See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/6462976

Treatment of Iatrogenic Cushing’s

Syndrome: Questions of
Glucocorticoid Withdrawal

Article in Orvosi Hetilap · March 2007

DOI: 10.1556/OH.2007.27964 · Source: PubMed

CITATIONS READS

7 3,862

5 authors, including:

Peter Igaz Károly Rácz

Semmelweis University Semmelweis University
173 PUBLICATIONS 1,359 CITATIONS 188 PUBLICATIONS 1,838
CITATIONS
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Diagnosis and treatment of malignant endocrine tumors View project

IL-6 bioactive domain View project

All content following this page was uploaded by Peter Igaz on 08 August 2015.

The user has requested enhancement of the downloaded file.

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 63

n ACTUALITIES IN CLINICAL PRACTICE n

Treatment of Iatrogenic Cushing’s

Syndrome: Questions of Glucocorticoid
Withdrawal
PETER IGAZ1, KÁROLY RÁCZ1, MIKLÓS TÓTH1,
EDITH GLÁZ1, ZSOLT TULASSAY1, 2

1 nd
2 Department of Medicine, Faculty of Medicine, Semmelweis University
2
Research Group of Gastroenterology and Endocrinology of the Hungarian Academy
of Sciences and Semmelweis University, Budapest, Hungary

Iatrogenic Cushing’s syndrome is the most common form of hypercortisolism. Glucocorticoids

are widely used for the treatment of various diseases, often in high doses that may lead to the
development of severe hypercortisolism. Iatrogenic hypercortisolism is unique, as the applica-
tion of exogenous glucocorticoids leads to the simultaneous presence of symptoms specific for
hypercortisolism and the suppression of the endogenous hypothalamic-pituitary-adrenal axis.
The principal question of its therapy is related to the problem of glucocorticoid withdrawal.
There is considerable interindividual variability in the suppression and recovery of the hypothal-
amic-pituitary-adrenal axis, therefore, glucocorticoid withdrawal and substitution can only be
conducted in a stepwise manner with careful clinical follow-up and regular laboratory examina-
tions regarding endogenous hypothalamic-pituitary-adrenal axis activity. Three major complica-
tions which can be associated with glucocorticoid withdrawal are: i. reactivation of the underly-
ing disease, ii. secondary adrenal insufficiency, iii. steroid withdrawal syndrome. Here, the
authors summarize the most important aspects of this area based on their clinical experience and
the available literature data.

Keywords: iatrogenic, hypercortisolism, glucocorticoid, withdrawal, secondary adrenal insuffi-

ciency

Abbreviatons

ACTH = adrenocorticotropin; CRH = corticotropin releasing hormone; HPA = hypothalamus-pituitary-adre-

nal system; IHM = insulin hypoglycaemia test; MT = Metyrapone test; ST = Synacthen (tetracosactid) test

Iatrogenic (exogenous) Cushing’s syndrome is the most frequently observed form of hypercorti-
solism in the clinical practice [1, 2]. Although it often results in severe clinical picture and may
appear following the treatment of many diseases where the anti-inflammatory, immunossuppressive
and apoptosis-inducing effects of glucocorticoids are exploited (e.g. autoimmune, haematologic,
inflammatory diseases), its treatment is still controversial. Whereas in the past years numerous stud-
ies have been published on the treatment of other forms of hypercortisolism (spontaneous, endoge-
nous forms), the treatment of iatrogenic forms is only discussed in a minority of studies.

Corresponding address: Peter Igaz MD PhD, 2nd Department of Medicine, Faculty of Medicine, Semmelweis
University, Szentkirályi u. 46, H-1088 Budapest, Hungary. E-mail: igapet@bel2.sote.hu

DOI: 10.1556/OH–HMJ.2007.27964 n 63 n 2007 n Volume 1, Number 1 n 63–72

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 64

n ACTUALITIES IN CLINICAL PRACTICE n

The clinical picture of iatrogenic hypercortisolism is not different from that of the other forms.
In view of the clinical picture and a history of glucocorticoid intake, its diagnosis is easy. As high,
supraphysiological doses of glucocorticoids are frequently applied, the iatrogenic Cushing’s syn-
drome is often very severe. Hypertension is frequent, but hypokalemia occurs more rarely than in
spontaneous forms, since the synthetic glucocorticoids have poorer mineralocorticoid activity in
comparison with the natural cortisol [1]. Virilization and hirsutism are also less frequent in iatro-
genic forms, as the exogenous steroids diminish the production of adrenal androgens by inhibiting
endogenous ACTH (corticotropin) secretion. On the other hand, glaucoma and other ophthal-
mological complications, e.g. posterior subcapsular cataract [3], avascular necrosis of bone [4] and
benign intracranial hypertension [5] occur more often in iatrogenic forms. A rare complication,
spinal epidural lipomatosis that may result in paraparesis is characteristic for patients suffering from
iatrogenic hypercortisolism [6]. Bone density is important to follow, since one of the major side-
effects of glucocorticoid therapy is osteoporosis. As a primary prevention measure, it can be pro-
posed to provide patients requiring long-term glucocorticoid therapy with calcium and vitamin-D
supplementation even at the beginning of steroid therapy [7].
An extraordinary feature of iatrogenic Cushing’s syndrome is the inhibition of the endoge-
nous hypothalamic-pituitary-adrenal (HPA) axis in patients with overt hypercortisolism.
Therefore, iatrogenic Cushing’s syndrome can be regarded as a duality of exogenous Cushing’s
syndrome and endogenous secondary adrenal insufficiency. Logically, a side-effect or disease
induced by an exogenous agent should be treated with the omission of the agent, however, this
strategy cannot be used here due to the insufficiency of the endogenous HPA axis and the under-
lying disease requiring steroid therapy. The sudden stop of glucocorticoid therapy would lead to
the reactivation of the underlying disease and the appearance of secondary adrenal insufficiency.
The long-term application of steroids in intraarticular, inhalative and ointment preparations
that also reach the systemic circulation, may also inhibit the endogenous HPA axis and result in
iatrogenic Cushing’s syndrome [1, 8, 9, 10]. Iatrogenic Cushing’s syndrome due to the use of
dexamethasone-containing nasal drops for many years has also been described [11]. In some
cases, the development of iatrogenic Cushing’s syndrome can be associated with the parallel
intake of medicaments inhibiting the metabolism of steroids, thus leading to the prolongation
of its biological half-life (e.g. agents inhibiting cytochrome p450 as the antimycotic itraconazole
or ritonavir used in the treatment of anti-HIV therapy to ameliorate the pharmacokinetics of
protease inhibitors) [12, 13]. The development of long-lasting Cushing’s syndrome following a
single intramuscular injection of a synthetic glucocorticoid (40 mg triamcinolon) has also been
reported [14]. In certain cases, agents not regarded as glucocorticoids but having glucocorti-
coid effects may also lead to iatrogenic Cushing’s syndrome, e.g. the progestagene megestrole
acetate used for the treatment of AIDS-cachexia and various malignancies including breast,
uterus and prostate neoplasms [15]. In patients with hypalbuminaemia (e.g. chronic liver dis-
ease), the free, not protein-bound fraction of glucocorticoid is higher, therefore smaller steroid
doses could result in iatrogenic Cushing’s syndrome [5].
The most significant question of the treatment of iatrogenic Cushing’s syndrome is the
problem of glucocorticoid withdrawal.
The withdrawal from a short glucocorticoid therapy lasting for some weeks or some months at
most, does not lead to major problems in general, complications are rare, therefore, the questions
of withdrawal from long-term (several months or years) glucocorticoid treatment will be present-
ed here. Unfortunately, no generally accepted consensus is known in this field [16], so we will
present the most rational possibilities available in the literature together with our experiences.

2007 n Volume 1, Number 1 n 64 n HUNGARIAN MEDICAL JOURNAL

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 65

n ACTUALITIES IN CLINICAL PRACTICE n

Questions of Withdrawal from Glucocorticoid Therapy

Some authors argue that glucorticoid treatments shorter than three weeks do not really sup-
press the HPA axis, independently of the doses applied, therefore treatment can be stopped
abruptly without a gradual dose diminution [17]. Other data, however, show that suppression
of the HPA axis can even occur by high-dose glucocorticoid treatments as short as 5–30 days
[18], but with doses near the physiological levels, clinically significant suppression is rare with
treatment shorter than one month [5]. Glucocorticoid treatment inhibits all participants of the
HPA axis. The atrophy of adrenal cortex can already be detected even after some weeks of treat-
ments due to the lack of the trophic effect of ACTH [7]. Although ACTH somewhat influences
the functioning of aldosterone-secreting zona glomerulosa as well, the suppression of HPA axis
does not really influence the secretion of aldosterone.
The development of Cushing’s syndrome-related symptoms shows strong correlation with
the length of steroid treatment: although Cushing’s syndrome is not expected after a four-week-
long steroid therapy, the dose of glucocorticoid should be withdrawn gradually. The longer the
time of steroid therapy and the higher the dose applied, the longer the time needed for with-
drawal. In some cases, the regeneration of the HPA axis takes more than one year [5, 19]. On the
other hand, the reactivity of the HPA axis may be preserved with a low steroid dose, given in the
morning (e.g. 5 mg or less prednisolone) [20]. It is very important to give steroids in the morn-
ing, as the suppression of early morning ACTH-secretion occurs less frequently by this protocol,
thus suppression of the whole HPA axis is rare [5, 21]. The type of glucocorticoid also influences
the development of HPA axis suppression, as preparations of longer duration, e.g. dexametha-
sone, suppress HPA-activity to a greater extent. In view of the severe side-effects of steroid thera-
py, we should attempt to give steroids for the shortest period possible and the withdrawal should
be initiated as soon as possible. If chronic steroid therapy is necessary, alternate day application
can be recommended after some weeks of higher initial doses, since the suppresion of HPA axis
occurs less frequently by this protocol and the withdrawal is easier to perform.
The most serious problem of glucocorticoid withdrawal is related to the high interindividual
variability regarding the time needed for the suppression of HPA axis and the capacity of its
regeneration [1, 21]. The reactivity of the hypothalamus and the pituitary returns earlier than
that of the adrenal cortex [5]. The clinical picture and laboratory examinations, dynamic tests
can be applied to judge the process of regeneration.

Tests for Assessing the Functional Activity of the HPA Axis

It is a difficult question to answer how the functional activity of the HPA axis looks like, or
when the substitution glucocorticoid therapy can be omitted. Although the clinical picture of
adrenal insufficiency (e.g. dizziness, orthostatic hypotension, nausea, vomiting, hyponatraemia,
hypoglycaemia, subfebrility, fever, confusion, abdominal pain, shock etc.) certainly indicates the
insufficiency of treatment, objective parameters are mandatory. For this purpose, we can use the
morning cortisol levels obtained from blood taken between 6.00 and 8.00 and dynamic tests.
For the correct interpretation of both morning cortisol values and dynamic tests, glucocorticoid
substitution should be suspended for at least 24 hours.
The reliability of the laboratory test used for determining cortisol concentration is an impor-
tant question. Steroid substances can cross-react with endogenous cortisol – depending on the

HUNGARIAN MEDICAL JOURNAL n 65 n 2007 n Volume 1, Number 1

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 66

n ACTUALITIES IN CLINICAL PRACTICE n

laboratory protocol used – that could disturb the interpretation of results. The longer the half-life
of the synthetic steroid is, the longer its application should be suspended before the examination.
If the morning cortisol level is under 3 µg/dl (85 nmol/l), substitution cannot be omitted
[1]. Cortisol levels over 20 µg/dl (550 nmol/l) indicate the appropriate functioning of the
HPA axis, in this case substitution can be omitted [1]. Cortisol concentrations between these
two values are not reliable as markers of HPA axis activity, therefore dynamic tests should be
performed.
Several dynamic tests are available for assessing the functional activity of the HPA axis.
Among these, the insulin-hypoglycaemia (IH) and the Metyrapone-test (MT) are used to assess
the whole HPA axis, the tetracosactid (Synacthen)-test is used for the study of the adrenal cor-
tex, whereas the CRH (corticotropin releasing hormone)-test can be applied for examining the
pituitary-adrenal cortex subsystem [7, 21]. As the details of these tests are included in textbooks
[22, 23], only the most relevant aspects are presented here. The IH-test can be regarded as the
golden standard, since it examines the whole HPA axis, in addition it is also suitable for study-
ing other hypothalamic-pituitary systems, e.g. growth hormone secretion. Hypoglycaemia
(blood glucose < 2.2 mmol/l) induced with 0.1–0.15 U/kg body weight regular insulin acti-
vates the HPA axis. Normally, cortisol serum concentration should surpass 20 µg/dl (550
nmol/l) in the blood sample taken during hypoglycaemia. The major drawbacks of IHT are: it
is dangerous due to the development of hypoglycaemia, therefore not to be performed on eld-
erly and patients suffering from cardiac problems and epilepsy, it is time-consuming, difficult
and requires surveillance in a hospital.
MT is also suitable for studying the whole HPA axis [24]. Metyrapone inhibits the enzyme
11β-hydroxylase that catalyzes the final step in cortisol biosynthesis, thus the reduced cortisol
production induces the secretion of ACTH from the pituitary. As a consequence of this, the
serum concentration of ACTH and the cortisol precursor 11-deoxycortisol (Compound S) will
increase [24]. Following the intake of 30 mg/kg Metyrapone between 23.00 and 00.00, the
concentration of 11-deoxycortisol in next morning’s blood sample should be over 7 µg/dl (200
nmol/l), whereas the concentration of cortisol should be lower than 5 µg/dl. Rarely MT can
lead to adrenal insufficiency, however, it is safe even for outpatient applications. Its most serious
difficulty is related to the limited availability of 11-deoxycortisol measurement and even that of
Metyrapone. Its sensitivity is near to that of IHT [24].
It is the short tetracosactid (Synacthen)-test that is most frequently used for the assessment of
the HPA axis. In case of adrenal insufficiency, serum cortisol concentration remains below 20
µg/dl (550 nmol/l) 30 minutes after the intravenous injection of 250 µg tetracosactid
(Synacthen, ACTH-analogue). As the sensitivity toward ACTH is reduced even after some days
after the cessation of ACTH-secretion, the Synacthen-test (ST) is not only suitable for the diag-
nosis of primary but also for secondary adrenal insufficiency. Thus, ST can be applied for the
assessment of the entire HPA axis. Nevertheless, the sensitivity of ST is less than that of IHT. Its
sensitivity can be enhanced, if the cortisol concentration to be surpassed is increased from 20
µg/dl (550 nmol/l) to 24 µg/dl (650 nmol/l) [25], however, its specificity may be reduced by
this modification. A significant problem with the classical ST is that the standard 250 µg tetracos-
actid dose results in a serum concentration that is many times higher than the physiological level,
thus it may not be sensitive enough for the diagnosis of milder forms of secondary adrenal insuf-
ficiency. Therefore, the application of a lower dose, i.e. 1 µg was raised (low-dose ST). In healthy
subjects, 1 µg tetracosactid also results in cortisol concentrations over 20 µg/dl (550 nmol/l) 30
minutes after injection. The utility of low-dose ST in secondary adrenal insufficiency is, however,

2007 n Volume 1, Number 1 n 66 n HUNGARIAN MEDICAL JOURNAL

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 67

n ACTUALITIES IN CLINICAL PRACTICE n

controversial. Some authors argue that its sensitivity reaches that of classical tests [26], whereas
others report that its sensitivity and specificity are much lower [17, 27]. A practical problem with
low-dose ST is related to the fact that tetracosactid is not available in 1 µg vials.
CRH test is suitable for the study of pituitary-adrenal subsystem by stimulating ACTH and
thereby cortisol. Following the intravenous injection of 1 µg/kg body weight CRH blood is
taken regularly in a 2-hour period (0., 15., 30., 60., 90. and 120. minutes). If the pituitary and
adrenal functioning is normal, cortisol peak concentrations are higher than 19 µg/dl (525
nmol/l). This test is safe and its sensitivity is good, but its routine application is limited by the
high costs of CRH [7, 21]. (The most important characteristics of dynamic tests are summa-
rized in Table 1.)
Table 1 Summary of dynamic tests used for assessing HPA-axis functioning
Function Normal
Test Method Advantages Disadvantages
studied response

IHT entire hypoglycaemia cortisol suitable for the dangerous in elderly,

HPA axis induced by > 20 µg/dl correct study patients suffering from
0.1–0.15 U/kg (550 nmol/l) of the entire HPA cardiac problems and epilepsy,
body weight during and other pituitary requires hospital surveillance
regular insulin hypoglycaemia hormones and medical presence
i.v. (glu < 2.2 (e.g. growth
mmol/l), blood hormone)
taken at 0., 15.,
30., 45. and
60. min.

MT entire 30 mg/kg 11-deoxy- suitable for the study – rarely provokes adrenal
HPA axis metyrapone cortisol of the entire HPA, insufficiency
at midnight, > 7 µg/dl comfortable, safe, – limited availability
blood taken (200 nmol/l), even in outpatients of 11-deoxycortisol
next morning whereas measurement and
cortisol metyrapone itself
< 5 µg/dl
(140 nmol/l)

CRH pituitary 1 µg/kg CRH cortisol suitable for the study – expensive
and i.v., blood taken > 19 µg/dl of pituitary – mild complications:
adrenal during 2 hours (525 nmol/l) and adrenal activity, flushing, tachypnea
(–15., 0., 15., safe
30., 60., 90.,
120. min.)

Classical primary + 250 µg tetra- cortisol simple, safe, suitable Mild secondary adrenal
ST secondary cosactid i.v., > 20 µg/dl for the detection insufficiency is not always
adrenal blood taken (550 nmol/l) of primary, and also detected due to its high dose.
insuffi- 30 and/or long-standing It does not show secondary
ciency 60 min. later secondary adrenal adrenal insufficiency after
insufficiency hypothalamus-pituitary injury
for a few days, as the adrenal
cortex is not yet atrophised.

Low- secondary 1 µg tetra- cortisol simple, safe, it can be – not suitable for the
dose ST adrenal cosactid i.v., > 20 µg/dl more sensitive for detection of primary adrenal
insuffici- blood taken (550 nmol/l) the detection of mild insufficiency
ency 30 and/or secondary adrenal – its sensitivity is controversial
60 min. later insufficiency – tetracosactid is not available
in 1 µg vials commercially

HUNGARIAN MEDICAL JOURNAL n 67 n 2007 n Volume 1, Number 1

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 68

n ACTUALITIES IN CLINICAL PRACTICE n

The Possible Consequences of Glucocorticoid Withdrawal

i. reactivation of the underlying disease

The underlying disease may be reactivated during glucocorticoid withdrawal that could warrant
the reintroduction of glucocortoid therapy. The patient should be regularly controlled during
withdrawal and even after its completion. Due to the possibility of the reactivation, the with-
drawal can only be conducted with the cooperation of the physician treating the underlying dis-
order and the endocrinologist.

ii. secondary adrenal insufficiency

Secondary adrenal insufficiency is the most feared complication of withdrawal, since it may result
in an acute life-threatening condition. Adrenal insufficiency is, however, rarely observed, and is in
general associated with severe stress situations. In contrast with the classical symptoms of addis-
onian crisis, electrolyte abnormalities are infrequent, due to the relative preservation of the min-
eralocorticoid system. High doses of parenteral glucocorticoids are required to treat it [21].

iii. steroid withdrawal syndrome

The pathogenesis of this syndrome is the least elucidated among the complications associated
with glucocorticoid withdrawal. This condition is characterized by aspecific phenomena that
may also occur in adrenal insufficiency, including nausea, malaise, weakness, fatigue, weight
loss, joint pains, subfebrility, fever etc. [1, 7]. In contrast with the complaints and symptoms
presented, glucocorticoid substitution is appropriate in this condition and the dynamic tests
also give normal results. Central nervous system mechanisms are suspected in the background
of this syndrome (noradrenerg, dopaminerg systems, CRH, opioid peptides, prostaglandins),
that may be similar to the pathological processes involved in drug withdrawal syndromes. Such
mechanisms appear to participate in withdrawal syndromes developing after stopping the
administration of other hormones (estrogenes/menopause, anabolic steroids) that may indicate
the possibility of common pathways [28]. Inflammatory cytokines, mostly interleukin-6
released in high quantities, being liberated from the suppressive effects of glucocorticoids may
be involved in its pathogenesis [29]. Some patients experience psychologic dependence on glu-
cocorticoids that may further complicate withdrawal [28]. Steroid withdrawal syndrome may
develop after the recovery from pituitary- or adrenal-related endogenous hypercortisolism [30].
Steroid substitution dose should be transiently increased for its treatment [1, 7, 21].

How to Withdraw Glucocorticoid Therapy – a Recommendation

Glucocorticoid withdrawal can only be started if the underlying disease is in a long-lasting inac-
tive phase or in case of an active disease, if other effective treatment modality to suppress the
underlying pathology is introduced instead of glucocorticoids (e.g. other forms of immunosup-

2007 n Volume 1, Number 1 n 68 n HUNGARIAN MEDICAL JOURNAL

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 69

n ACTUALITIES IN CLINICAL PRACTICE n

pressive medicine in autoimmune diseases). The dose of the applied glucocorticoid should be
reduced to the substitution dose characteristic for the particular steroid (Table 2).
Glucocorticoid withdrawal can thus be initiated if the underlying disease is in a long-lasting
inactive phase with a steroid dose corresponding to the daily substitution dose or only slightly
surpassing it (e.g. 4–6 mg prednisolone/day).

Table 2 Most important characteristics of glucocorticoid drugs applied in clinical practice

Anti-inflammatory activity Daily substitution dose in

Glucocorticoid Biological half-life (hour)
in relation to cortisol adults (mg)

Hydrocortisone (cortisol) < 8–12 1 20

Prednisone 18–36 4 5

Prednisolone 18–26 4 5

Methylprednisolone 18–36 5 4

Triamcinolone 18–36 5 4

Dexamethasone 36–54 25–30 0.5–1

It is frequently observed that patients report complaints corresponding to steroid withdraw-

al syndrome, although the substitution dose is appropriate. In this case, glucocorticoid dose
should transiently be increased depending on the clinical picture, then parallel to the ameliora-
tion of symptoms, withdrawal can again be attempted.
There is no consensus regarding the protocol of steroid withdrawal. It can be stated that
withdrawal can only be performed gradually, by reducing glucocorticoid doses in a stepwise
manner. There are two major standpoints concerning the type of glucocorticoid applied. One
proposes the introduction of alternate day steroid intake regimen, without changing the gluco-
corticoid medication used before. The steroid should be taken in the morning, as the endoge-
nous ACTH-secretion is thus less influenced [21]. It is not recommended to turn to the alter-
nate-day abruptly, since symptoms of adrenal insufficiency may develop on days without the
steroid [5]. We could propose the following protocol for example: the daily 4 mg methylpred-
nisolone is first changed to an alternate day regimen of 2 and 4 mg, later, by omitting the 2 mg
dose, an alternate day 4 mg, and lastly an alternate day 2 mg regimen is followed.
The other standpoint suggests the use of hydrocortisone instead of the previously used other
synthetic glucocorticoids. Hydrocortisone that has the same chemical structure as endogenous
cortisol, suppresses HPA axis functioning to a lesser extent than other synthetic glucocorticoids.
Due to its shorter biological half-life, hydrocortisone diminishes the symptoms of Cushing’s
syndrome, on the other hand, it should be given at least twice a day. The evening dose should
be taken in the late afternoon, in order to promote ACTH-secretion. A 24 h suspension of its
administration is sufficient to enable a correct interpretation of morning cortisol determination
and dynamic tests. A further advantage over its synthetic counterparts is its higher substitution
dose (15–20 mg), thus its dose could be reduced by lesser steps, more gradually. Some authors
propose that the dose of hydrocortisone can be tapered by 2.5 mg every one-two weeks [17,
21]. As the renin-aldosterone system is hardly affected, drugs with mineralocorticoid activity
are not necessary to give. (The flow-chart of glucocorticoid withdrawal is depicted in Fig. 1.)

HUNGARIAN MEDICAL JOURNAL n 69 n 2007 n Volume 1, Number 1

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 70

n ACTUALITIES IN CLINICAL PRACTICE n

Initial condition for starting glucocorticoid withdrawal:

The underlying disease is inactive with a steroid dose corresponding to or only minimally surpassing
the substitution dose (or another treatment modality effective in suppressing the underlying disease
was introduced beside the substitution steroid dose)

Alternate day regimen Change to the short-acting hydrocortisone

with the same steroid drug as before that suppresses HPA to a lesser extent

Gradual reduction of dose during several weeks, months

with careful follow-up of the clinical picture

Determination of morning cortisol level after

suspending steroid substitution for at least 24 h

10–20 µg/dl
3–10 µg/dl
according to others > 20 µg/dl
according to others < 3 µg/dl
3-20 µg/dl

Dynamic test (ST, if indicated

IHT) to be performed

Subnormal response Normal response

HPA-functioning insufficient,
HPA-functioning appropriate
substitution should be
substitution can be omitted
continued

Fig. 1 A flow-chart depicting a suggested protocol for glucocorticoid withdrawal

Some authors propose that the clinical picture is sufficient for monitoring the withdrawal
process, others emphasize the necessity of morning cortisol level and/or dynamic tests [1, 7].
Laboratory examinations should only then be applied if the patient feels well even at a reduced
substitution dose.
If the morning cortisol level is under 3 µg/dl, substitution should be continued, if it is over
20 µg/dl, the withdrawal process can be regarded as finished and the substitution can be omit-
ted. If cortisol concentrations are between 3 and 20 µg/dl, dynamic tests should be conducted
[1]. It should be more rational, however, to perform dynamic tests if the morning cortisol levels
are over 10 µg/dl (270 nmol/l) [21]. If the response is normal, substitution should be
stopped. If a hyporesponsive phenomenon is observed, the dose should be maintained or it
should be carefully, slowly reduced further.
We propose short ST test to be performed based on its comfortability, safety and easy acces-
sibility. As the adrenal cortex regenerates later, as hypothalamic-pituitary functioning, a normal

2007 n Volume 1, Number 1 n 70 n HUNGARIAN MEDICAL JOURNAL

HMJ_2007_1_3tord 4/5/07 1:44 PM Page 71

n ACTUALITIES IN CLINICAL PRACTICE n

ST-response may indicate the normal functioning of the HPA axis. If ST shows a normal
response, but the clinical picture does not correlate with this – if not contraindicated –, IHT
should be performed. Some authors suggest MT to be performed [1], but its application is dif-
ficult due to its limited accessibility. Others recommend the CRH-test, which is also sensitive
and safe [21], but its costs do not enable its routine application.
Similar to patients suffering from adrenal insufficiency, during glucocorticoid withdrawal,
the substitution dose should be doubled or tripled in case of severe stress situations (e.g. infec-
tion, surgical intervention, accident), as the endogenous HPA axis is not capable of this on its
own. Moreover, we should know, and the patient should also be reminded to the fact that HPA
axis functioning could be insufficient even after one year after stopping glucocorticoid substitu-
tion, thus the patient should receive supplementary steroid during acute events [21]. Patients
should be provided with a card indicating the fact of steroid substitution and the necessity of
steroid supplementation during acute stress situations.

Concluding Remarks
The diagnostics and treatment of hypercortisolism belong to the most sophisticated problems
of clinical endocrinology. Iatrogenic (exogenous) forms of hypercortisolism are by far more fre-
quent than endogenous forms, nevertheless, the literature dealing with it is scarce in relation to
the other manifestations. In contrast with its outstanding clinical relevance, there is no consen-
sus and generally accepted guideline for the protocol of glucocorticoid withdrawal. Due to the
high interindividual variation, the protocol and timetable of glucocorticoid withdrawal can only
be determined individually, with the parallel examination of the clinical picture and laboratory
results. As severe complications, i.e. reactivation of the underlying disease, secondary adrenal
insufficiency and steroid withdrawal syndrome may appear during glucocorticoid withdrawal, it
can only be performed by the cooperation of the physician treating the underlying disease and
the endocrinologist, especially in those cases when the patient has been treated with high-dose
glucocorticoids for several years.

References
[1] Hopkins, R. L., Leinung, M. C.: Exogenous Cushing’s syndrome and glucocorticoid withdraw-
al. Endocrinol. Metab. Clin. North. Am., 2005, 34, 371–384.
[2] Newell-Price, J., Bertagna, X., Grossman, A. B. et al.: Cushing’s syndrome. Lancet, 2006, 367,
1605–1617.
[3] Jobling, A. I., Augusteyn, R. C.: What causes steroid cataracts? A review of steroid-induced pos-
terior subcapsular cataracts. Clin. Exp. Optom., 2002, 85, 61–75.
[4] Mankin, H. J.: Nontraumatic necrosis of bone (osteonecrosis). N. Eng. J. Med., 1992, 326,
1473–1479.
[5] Axelrod, L.: Glucocorticoid therapy. In: DeGroot, L. J., Jameson, J. L. Eds. Endocrinology.
Elsevier Saunders, Philadelphia, 2006, pp. 2329–2342.
[6] Kaplan, J. G., Barasch, E., Hirschfeld, A. et al.: Spinal epidural lipomatosis: a serious complica-
tion of iatrogenic Cushing’s syndrome. Neurology, 1989, 39, 1031–1034.
[7] Vollenweider, P., Waeber, G.: Planifier un sevrage aux glucocorticoïdes: stratégie diagnostique et
thérapeutique. Revue Praxis, 2003, 92, 1675–1682.
[8] Ermis, B., Ors, R., Tastekin, A. et al.: Cushing’s syndrome secondary to topical corticosteroids
abuse. Clin. Endocrinol., 2003, 58, 795–796.

HUNGARIAN MEDICAL JOURNAL n 71 n 2007 n Volume 1, Number 1

 HMJ_2007_1_3tord 4/5/07 1:44 PM Page 72

n ACTUALITIES IN CLINICAL PRACTICE n

[9] Kaliner, M. A.: Pharmacologic characteristics and adrenal suppression with newer inhaled corticos-
teroids: a comparison of ciclesonide and fluticasone propionate. Clin. Ther., 2006, 28, 319–331.
[10] Suliman, A. M., Smith, T. P., Labib, M. et al.: The low-dose ACTH test does not provide a use-
ful assessment of the hypothalamic-pituitary-adrenal-axis in secondary adrenal insufficiency.
Clin. Endocrinol., 2002, 56, 533–539.
[11] Fuchs, M., Wetzig, H., Kertscher, F. et al.: Iatrogenes Cushing-Syndrom und Mutatio tarda
durch Dexamethason-haltige Nasentropfen. HNO, 1999, 47, 647–650.
[12] Bolland, M. J., Bagg, W., Thomas, M. G. et al.: Cushing’s syndrome due to interaction between
inhaled corticosteroids and itraconazole. Ann. Pharmacother., 2004, 38, 46–49.
[13] Clevenbergh, P., Corcostegui, M., Gerard, D. et al. Iatrogenic Cushing’s syndrome in an HIV
infected patient treated with inhaled corticosteroids (fluticasone propionate) and low-dose
ritonavir enhanced PI containing regimen. J. Infect., 2002, 44, 194–195.
[14] Iglesias, P., Gonzalez, J., Diez, J. J.: Acute and persistent iatrogenic Cushing’s syndrome after a
single dose of triamcinolone acetonide. J. Endocrinol. Invest., 2005, 28, 1019–1023.
[15] Mann, M., Koller, E., Murgo, A. et al.: Glucocorticoidlike activity of megestrol. A summary of
Food and Drug Administration experience and a review of the literature. Arch. Intern. Med.,
1997, 157, 1651–1656.
[16] Richter, B., Neises, G., Clar, C.: Glucocorticoid withdrawal schemes in chronic medical disor-
ders. A systematic review. Endocrinol. Metab. Clin. North. Am., 2002, 31, 751–778.
[17] Stewart, P. M.: Chronic corticosteroid therapy, hypothalamo-pituitary-adrenal axis suppression,
and steroid withdrawal. In Larsen, P. R., Kronenberg, H. M., Melmed, S., Polonsky, K. S. Eds.
Williams Textbook of Endocrinology, 2003, Saunders, Philadelphia, 507–508.
[18] Henzen, C., Suter, A., Lerch, E. et al.: Suppression and recovery of adrenal response after short-
term high-dose glucocorticoid treatment. Lancet, 2000, 355, 542–545.
[19] Graber, A. L., Ney, R. L., Nicholson, W. E. et al.: Natural history of pituitary-adrenal recovery fol-
lowing long-term suppression with corticosteroids. J. Clin. Endocrinol. Metab., 1965, 25, 11–16.
[20] LaRochelle G. E., LaRochelle, A. G., Ratner, R. E. et al.: Recovery of the hypothalamic-pituitary-
adrenal (HPA) axis in patients with rheumatic diseases receiving low-dose prednisone. Am. J.
Med., 1993, 95, 258–264.
[21] Krasner, A. S.: Glucocorticoid-induced adrenal insufficiency. JAMA, 1999, 282, 671–676.
[22] Stewart, P. M.: Assessing adequacy of function of the hypothalamo-pituitary-adrenal axis. In
Larsen, P. R., Kronenberg, H. M., Melmed, S., Polonsky, K. S. Eds. Williams Textbook of
Endocrinology, 2003, Saunders, Philadelphia, pp. 530–531.
[23] Aron, D. C.: Diagnostic implications of adrenal physiology and clinical epidemiology for evalua-
tion of glucocorticoid excess and deficiency. In DeGroot, L. J., Jameson, J. L. Eds.
Endocrinology. Elsevier Saunders, Philadelphia, 2001, pp. 1655–1670.
[24] Fiad, T. M., Kirby, J. M., Cunningham, S. K. et al.: The overnight single-dose Metopyrone test
is a simple and reliable index of the hypothalamic-pituitary-adrenal axis. Clin. Endocrinol.,
1994, 40, 603–609.
[25] Bangar, V., Clayton, R. N.: How reliable is the short synacthen test for the investigation of the
hypothalamic-pituitary-adrenal axis? Eur. J. Endocrinol., 1998, 139, 580–583.
[26] Thaler, L. M., Blevins, L. S.: The low dose (1-µg) adrenocorticotropin stimulation test in the
evaluation of patients with suspected central adrenal insufficiency. J. Clin. Endocrinol. Metab.,
1998, 83, 2726–2729.
[27] Soule, S., van Zyl Smit, C., Parolis, G. et al.: The low dose ACTH stimulation test is less sensitive
than the overnight metyrapone test for the diagnosis of secondary hypoadrenalism. Clin.
Endocrinol., 2000, 53, 221–227.
[28] Hochberg, Z., Pacak, K., Chrousos, G. P.: Endocrine withdrawal syndromes. Endocr. Rev., 2003,
24, 523–538.
[29] Papanicolaou, D. A., Tsigos, C., Oldfield, E. H. et al.: Acute glucocorticoid deficiency is associat-
ed with plasma elevations of interleukin-6: does the latter participate in the symptomatology of
the steroid withdrawal syndrome and adrenal insufficiency? J. Clin. Endocrinol. Metab., 1996,
81, 2303–2306.
[30] Bhattachayya, A., Kaushal, K., Tymms, D. J. et al.: Steroid withdrawal syndrome after successful
treatment of Cushing’s syndrome: a reminder. Eur. J. Endocrinol., 2005, 153, 207–210.

2007 n Volume 1, Number 1 n 72 n HUNGARIAN MEDICAL JOURNAL

View publication stats