seizures and adverse fetal
individuals even when AEDs are taken.
outcomes are less clear.
3. Drug metabolism and drug effects are dif- It is the generally held view that the occur-
ferent in pregnancy. The state of pregnancy rence of generalized tonicclonic seizures dur-
induces significant changes in protein ing pregnancy may harm the fetus; although
binding of hormones and exogenous com- the absolute risk is low, it is likely to depend on
pounds such as AEDs leading to a fall the frequency and severity of seizures. There
in free drug concentrations particularly is insufficient evidence to quantify this risk.
in the final trimester. Additional pharma- Partial seizures (simple or complex), absence
cokinetic changes in drug clearance may seizures, or myoclonus are not harmful to the
result in reduction in available drug. Cer- fetus43.
tain AEDs, such as lamotrigine (LTG) are Nevertheless, a more recently published pro-
particularly prone to this effect53. spective study of cognitive function in children
exposed to AEDs in utero reported that the
Delivery is a time of particular concern. Sei- type of seizure (focal or generalized) or occur-
zures occurring around the time of delivery rence of more than five convulsive seizures
can result in both maternal and fetal harm54 during pregnancy was not significant within
compounding the hemodynamic and physical a regression model59. This finding contrasts
stresses associated with labor. It is commonly with previous data60 and its significance is fur-
quoted that for 24% of women with epilepsy, ther limited because in common with all previ-
delivery will be associated with seizures during ous work the study was not statistically pow-
labor or in the following 24 hours, although ered to consider this association, indeed the
the evidence for this statement is approxi- proportion of patients with poorly controlled
mately 20 years old55 and may reflect previous epilepsy was small: only eight pregnancies of
obstetric practice with respect to women with 303 mothers were exposed to more than eight
epilepsy. convulsive seizures.

Neurological disorders in pregnancy

The effects that might be induced by pro- a first-choice treatment for many idiopathic
longed seizures are also not well documented. generalized epilepsies, which may relapse if
For example, early reports suggested prolonged treatment is changed or stopped. The failure
seizures in the form of status epilepticus may of the reported studies to account adequately
result in significant fetal (and maternal) mor- for the effect of seizure frequency and severity
tality rates. Status epilepticus is defined as a on pregnancy outcome makes counseling indi-
seizure persisting for more than 30 minutes. vidual patients particularly difficult.
An early report by Teramo et al.54 documented In general, however, the main aim for women
29 cases from the literature of which nine of with epilepsy in pregnancy is to remain sei-
the mothers and 14 fetuses died. zure free, and they are therefore advised to
More recently, the EURAP study61 reported continue AEDs during pregnancy to avoid sei-
that seizure frequency presented a low risk zures. Exceptions occur if the risk of seizures
of adverse pregnancy outcomes such as spon- is very low, the seizures can be avoided in
taneous abortions, stillbirth and perinatal some other way or the seizures are mild (i.e.
deaths. Of 36 cases of status epilepticus (12 non-convulsive seizures)43. This decision mak-
convulsive) there was one stillbirth but no ing is summarized below.
cases of miscarriage or maternal mortality.
Although they reflect the findings of a large
and respected pregnancy registry, these find- Significance of epilepsy type in pregnancy
ings provide only limited guidance to neurolo-
gists and women with epilepsy: status epi- Some epilepsies are associated with underly-
lepticus is usually associated with significant ing conditions that have a strong genetic pre-
metabolic and hemodynamic compromise, disposition. In these cases, women will need
and risk of death is high. For example, a large to be given genetic counseling about the likeli-
North American epidemiological study dem- hood that their children may suffer epilepsy, or
onstrated overall mortality rates associated the condition that may underlie the womans
with status epilepticus were 22%62 similar to epileptic condition.
results obtained from other studies over many For example, women whose epilepsy is
years. In contrast the EURAP study recorded caused by Mendelian genetic conditions, such
a mortality rate of 0%, which suggests that as subependymal heterotopia, neurofibro-
the episodes of status epilepticus recorded in matosis or tuberous sclerosis will need to be
the EURAP study differed significantly from aware that these disorders may have up to a
those considered in previous studies. It must 50% chance of being expressed in any child.
be assumed that the cases of status epilepticus Women with learning disabilities for whom
in EURAP were somehow milder, or differed the underlying diagnosis is not clear and
in definition from other studies. whose neurological problems are likely to be
Overall, one of the key issues that doctors due to a combination of the effects of several
and women with epilepsy struggle with is the genes and environment are more difficult to
need to balance control of seizure frequency counsel. Nevertheless, maternal IQ and home
and severity with the potential adverse effects social environment are the strongest predic-
of AEDs during pregnancy. In particular, tors of a childs IQ63: maternal learning diffi-
women on high doses of AEDs, on non-first culties are likely to adversely affect child neu-
line AEDs and/or polytherapy are generally rodevelopment in both respects.
those with the most severe or enduring epi- Women with idiopathic generalized epi-
lepsy. VPA presents a particular challenge for lepsies, particularly absence epilepsies and
clinicians and women with epilepsy. VPA is juvenile myoclonic epilepsy are told that the


likelihood of their children developing a simi- whether they had been placed in foster care
lar disorder is in the order of 520%. It is not or remained with parents. Children placed in
normal practice to counsel women about the foster care were likely to have been exposed
subtle cognitive abnormalities that have been to higher doses of cocaine. But postnatal neu-
noted in some neuropsychometric studies of rodevelopment among the fostered group was
women with these conditions as too little is superior (although not equivalent to nor-
known about the magnitude of these effects. mal) highlighting the importance of the early
Certain behaviors such as alcohol consump-
Effect of co-morbidities and other health- tion and smoking are well known to be det-
related issues on pregnancy outcome rimental to pregnancy outcome67. Folic acid
should be taken68, although whether it pro-
Women with epilepsy who are pregnant or tects the fetus from neural tube defects in epi-
who are planning pregnancy may present with lepsy is less clear.
non-epilepsy related factors that may affect The above non-epilepsy related factors can
pregnancy outcome. have a significant effect on the outcome of a
Many of the co-morbidities frequently seen pregnancy. It is important that those inter-
in people with epilepsy (such as mental health preting the research data and guidelines, and
disorders, learning difficulties and osteopenia) counseling women with epilepsy about preg-
and their treatments may be relevant in terms nancy are aware that non-epilepsy related fac-
of predicting or counseling about pregnancy tors may modify risks presented to women in
outcome. For example, some antidepressants this context. The extent to which this is the
or antipsychotics may have established terato- case in neurology (and obstetric) practice in
genic potential. the UK is not certain.
Women may also present with specific
obstetric risk factors (such as history of pre-
mature labor, spina bifida or occurrence of Effect of anti-epileptic drugs
other birth defects), which may modify advice on pregnancy outcome
given in relation to epilepsy and AEDs.
Low SES is a further risk factor that is Without epilepsy, and without AEDs, on aver-
known to be associated with poor pregnancy age women face an overall risk of approxi-
outcomes. As has been described, women with mately 2% of bearing a child with a major con-
epilepsy frequently come from low SES house- genital malformation (MCM). Furthermore,
holds (i.e. partners may be of low SES) or it is estimated that 3.6% of children at school
neighborhoods. Low SES is known to be associ- age will be identified as having a primary spe-
ated with adverse obstetric and fetal outcomes cial educational need associated with learning
including low birth weight, perinatal, neonatal difficulties. A higher proportion, in the region
and postnatal mortality, and also of non-chro- of 20%, will require some form of learning
mosomal congenital malformations6466. support at school.
Furthermore, low SES and poor early house- The possibility that AEDs may adversely
hold environment are also associated with affect the development of the fetus to increase
poor early neurodevelopment. This issue is the risk of congenital malformations has been
relevant when considering the teratogenic recognized for many years in a range of stud-
effects of drugs used during pregnancy. For ies, published since before the 1970s69. In par-
example, children exposed to cocaine in utero ticular, a number of pregnancy registries in
were assessed and consideration given to North America, Europe, UK and Australia have

Neurological disorders in pregnancy

been set up to monitor pregnancy outcomes But over the past 5 years, there has been
in women taking AEDs7073 and one company increasing emphasis on the possible differen-
register was set up to monitor LTG only74. tial effect of VPA.
These registries are described in the report of The pregnancy registries, together with
Morrow71. These registries share in common other published series, also imply a differen-
the fact that only first trimester exposures tial effect of AEDs on fetal outcomes. This has
are considered, and that the greatest focus is intuitive appeal, as although AEDs all have
on the presence of MCMs. Meta-analysis of in common a capacity to suppress or prevent
their findings suggests that the risk of MCM is seizures, they are often very different chemical
increased approximately three-fold75. In partic- entities. In all the registry studies, consistently
ular women taking two or more drugs (poly- higher rates of congenital malformations are
therapy) have more than a 10% chance of seen with VPA compared with CBZ or LTG70
having a child with a congenital malformation. 73,80
. The data provided do not indicate any
It has been this range of studies, combined particular pattern of malformations, with the
with the warnings about potential teratogenic- exception of neural tube defects being more
ity from the manufacturer, and latterly data strongly associated with VPA, an association
from the registry studies that has led clinicians which has long been suspected81,82.
to warn patients about potential teratogenic There are many difficulties with the inter-
effects and, where possible, to restrict the dose pretation of these data and their application
of AED treatment to a single agent, and, if pos- to clinical practice. The most commonly cited
sible, to review dose or consider withdrawal. problem is indication bias and the possibil-
This strategy depends significantly on the ity that outcomes measured by the epilepsy
patients condition. For example, many of the registries, or even prospective cohorts may be
idiopathic epilepsies and particularly juve- confounded by factors including co-morbidi-
nile myoclonic epilepsy are likely to relapse ties, parental (including paternal) genetic and
if the dose is reduced too much or treatment is mental health, and socioeconomic status.
stopped. Importantly, it is also these epilepsies Three specific areas of concern about indica-
that best respond to VPA. tion bias have been expressed. First, women
Over the past 15 years, new AEDs have treated with VPA differ in terms of the under-
become available to treat different types of epi- lying epileptic condition. Until recently, VPA
lepsy. With greater choice of AEDs, there has was first choice treatment for women with
been corresponding interest in the differen- idiopathic generalized epilepsy, as it was per-
tial effects individual AEDs may have on fetal ceived to be more effective, whereas partial
development, in terms of rates of both con- onset epilepsies (with or without secondary
genital malformations and neurodevelopment. generalization) were treated with CBZ (or lat-
It is hoped that women who may become preg- terly LTG). Idiopathic generalized epilepsies
nant can be commenced on, or switched to, have a strong genetic propensity, and concerns
AEDs with potentially less teratogenic effects. have been expressed that any increased rate
Data have been derived from animal stud- of MCM or neurodevelopmental delay may
ies76, from a range of case reports, retrospec- reflect the underlying condition rather than a
tive observational studies and prospectively specific teratogenic effect.
collected from women with epilepsy. Over Second, there is clear evidence that use of
many years, many AEDs have been implicated VPA has fallen significantly since concerns
in causing additional MCM and neurodevel- have been raised about its teratogenicity. Pub-
opmental delay for example phenytoin77,78, lications and guidelines now routinely high-
phenobarbital and carbamazepine (CBZ)79. light VPA as having poorer fetal outcomes,


and practice has been observed to change83. neurological conditions accounted for 37 out
These trends in prescribing have also been of a total 87 maternal deaths, with epilepsy (n
demonstrated by the pregnancy registries. = 11)1 accounting for a significant proportion
For example, the Non-Epileptic Attack Dis- of these deaths.
order (NEAD) registry reported that whereas These statistics underpin general advice
in 1999, 17% of subjects reported using VPA that the obstetric care of women with epilepsy
in the first trimester, by 2008 this proportion should include close liaison between obstetri-
had fallen to just 3%. Although no study has cian and neurologist.
investigated the types of patient now treated In common with many other epilepsy out-
with VPA, it must strongly be suspected that comes clinicians and patients struggle to
those who have continued to be treated with extrapolate the observations in published
VPA over the past 10 years must differ sig- research and of general guidelines to their own
nificantly from those who are not. It is likely specific cases. Many women with epilepsy seek
they have more difficult to treat epilepsy (with to avoid an over medicalized pregnancy, which
co-morbidities) and do not tolerate or achieve they may perceive to be spoilt by numerous
seizure control on alternative drugs. Poorly medical appointments and investigations.
educated women, with poor access to medi- A recent authoritative consensus docu-
cal services who might be expected to have ment67,84 reviewed all evidence in this area.
poorer outcome in any case might remain The conclusions were vague, and were unable
on VPA through omission to change the AED, to distinguish between women with mild epi-
although evidence to support this assertion is lepsy and those who suffered frequent gen-
only anecdotal. eralized seizures: for WWE [women with
Third, the dose-dependent effect observed epilepsy] who are taking antiepileptic drugs
for VPA (and some other AEDs) may also (AEDs), there is probably no substantially
in part be due to patients with more severe increased risk (>2 times expected) of cesar-
underlying conditions requiring a greater dose ean delivery or late pregnancy bleeding, and
of VPA. Plasma concentrations are not help- probably no moderately increased risk (>1.5
ful in this regard as serum levels of VPA vary times expected) of premature contractions or
by more than 100% through a 24-hour period premature labor and delivery. There is possi-
and their measurement is of little clinical util- bly a substantially increased risk of premature
ity save for the purpose of identifying possible contractions and premature labor and delivery
non-compliance with treatment regimens. during pregnancy for WWE who smoke...
Significant confounders, such as co-morbid- Thus in everyday practice, it is recommended
ity, may be relevant. that there is good liaison between obstetri-
cian and epilepsy specialist, that care should
be taken to optimize AED doses at different
Obstetric outcomes stages through the pregnancy, and that deliv-
ery should take place in an obstetric unit with
Poorly controlled epilepsy is generally held to facilities for maternal and neonatal resuscita-
be associated with poor fetal and obstetric out- tion and treating maternal seizures43. The opti-
comes. For example, an authoritative review mization of AED dose is case specific. Individ-
stated that women with epilepsy have been uals vary in terms of their seizure threshold
observed to have a greater incidence of com- and metabolism of drugs. The consequence of
plications such as eclampsia, preterm delivery, a seizure for that individual must also be con-
spontaneous abortion and induced labor68. In sidered. For example, a woman with a history
the UK between 2003 and 2005, women with of status epilepticus, or who has been seizure

Neurological disorders in pregnancy

free for more than 12 months and holds a driv- pregnancy-related relapse in multiple sclerosis.
ing licence, is usually extremely reluctant to Pregnancy in Multiple Sclerosis Group. N Engl J
reduce drug doses to a level that may compro- Med 1998;339:28591
mise seizure threshold. 13. Vukusic S, Hutchinson M, Hours M, et al.
Pregnancy and multiple sclerosis (the PRIMS
study): clinical predictors of post-partum
relapse. Brain 2004;127:135360
14. Achiron A, Kishner I, Dolev M, et al. Effect of
intravenous immunoglobulin treatment on
1. Lewis G. Saving Mothers Lives: reviewing maternal
pregnancy and postpartum-related relapses in
deaths to make motherhood safer - 2003. London:
multiple sclerosis. J Neurol 2004;251:11337
CEMACH, 2007
15. Petitti DB, Sidney S, Quesenberry CP Jr,
2. RCOG. Mothers Die. Sixth Report of Confidential
Bernstein A. Incidence of stroke and myocar-
Enquiries into Maternal Deaths in the UK.
dial infarction in women of reproductive age.
London: RCOG Press, 2004
Stroke 1997;28:2803
3. American College of Radiologists. ARC practice
16. Kittner SJ, Stern BJ, Feeser BR, et al.
guideline for imaging pregnant or potentially preg-
Pregnancy and the risk of stroke. N Engl J Med
nant adolescents and women with ionizing radia-
tion. American College of Radiologists Council
17. Lanska DJ, Kryscio RJ. Stroke and intracranial
venous thrombosis during pregnancy and puer-
4. Chen MM, Coakley FV, Kaimal A, Laros RK Jr.
perium. Neurology 1998;51:16228
Guidelines for computed tomography and mag-
netic resonance imaging use during pregnancy 18. Witlin AG, Mattar F, Sibai BM. Postpartum
and lactation. Obstet Gynecol 2008;112:33340 stroke: a twenty-year experience. Am J Obstet
5. Alonso A, Hernan MA. Temporal trends in the Gynecol 2000;183:838
incidence of multiple sclerosis: a systematic 19. Tiecks FP, Lam AM, Matta BF, et al. Effects of
review. Neurology 2008;71:12935 the valsalva maneuver on cerebral circulation
6. FDA. Glatiramer data. 2001, http://www. in healthy adults. A transcranial Doppler Study. Stroke 1995;26:138692
label/2001/20622s15lbl.pdf 20. Lamy C, Hamon JB, Coste J, Mas JL. Ischemic
7. Lee M, OBrien P. Pregnancy and mul- stroke in young women: risk of recurrence
tiple sclerosis. J Neurol Neurosurg Psychiatry during subsequent pregnancies. French Study
2008;79:130811 Group on Stroke in Pregnancy. Neurology
8. Dalton CM, Keenan E, Jarrett L, Buckley L, 2000;55:26974
Stevenson VL. The safety of baclofen in preg- 21. Jaigobin C, Silver FL. Stroke and pregnancy.
nancy: intrathecal therapy in multiple sclerosis. Stroke 2000;31:294851
Mult Scler 2008;14:57172 22. Jeng JS, Tang SC, Yip PK. Incidence and eti-
9. Carton H, Vlietinck R, Debruyne J, et al. Risks ologies of stroke during pregnancy and puer-
of multiple sclerosis in relatives of patients in perium as evidenced in Taiwanese women.
Flanders, Belgium. J Neurol Neurosurg Psychiatry Cerebrovasc Dis 2004;18:2905
1997;62:32933 23. Cronin CA, Weisman CJ, Llinas RH. Stroke
10. Confavreux C, Hutchinson M, Hours MM, treatment: beyond the three-hour window
Cortinovis-Tourniaire P, Moreau T. Rate of and in the pregnant patient. Ann N Y Acad Sci
pregnancy-related relapse in multiple sclerosis. 2008;1142:15978
Pregnancy in Multiple Sclerosis Group. N Engl J 24. WHO International Collaborative Study
Med 1998;339:28591 of Hypertensive Disorders of Pregnancy.
11. Confavreux C. Intravenous immunoglobu- Geographic variation in the incidence of hyper-
lins, pregnancy and multiple sclerosis. J Neurol tension in pregnancy. Am J Obstet Gynecol
2004;251:11389 1998;158:803
12. Confavreux C, Hutchinson M, Hours MM, 25. Sharshar T, Lamy C, Mas JL. Incidence and
Cortinovis-Tourniaire P, Moreau T. Rate of causes of strokes associated with pregnancy


and puerperium. A study in public hospitals hemorrhage from cerebral arteriovenous mal-
of Ile de France. Stroke in Pregnancy Study formations. Neurosurgery 1990;27:86771
Group. Stroke 1995;26:9306 39. Uchide K, Terada S, Akasofu K, Higashi S.
26. Roberts JM. Endothelial dysfunction in pre- Cerebral arteriovenous malformations in a
eclampsia. Semin Reprod Endocrinol 1998;16:515 pregnancy with twins: case report. Neurosurgery
27. Ducros A, Boukobza M, Porcher R, et al. The 1992;31:7802
clinical and radiological spectrum of revers- 40. Finnerty JJ, Chisholm CA, Chapple H, Login
ible cerebral vasoconstriction syndrome. IS, Pinkerton JV. Cerebral arteriovenous mal-
A prospective series of 67 patients. Brain formation in pregnancy: presentation and neu-
2007;130:3091101 rologic, obstetric, and ethical significance. Am J
28. Barrett JM, Van Hooydonk JE, Boehm FH. Obstet Gynecol 1999;181:296303
Pregnancy-related rupture of arterial aneu- 41. Viscomi CM, Wilson J, Bernstein I. Anesthetic
rysms. Obstet Gynecol Surv 1982;37:55766. management of a parturient with an incom-
29. Dias MS. Neurovascular emergencies in preg- pletely resected cerebral arteriovenous malfor-
nancy. Clin Obstet Gynecol 1994;37:33754 mation. Reg Anesth 1997;22:1927
30. Pickard JD, Murray GD, Illingworth R, et al. 42. Overell JR, Bone I, Lees KR. Interatrial septal
Effect of oral nimodipine on cerebral infarc- abnormalities and stroke: a meta-analysis of
tion and outcome after subarachnoid haemor- case-control studies. Neurology 2000;55:11729
rhage: British aneurysm nimodipine trial. BMJ 43. NICE. Epilepsy. 2011.
1989;298:63642 CG020
31. Dias MS, Sekhar LN. Intracranial hemorrhage 44. Jacoby A, Baker GA. Quality of Life Trajectories
from aneurysms and arteriovenous malforma- in Epilepsy. Epilepsy Behav 2008;12:55771
tions during pregnancy and the puerperium. 45. Bauer J, Isojrvi JI, Herzog AG. Reproductive
Neurosurgery 1990;27:85565 dysfunction in women with epilepsy: recom-
32. Stoodley MA, Macdonald RL, Weir BK. mendations for evaluation and management. J
Pregnancy and intracranial aneurysms. Neurol Neurosurg Psychiatry 2002;73:1215
Neurosurg Clin North Am 1998;9:54956 46. Bilo L, Meo R. Epilepsy and polycystic ovary
33. Weir BK, Drake CG. Rapid growth of resid- syndrome: where is the link? Neurol Sci
ual aneurysmal neck during pregnancy. Case 2006;27:22130
report. J Neurosurg 1991;75:7802 47. Duncan S. Polycystic ovarian syndrome in
34. Molyneux AJ, Kerr RS, Yu LM, et al. International women with epilepsy: a review. Epilepsia
subarachnoid aneurysm trial (ISAT) of neuro- 2001;42(Suppl 3):605
surgical clipping versus endovascular coiling in 48. Genton P, Bauer J, Duncan S, Taylor AE, et al. On
2143 patients with ruptured intracranial aneu- the association between valproate and polycys-
rysms: a randomised comparison of effects tic ovary syndrome. Epilepsia 2001;42:295304
on survival, dependency, seizures, rebleeding, 49. Harden CL. Polycystic ovaries and polycystic
subgroups, and aneurysm occlusion. Lancet ovary syndrome in epilepsy: evidence for neu-
2005;366:80917 rogonadal disease. Epilepsy Curr 2005;5:1426
35. Wilson SR, Hirsch NP, Appleby I. Management 50. Herzog AG. Polycystic ovarian syndrome in
of subarachnoid haemorrhage in a non-neuro- women with epilepsy: epileptic or iatrogenic?
surgical centre. Anaesthesia 2005;60:4708524. Ann Neurol 1996;39:55960
36. Eggert SM, Eggers KA. Subarachnoid haemor- 51. Harden CL, Meador KJ, Pennell PB, et al.
rhage following spinal anaesthesia in an obstet- Management issues for women with epi-
ric patient. Br J Anaesth 2001;86:44244 lepsy-Focus on pregnancy (an evidence-based
37. Wiebers DO. Unruptured intracranial aneu- review): II. Teratogenesis and perinatal out-
rysms: natural history, clinical outcome, and comes: Report of the Quality Standards
risks of surgical and endovascular treatment. Subcommittee and Therapeutics and
Lancet 2003;362:10310 Technology Subcommittee of the American
38. Horton JC, Chambers WA, Lyons SL, Adams Academy of Neurology and the American
RD, Kjellberg RN. Pregnancy and the risk of Epilepsy Society. Epilepsia 2009;50:123746

Neurological disorders in pregnancy

52. Schmidt D. [Pharmacotherapy of epilepsy-- parison between rural and urban areas. BMJ
current problems and controversies]. Fortschr 1993;307:145862
Neurol Psychiatr 1983;51:36386 66. Vrijheid M, Dolk H, Stone D, Abramsky L,
53. Pennell PB. Antiepileptic drugs during preg- Alberman E, Scott JE. Socioeconomic inequali-
nancy: what is known and which AEDs seem ties in risk of congenital anomaly. Arch Dis Child
to be safest? Epilepsia 2008;49(Suppl 9):4355 2000;82:34952
54. Teramo K, Hiilesmaa V, Bardy A, Saarikoski S. 67. Harden CL, Pennell PB, Koppel BS, et al.
Fetal heart rate during a maternal grand mal Management issues for women with epile-
epileptic seizure. J Perinat Med 1979;7:36 psy--focus on pregnancy (an evidence-based
55. Delgado-Escueta AV, Janz D. Consensus guide- review): III. Vitamin K, folic acid, blood levels,
lines: preconception counseling, management, and breast-feeding: Report of the Quality
and care of the pregnant woman with epilepsy. Standards Subcommittee and Therapeutics
Neurology 1992;42(Suppl 5):14960 and Technology Assessment Subcommittee
56. Nei M, Daly S, Liporace J. A maternal complex of the American Academy of Neurology
partial seizure in labor can affect fetal heart and the American Epilepsy Society. Epilepsia
rate. Neurology 1998;51:9046 2009;50:124755
57. Sahoo S, Klein P. Maternal complex partial sei- 68. OBrien MD, Gilmour-White SK. Management
zure associated with fetal distress. Arch Neurol of epilepsy in women. Postgrad Med J
2005;62:13045 2005;81:27885
58. Schupf N, Ottman R. Reproduction among 69. Janz D. The teratogenic risk of antiepileptic
individuals with idiopathic/cryptogenic epi- drugs. Epilepsia 1975;16:15969
lepsy: risk factors for spontaneous abortion. 70. Holmes LB, Wyszynski DF. North American
Epilepsia 1997;38:8249 antiepileptic drug pregnancy registry. Epilepsia
59. Meador KJ, Baker GA, Browning N, et al. 2004;45:1465
Cognitive function at 3 years of age after fetal 71. Morrow J, Russell A, Guthrie E, et al.
exposure to antiepileptic drugs. N Engl J Med Malformation risks of antiepileptic drugs
2009;360:1597605 in pregnancy: a prospective study from the
60. Adab N, Kini U, Vinten J, et al. The longer UK Epilepsy and Pregnancy Register. J Neurol
term outcome of children born to moth- Neurosurg Psychiatry 2006;77:1938
ers with epilepsy. J Neurol Neurosurg Psychiatry 72. Tomson T, Battino D, Bonizzoni E, et al. EURAP:
2004;75:157583 an international registry of antiepileptic drugs
61. EURAP Study Group. Seizure control and and pregnancy. Epilepsia 2004;45:14634
treatment in pregnancy: observations from the 73. Vajda FJ, Hitchcock A, Graham J, et al. Foetal
EURAP epilepsy pregnancy registry. Neurology malformations and seizure control: 52 months
2006;66:35460 data of the Australian Pregnancy Registry. Eur J
62. DeLorenzo RJ, Hauser WA, Towne AR, et al. Neurol 2006;13:64554
A prospective, population-based epidemio- 74. Cunnington MC. The International Lamotrigine
logic study of status epilepticus in Richmond, pregnancy registry update for the epilepsy
Virginia. Neurology 1996;46:102935 foundation. Epilepsia 2004;45:1468
63. Tong S, Baghurst P, Vimpani G, McMichael A. 75. Meador K, Reynolds MW, Crean S, Fahrbach K,
Socioeconomic position, maternal IQ, home Probst C. Pregnancy outcomes in women with
environment, and cognitive development. J epilepsy: a systematic review and meta-analysis
Pediatr 2007;151:2848 of published pregnancy registries and cohorts.
64. Petrou S, Kupek E. Socioeconomic differences Epilepsy Res 2008;81:113
in childhood hospital inpatient service utili- 76. Finnell RH, Dansky LV. Parental epilepsy, anti-
sation and costs: prospective cohort study. J convulsant drugs, and reproductive outcome:
Epidemiol Commun Health 2005;59:5917 epidemiologic and experimental findings span-
65. Reading R, Raybould S, Jarvis S. Deprivation, ning three decades; 1: Animal studies. Reprod
low birth weight, and childrens height: a com- Toxicol 1991;5:28199


77. Scolnik D, Nulman I, Rovet J, et al. 81. Lindhout D, Schmidt D. In-utero exposure
Neurodevelopment of children exposed in to valproate and neural tube defects. Lancet
utero to phenytoin and carbamazepine mono- 1986;1:13923
therapy. JAMA 1994;271:76770 82. Omtzigt JG, Los FJ, Hagenaars AM, Stewart
78. Wide K, Henning E, Tomson T, Winbladh B. PA, Sachs ES, Lindhout D. Prenatal diagnosis
Psychomotor development in preschool chil- of spina bifida aperta after first-trimester val-
dren exposed to antiepileptic drugs in utero. proate exposure. Prenat Diagn 1992;12:8937
Acta Paediatr 2002;91:40914 83. Ackers R, Besag FM, Wade A, Murray ML,
79. Ornoy A, Cohen E. Outcome of children Wong IC. Changing trends in antiepileptic drug
born to epileptic mothers treated with car- prescribing in girls of child-bearing potential.
Arch Dis Child 2009;94:4437
bamazepine during pregnancy. Arch Dis Child
84. Chen DK, So YT, Fisher RS. Use of Serum
Prolactin in Diagnosing Epileptic Seizures. Am
80. Veiby G, Daltveit AK, Engelsen BA, Gilhus NE.
Acad Neurol 2005;65:668
Pregnancy, delivery, and outcome for the child
in maternal epilepsy. Epilepsia 2009;50:21309