ORIGINAL ARTICLE

E n d o c r i n e C a r e

Ketoconazole in Cushing’s Disease: Is It Worth a Try?

Frederic Castinetti, Laurence Guignat, Pauline Giraud, Marie Muller, Peter Kamenicky,
Delphine Drui, Philippe Caron, Fiorina Luca, Bruno Donadille,

Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
Marie Christine Vantyghem, Helene Bihan, Brigitte Delemer, Gerald Raverot,
Emmanuelle Motte, Melanie Philippon, Isabelle Morange, Bernard Conte-Devolx,
Laurent Quinquis, Monique Martinie, Delphine Vezzosi, Maelle Le Bras,
Camille Baudry, Sophie Christin-Maitre, Bernard Goichot, Philippe Chanson,
Jacques Young, Olivier Chabre, Antoine Tabarin, Jerome Bertherat, and Thierry Brue
Aix Marseille Université, Hopital de la Timone (F.C., M.P., I.M., B.C.-D., T.B.), Service d’Endocrinologie, 13005,
Marseille, France; Hôpital Cochin Service d’Endocrinologie et Maladies Métaboliques (L.G., L.Q., C.B., J.B.),
75014, Paris, France; Centre Hospitalier Universitaire (CHU) de Bordeaux Hôpital du Haut Lévêque Service
d’Endocrinologie-Diabétologie et Maladies Métaboliques (P.G., A.T.), 33600 Pessac, France; CHU de
Grenoble Hôpital Albert Michallon Service d’Endocrinologie-Diabétologie-Nutrition (M.Mu., M.Ma., O.C.),
38043, Grenoble, France; Université Paris-Sud Service d’Endocrinologie et Maladies de la Reproduction (P.K.,
P.Ch., J.Y.), 94270, Le Kremlin Bicetre, France; CHU de Nantes Hôpital G & R Laënnec St-Herblain
Endocrinologie, Maladies Métaboliques et Nutrition (D.D., M.L.B.), 44093, Nantes, France; Hôpital Larrey
Service d’Endocrinologie-Maladies Métaboliques-Nutrition (P.Ca., D.V.), 31400, Toulouse, France; CHU de
Strasbourg Hôpital de Hautepierre (F.L., B.G.), Service de Médecine Interne et de Nutrition, 67100 Strasbourg,
France; Assistance Publique-Hôpitaux de Paris, Hôpital St-Antoine Service d’Endocrinologie-Diabétologie et
Médecine de la Reproduction (B.Do., S.C.-M.), 75012, Paris, France; Centre Hospitalier Regional Universitaire
de Lille Hôpital Claude Huriez, Service d’Endocrinologie Métabolisme (M.C.V.), 59000 Lille, France; Hôpital
Avicenne Service d’Endocrinologie, Diabétologie, et Maladies Métaboliques (H.B.), 93000 Bobigny, France;
CHU de Reims Hôpital Robert Debré (B.De.), Service d’Endocrinologie-Diabète-Nutrition, 51092, Reims,
France; Hôpital Neuro-cardiologique Fédération d’Endocrinologie du Pôle Est (G.R.), 39500 Bron, France; and
Hôpital de Bicêtre Endocrinologie Pédiatrique (E.M.), 94270, Le Kremlin Bicetre, France

Background: The use of ketoconazole has been recently questioned after warnings from the European
Medicine Agencies and the Food and Drug Administration due to potential hepatotoxicity. However,
ketoconazole is frequently used as a drug to lower circulating cortisol levels. Several pharmacological
agents have recently been approved for the treatment of Cushing’s disease (CD) despite limited efficacy
or significant side effects. Ketoconazole has been used worldwide for more than 30 years in CD, but in
the absence of a large-scale study, its efficacy and tolerance are still under debate.
Patients and methods: We conducted a French retrospective multicenter study reviewing data
from patients treated by ketoconazole as a single agent for CD, with the aim of clarifying efficacy
and tolerance to better determine the benefit/risk balance.
Results: Data from 200 patients were included in this study. At the last follow-up, 49.3% of patients
had normal urinary free cortisol (UFC) levels, 25.6% had at least a 50% decrease, and 25.4% had
unchanged UFC levels. The median final dose of ketoconazole was 600 mg/d. Forty patients (20%)
received ketoconazole as a presurgical treatment; 40% to 50% of these patients showed improve-
ment of hypertension, hypokalemia, and diabetes, and 48.7% had normal UFC before surgery.
Overall, 41 patients (20.5%) stopped the treatment due to poor tolerance. Mild (⬍5N, inferior to
5-fold normal values) and major (⬎5N, superior to 5-fold normal values) increases in liver enzymes
were observed in 13.5% and 2.5% of patients, respectively. No fatal hepatitis was observed.
Conclusions: Ketoconazole is an effective drug with acceptable side effects. It should be used under
close liver enzyme monitoring. Hepatotoxicity is usually mild and resolves after drug withdrawal.
(J Clin Endocrinol Metab 99: 1623–1630, 2014)

ISSN Print 0021-972X ISSN Online 1945-7197 Abbreviations: BP, blood pressure; CD, Cushing’s disease; MRI, magnetic resonance im-
Printed in U.S.A. aging; UFC, urinary free cortisol; ULN, upper limit of normal.
Copyright © 2014 by the Endocrine Society
Received September 28, 2013. Accepted January 16, 2014.
First Published Online January 28, 2014

For editorial see page 1586

doi: 10.1210/jc.2013-3628 J Clin Endocrinol Metab, May 2014, 99(5):1623–1630 jcem.endojournals.org 1623
1624 Castinetti et al Ketoconazole in Cushing’s Disease
ecently, the Food and Drug Administration and the
R European Medicines Agency have either suspended or
released strong warnings regarding the use of ketoconazole
as an antifungal agent due to its potential association with
cases of severe hepatotoxicity. Ketoconazole is an imidazole
derivative primarily used for its antifungal activity. How-
ever, in the early 1980s, the drug was also shown to have
effects on steroid synthesis in humans by inhibiting cyto-
chrome P450 enzymes and significantly reducing cortisol lev-
els (1–5). This drug has thus increasingly been used in the
management of Cushing’s disease (CD) (6).
However, treatment of CD still remains challenging (7,
8). Although transsphenoidal surgery induces short-term
remission in 60% to 80% of cases, long-term follow-up
studies have demonstrated an overall risk of 20% to 30%
for recurrence, even in patients with immediate postsur-
gical corticotroph deficiency (9). Previously published
studies suggest that ketoconazole has potent antisecretory
efficacy in 50% to 90% of cases, with a relatively good
tolerance profile (10 –19). However, most of these studies
were small, with only 2 studies enrolling more than 10
patients each, and only 100 patients being studied in total
(11, 18). Moreover, follow-up was limited. The benefit/
risk balance for use of ketoconazole in CD is therefore
difficult to determine. The question of whether the hepa-
totoxicity warnings in place recommending against use of
ketoconazole as an antifungal agent should also apply to
its use in CD remains unanswered.
To further improve our knowledge on the efficacy and
tolerance of ketoconazole in CD, we initiated the French
Retrospective Study on Ketoconazole Outcome (FReSKO).
Data were included from CD patients who had been treated
with ketoconazole and followed up in tertiary reference cen-
ters. Our main objectives were to evaluate efficacy and tol-
erance, particularly hepatic tolerance, to clarify whether this
drug should still be considered in therapeutic algorithms of
hypercortisolism.
Patients and Methods
Data from 200 patients followed in 14 French centers were
included in this retrospective study. All patients were treated
with ketoconazole as a single treatment for active CD be-
tween 1995 and 2012. The diagnosis of CD was based on
criteria defined according to current guidelines (20). The di-
agnosis of CD was confirmed by intrapetrosal sinus sampling
stimulated with CRH or the desmopressin test (21) when
pituitary magnetic resonance imaging (MRI) was considered
negative. Before treatment with ketoconazole, each patient
underwent a complete clinical and hormonal evaluation, in-
cluding measurements of circulating levels of ACTH and cor-
J Clin Endocrinol Metab, May 2014, 99(5):1623–1630
tisol at 8:00 AM and midnight and of at least 2 samples of
24-hour urinary free cortisol (UFC). Mean UFC was nor-
malized to maximal normal value for each kit available in
each center (upper limit of normal [ULN]). For each patient,
the following data were recorded: sex, age at diagnosis, pre-
vious treatments (surgery or radiotherapy), initial clinical
Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
signs, hypertension, diabetes, hypokalemia, initial 24-hour
UFC (mean of 2–3 samples), reason for ketoconazole treat-
ment, and initial dose.
The first evaluation of UFC (mean of 2–3 samples) was
carried out in all centers after 0.25 to 1 month of treat-
ment. Urinary cortisol secretion was usually monitored at
1- to 4-month intervals, and if necessary, ketoconazole
dosage was increased by 200 mg/d every 7 to 28 days
depending on the investigator’s judgment until normal-
ization was achieved. During ketoconazole treatment,
complete clinical evaluation (clinical signs of hypercorti-
solism, weight, and blood pressure [BP]) as well as stan-
dard biological evaluations including blood glucose and
liver function tests (repeated at each titration step) was
periodically performed. For each patient, the following
data were recorded: final UFC, final dose, length of keto-
conazole treatment, length of follow-up period, and rea-
son for withdrawal. Patients were considered controlled if
they had normal 24-hour UFC at 2 consecutive evalua-
tions. Partial control was defined as a decrease in UFC of
more than 50% without normalization. A lack of control
was defined by a decrease in UFC of less than 50% and/or
immediate clinical or biological intolerance leading to ke-
toconazole discontinuation.
When the dose of ketoconazole was finally established,
biological evaluation of hypercortisolism was performed at
3- to 6-month intervals. Patients who had signs of immediate
clinical or biological intolerance were considered uncon-
trolled, and the ketoconazole treatment was stopped. When
ketoconazole was given for more than 1 year, MRI was per-
formed at 6- to 12-month intervals, depending on the inves-
tigator’s judgment, until treatment was stopped. Each inves-
tigator also evaluated the changes in clinical signs of
hypercortisolism including BP as well as plasma potassium
and glucose tolerance. Improvement in hypertension was de-
fined as a decrease of at least 10 mm Hg of systolic and/or
diastolic BP in patients with hypertension. Improved glyce-
mic control was defined as 1 or more of the following: a
decrease of insulin dose (⬎10% of the total dose), a decrease
in the number of antidiabetic drugs, and an improvement of
hemoglobin A1c (⬎0.5% when available) without addition
of other antidiabetic drugs.
Regular monitoring (every 7–15 days) of aspartate ami-
notransferase, alanine aminotransferase, and ␥-glutamyl
transpeptidase was performed during the first month of
prescription and then at each dose change. Increases in any
doi: 10.1210/jc.2013-3628
of these parameters as well as any other adverse effects
were recorded.
All patients had received detailed information on the side
effects and potential benefits of ketoconazole use in CD.
Each patient had given informed consent allowing retrospec-
tive analyses of their medical records for scientific purposes,
as approved by the ethics committees of each institution. Of
note, since 2010, ketoconazole can be prescribed in France
only after obtaining a temporary authorization of use from
the French Health Authorities. The authorities must also be
notified of data from regular monitoring of efficacy and tol-
erability, so that the treatment remains authorized.
Statistical analysis was performed with XLStat version
2013.1.01 (Addinsoft). Quantitative data are presented as
mean ⫾SD (range). Statistical comparison of quantitative
data was performed by Student’s t test or by ANOVA.
Statistical comparison of qualitative data was performed
by ␹2 test or Fisher’s exact test when the theoretical num-
ber of patients was less than 5. All statistical tests were
two-tailed, and P ⬍ .05 was considered significant.
Results
Characteristics of the population before
ketoconazole use
In total, data from 200 patients (44 males and 156
females) were included in this retrospective study (Figure
1). The mean age at diagnosis of CD was 41.9 ⫾ 15.8
(range, 8 – 87) years. Pituitary MRI reported a microad-
enoma in 106 patients (53.4%), a macroadenoma in 36
patients (18.2%), and a lack of obvious adenoma in 58
patients (29.4%). Most patients (n ⫽ 144; 72%) were
treated by transsphenoidal surgery (mean age of 39.3 ⫾
18.3 years), whereas 47 patients (23.6%) with a mean age
Figure 1. Flow chart summarizing the main characteristics of the cohort and the main results. .0001). Antisecretory efficacy could
Abbreviation: CI, contraindication.
jcem.endojournals.org 1625
of 43 ⫾ 13 years were treated by radiotherapy. Sixteen
patients (8%) had a second surgery.
At the time of ketoconazole initiation, all patients had
clinical CD, 116 of 174 patients (66.7%) had hyperten-
sion, 39 of 174 (22.4%) hypokalemia, and 55 of 174 pa-
tients (31.8%) had diabetes mellitus. Data were not de-
Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
tailed for 26 patients. Mean initial 24-hour UFC was 4.1 ⫾
5.3 times the ULN (range, 1.1– 40).
Ketoconazole as a presurgical treatment
Forty patients were treated with ketoconazole as a pre-
surgical treatment. Initial UFC was statistically increased
in this subgroup of patients compared with patients
treated with ketoconazole given for other indications
(mean 6.89 ⫾ 10.4 vs 3.3 ⫾ 2.14 times the ULN, P ⫽
.003). The mean initial ketoconazole dose in these patients
was 585 ⫾ 242.4 (range 200 –1200) mg/d, and the mean
final dose was 755 ⫾ 284 (range 200 –1200) mg/d.
At the end of the treatment period (mean duration of
4.05 ⫾ 4.1 months, range 0.03–15 months), ie, before
surgery, 48.7% of 39 patients had normal UFC, whereas
UFC remained above the normal range despite a 50%
decrease in 35.9% of patients. One patient who had nor-
mal UFC at diagnosis but increased midnight salivary cor-
tisol was not evaluated for UFC at the last follow-up.
Mean final UFC was 2.5 ⫾ 6.4 (range 1– 40) times the
ULN, P ⫽ .028 vs initial UFC). Clinical signs were im-
proved in 16 of 38 patients (42.1%), hypertension was
improved in 13 of 26 patients (50%), diabetes was im-
proved in 8 of 16 patients (50%), and hypokalemia was
improved in 5 of 13 patients (38.4%) (Figure 1).
Ketoconazole as a primary or secondary treatment
Ketoconazole was given to 160 patients as a primary
treatment because of contraindica-
tion or refusal of surgery (n ⫽ 32,
20%) and as a secondary treatment
because of unsuccessful surgery (n ⫽
93, 58.1%) or to control hypersecre-
tion while waiting for radiotherapy
to be effective (n ⫽ 35, 21.8%) (Fig-
ure 1). In these patients, ketocona-
zole was initiated at a mean dose of
542.7 ⫾ 198.7 (range 200 –1200)
mg/d for a mean period of 24.8⫾
33.6 (range 0.2–135) months.
At the last follow-up, the mean fi-
nal dose of ketoconazole was
779.5.1 ⫾ 292.2 (range 200 –1200)
mg/d (final vs initial dose, P ⬍
be evaluated in 158 patients; 2 pa-
1626 Castinetti et al Ketoconazole in Cushing’s Disease J Clin Endocrinol Metab, May 2014, 99(5):1623–1630

months). Eight patients had contraindications to surgery

and were treated by ketoconazole as a first-line treatment.
At the last follow-up, mean final UFC was 1.48 ⫾ 1.27-
fold ULN. At the last visit, hypertension was improved in
15 of 27 patients (55.5%), diabetes in 7 of 14 patients
(50%), and hypokalemia in 7 of 8 patients (87.5%). UFC

Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
was normalized in 33 of 51 patients (64.7%), and it had
decreased by at least 50% in 12 of 51 patients (23.5%).
The 6 remaining patients (11.7%) had a final increase in
UFC despite initial control after a mean time of 64.3 ⫾
27.3 (range 24.4 –105.6) months, and treatment had to be
modified.

Antisecretory efficacy and predictive factors of

Figure 2. The 24-hour UFC (fold upper limit normal level ULN) before
(E) and after (‚) ketoconazole use. Patients are classified based on
final dose of ketoconazole (represented by the numbers above UFC
values, in milligrams per day) and then initial UFC.
tients who had normal UFC at diagnosis but increased
midnight salivary cortisol were not evaluated for UFC at
the last follow-up. At the last visit, 78 patients (49.3%)
were controlled, 37 patients (23.4%) had partial control
with at least 50% decrease of UFC (without normaliza-
tion), and 43 patients (27.2%) had unchanged UFC levels.
Mean final 24-hour UFC was 1.8 ⫾ 1.6-fold ULN (range
1– 40) (final vs initial UFC, P ⬍ .0001). More precisely,
mean final UFC was 1 ⫾ 0.2, 1.8 ⫾ 1, and 4.4 ⫾ 6 in the
controlled, partially controlled, and uncontrolled groups,
respectively (P ⬍ .001). At the last follow-up, clinical signs
were improved in 74 of 134 patients (55.2%), hyperten-
sion in 36 of 90 patients (40%), hypokalemia in 10 of 26
patients (38.4%), and diabetes in 23 of 39 patients (59%)
(Figure 2).
Long-term treatment (more than 24 months)
Fifty-one of the 160 patients were treated with keto-
conazole for more than 24 months (mean treatment du-
ration, 108.5 ⫾ 244.4 months, ranging from 24.1–135
control
Descriptive data between the 3 groups (control, partial
control, and failure) and predictive factors for secretion
control are shown in Table 1. No significant difference
was observed in terms of age at diagnosis, previous treat-
ments, and initial dose. Surprisingly, gender appeared to
be a predictive factor despite the fact that the maximal
dose was not statistically different between both groups
(750 ⫾ 236.7 vs 716 ⫾ 281.5 mg/d in males vs females,
respectively, P ⫽ .471). As expected, the total duration of
the treatment was significantly lower in uncured patients
(P ⬍ .001). Interestingly, initial UFC was not statistically
different between each group (P ⫽ .118). The final post-
ketoconazole UFC for each patient is shown in Figure 2.
Drug withdrawal and tolerance
The main reasons for patients’ withdrawal from keto-
conazole are given in Table 2 for 118 patients. This anal-
ysis does not include the 40 patients taking ketoconazole
as a presurgery treatment or the 42 patients still taking
ketoconazole at the last visit.
As shown in Table 3, intolerance data were available
for 190 patients. Forty-one patients (20.5%) stopped the
treatment due to poor tolerance.

Table 1. Characteristics of the Patients Depending on Their Final Status in Terms of Control of Hypersecretiona
Control Partial Control No Efficacy P
Sex ratio (female/male) 80/17 (82.4/17.5%) 42/9 (82.3/17.7%) 31/18 (63.2/36.7%) .013
Initial MRI 52 (53.6%) 30 (58.8%) 21 (42.8%) .216
Microadenoma 13 (13.4%) 11 (21.6%) 12 (24.5%)
Macroadenoma 32 (33%) 10 (19.6%) 16 (32.6%)
No image
Previous surgery 66 (68%) 39 (76.4%) 36 (73.4%) .472
Previous radiotherapy 25 (25.7%) 10 (19.7%) 12 (25%) .723
Age at diagnosis, y 43.5 ⫾ 14.7 (19 – 87) 40.5 ⫾ 17.4 (13– 85) 40.3 ⫾ 16.2 (8 –72) .402
Initial UFC (⫻ULN) 3.4 ⫾ 2.4 (1.1–11) 5.2 ⫾ 7.8 (1.1–11.6) 4.5 ⫾ 6.1 (1.1– 40) .118
Final dose, mg/d 668 ⫾ 264 (200 –1200) 776 ⫾ 248 (200 –1200) 780 ⫾ 292 (200 –1200) .424
Duration of treatment, mo 27.6 ⫾ 36.4 (0.4 –135) 18.2 ⫾ 29.2 (0.3–135) 9.7 ⫾ 14.9 (0.03– 66) ⬍.001
a
Partial control was defined by at least a 50% decrease of UFC without normalization and control by normal UFC at last follow-up. P ⬍ .05 was
considered significant. Quantitative data are given as mean ⫾ SD (range); n ⫽ 197.
doi: 10.1210/jc.2013-3628 jcem.endojournals.org 1627

Table 2. Reasons for Ketoconazole Withdrawala whatever the severity of hepatic enzyme disturbance. No
increase in liver enzyme levels was observed after the first
Reasons for ketoconazole withdrawal month of treatment or dose increase performed after the
(118/160 patients) n (%)
first month of treatment. As a consequence, the duration
Lack of efficacy 43 (26.8)
Patients with initial UFC control and 11 (6.9)
of treatment was significantly lower in patients with vs
without hepatic enzyme abnormalities (P ⬍ .001).

Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
secondary failure of the drug
Adverse effects (see Table 3) 41 (25.6) Other nonhepatic side effects reported are shown in
Decision to perform another treatment despite 22 (13.7) Table 3. Overall, these side effects were observed at a mean
ketoconazole efficacy
Bilateral adrenalectomy 7 (4.4) dose of 713.7 ⫾ 294 mg/d. Interestingly, 10 patients
Transsphenoidal surgery 15 (9.3) (5.4%) with a mean dose of 700 ⫾ 256 mg/d (ranging from
Visualization of adenoma on MRI (considered 8 (5) 400 –1200 mg/d) presented clinical and biological adrenal
normal before ketoconazole)
Radiotherapy efficacy 8 (5)
insufficiency. Three of these patients were being treated
Patient’s decision 3 (1.8) with a block and replace strategy; they had moderately
Pregnancy 1 (0.6) increased initial UFC (1.1–3 times the ULN) and normal
a
At their last visit, 42 patients were still on treatment. Forty patients UFC on ketoconazole; however, the dose was increased by
with ketoconazole as a presurgical treatment are not included in this 200 mg/d (up to 1200 mg/d) and hydrocortisone was
table.
added (20 mg/d).

Increase in liver enzyme levels up to 5-fold normal val-

ues was reported in 30 of 190 patients (15.8%). Four
Discussion
patients presented 5- to 10-fold increase, and 1 patient
(treated at a dose of 600 mg/d) presented a 40-fold increase Our data showed that ketoconazole is a highly effective
of aspartate aminotransferase, alanine aminotransferase, drug to reduce hypercortisolemia; 75.2% of patients had
and ␥-glutamyl transpeptidase, which remained high for 3 at least a 50% decrease of UFC, including 49% with nor-
weeks after ketoconazole withdrawal. This patient was mal UFC at the last follow-up. Surprisingly, of the 49
also concomitantly taking alcohol and acetaminophen. patients with lack of control, only 9 reached the final dose
Liver enzyme levels in this patient eventually returned to of 1200 mg/d. Because 17 patients did not have a dose
normal 90 days after ketoconazole withdrawal. In all increase because of intolerance, this means that 32 pa-
other patients, liver enzyme levels returned to normal tients did not have the maximal dose despite incomplete
within 2 to 4 weeks after a dose decrease (50% of cases) efficacy and good tolerance, based on investigators’ deci-
or withdrawal (50% of cases). No fatal hepatitis was ob-
sions. The overall results of efficacy in our study were
served. The mean ketoconazole dose at the time of the liver
concordant with smaller studies previously reported (11,
enzyme increase was not different between patients with
13, 15, 16, 18, 19) based on ketoconazole use as a single
(765 ⫾ 293 mg/d, range 200 –1200) and without (716 ⫾
drug or associated with other drugs lowering circulating
268 mg/d, range 200 –1200) liver intolerance (P ⫽ .491),
cortisol levels (22, 23). Surprisingly, gender appeared to be
a predictive factor of control, because the proportion of
Table 3. Adverse Effects Induced by Ketoconazolea males was higher in the uncontrolled than in the controlled
Mean Dose group. This was not due to a dose effect because there was
Frequency (mg/d) Min–Max no significant difference between the doses given to males
Liver enzyme increase 30 (15.8%) 772.4 ⫾ 305.7 400 –1200 and females (P ⫽ .471). Because most males (59%) had at
Gastrointestinal 25 (13.1%) 625 ⫾ 258.3 400 –1200 least a 50% decrease of UFC (including 39% with normal
complaints
UFC at the last visit), restricting the use of ketoconazole in
Adrenal insufficiency 10 (5.4%) 700 ⫾ 256 400 –1200
Pruritus 7 (3.7%) 700 ⫾ 385.6 400 –1200 males is not recommended. We could not identify any
Intense fatigue 2 (1.25%) 700 600 – 800 other predictive factor of control of hypercortisolism, in-
Hair loss 2 (1.25%) 700 600 – 800 cluding initial UFC. This finding is in contrast to pituitary-
Leg edema 2 (1.25%) 800 800 – 800
Muscle pain 2 (1.25%) 700 200 –1200 targeted drugs lowering ACTH and cortisol circulating
Dyspnea 1 (0.6%) 400 400 levels. For example, moderate hypercortisolism is thought
Hypertriglyceridemia 1 (0.6%) 800 800 to be a predictive factor for pasireotide and cabergoline
Leukoneutropenia 1 (0.6%) 600 600
Dizziness 1 (0.6%) 1200 1200 efficacy (24, 25). The antisecretory effect was usually
Increased creatinine 1 (0.6%) 600 600 maintained in the long term, which had been very rarely
level reported in the literature due to the lack of long-term fol-
a
For these results, n ⫽ 190. low-up data. In most patients, the increase of ACTH in-
1628 Castinetti et al Ketoconazole in Cushing’s Disease
duced by long-term cortisol inhibition may be expected to
eventually lead to cortisol escape. However, in this study,
only 15% of the patients treated for more than 24 months
had a final increase in UFC, despite initial control. Most of
them were finally treated by bilateral adrenalectomy. This
lack of escape could be explained by a pituitary effect of
ketoconazole with ACTH blockade, as previously shown
(26, 27), even if this central effect remains a matter of debate
(28). However, it is noticeable that the escape phenomenon
is almost never seen in patients with CD treated with me-
tyrapone once the effective dose has been found (29).
Interestingly, 40 patients were treated with ketocona-
zole as a presurgical treatment. Although this approach
remains a matter of debate due to the lack of large pro-
spective studies on potential benefits, approximately half
of the patients had improvement in diabetes, hyperten-
sion, and hypokalemia before surgery, confirming the
rapid efficacy of the drug. To our knowledge, only 2 stud-
ies reported the potential effects of ketoconazole as a pre-
surgical treatment of CD: in the first, of a total of 52 pa-
tients, 17 were treated with ketoconazole for a mean
period of 4 months. Patients had a decrease in BP and
hemoglobin A1c levels before surgery. It was also sug-
gested that pretreatment could lead to a lower risk of post-
surgical ACTH deficiency, but this might make immediate
postsurgical evaluation difficult to interpret (30). The sec-
ond, based on 20 patients treated with ketoconazole, re-
ported a higher long-term rate of remission in patients
with pretreatment (31). Our study was not designed to
define the benefits of such an approach, which might be of
importance in patients with uncontrolled side effects of
severe hypercortisolism.
The most significant issue with ketoconazole treatment
is a potentially increased risk of hepatotoxicity. Fatal hep-
atitis has indeed been reported in a few patients treated
with ketoconazole as an antifungal agent (dose range,
200 – 400 mg daily) (32– 40). Here, increased liver enzyme
levels were reported in 18.4% of patients. However, the
majority (85%) had less than a 5-fold increase, and liver
enzymes returned to normal levels within 1 to 2 weeks
after a dose decrease of 200 mg/d or treatment with-
drawal. It is of note, however, that 5 patients had increases
of more than 5-fold normal values, including 1 patient
who had up to a 40-fold increase. However, in this latter
case, hepatotoxicity was multifactorial; the patient was
also taking alcohol and high doses of acetaminophen.
Liver enzyme levels returned to normal within 1 to 3
months after drug withdrawal in all cases. No fatal hep-
atitis was observed. Of note, hepatotoxicity was not dose-
dependent, but it occurred at the beginning of the treat-
ment or rapidly after the dose change. Therefore, this
suggests that a close follow-up of liver enzymes (once
J Clin Endocrinol Metab, May 2014, 99(5):1623–1630
weekly) for 1 month after a dose change could potentially
avoid any risks of severe hepatotoxicity. All potential as-
sociated factors of hepatotoxicity and abnormal liver en-
zyme levels detected before ketoconazole initiation should
be taken into account before considering ketoconazole as
a treatment. Our results suggest that, first, in case of an
Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
increase in liver enzyme levels less than 5-fold normal val-
ues, the dose should be decreased to at least 200 mg under
weekly surveillance and, second, that treatment should be
stopped if liver enzyme levels increase up to 5-fold normal
values with close monitoring until normalization. Of note,
because this was a retrospective study, monitoring of liver
enzymes was not homogeneous across all the centers; each
investigator thus decided to decrease the dose and follow
closely (and stop in case of further increase) when there
was a moderate liver enzyme increase (less than 5-fold)
because they considered that the benefit/risk balance was
in favor of maintaining ketoconazole. Because our study
was not a prospective study primarily aiming at better
defining this side effect, there is obviously no evidence that
such an approach should be the preferred one in compar-
ison with a systematic withdrawal when liver enzymes
increase moderately. Other side effects reported at a me-
dian dose of 600 mg/d (which was the standard dose used
long-term in most patients) included a potential risk of
adrenal insufficiency in 3.7% of cases (excluding patients
with a block and replace strategy), reported here for the
first time in the treatment of hypercortisolism (41).
The present study had obvious limitations due to its
retrospective nature. However, a prospective randomized
study design is difficult to consider here because of the
rarity of CD and because ketoconazole had been used for
this indication for more than 30 years. Moreover, identi-
fying a drug that should be used as the reference treatment
currently remains a matter of debate. Because the moni-
toring data were extracted retrospectively in this study,
each investigator followed their patients according to their
own judgment, and there was no standardized follow-up.
Some data were therefore difficult to recover. For exam-
ple, although we asked each physician to assess whether
clinical signs, glycemic control, hypertension, and hypo-
kalemia had improved, worsened, or remained unchanged
during the treatment, we were not able to quantitatively
evaluate these modifications. Finally, we chose to use UFC
measurements as a way to define the control of the disease
because midnight salivary cortisol was not available in all
patients due to the long periods of treatment. Reproduc-
ibility in UFC measurements can sometimes be an issue,
but in this study, it was resolved by averaging 2 to 3 UFC
samples.
Formal comparison with other drugs is difficult. Anal-
ysis of regular clinical practice provides an estimate that is
doi: 10.1210/jc.2013-3628
not based on robust analysis and is subject to many biases.
However, it may be representative of the response and side
effects rates expected in a tertiary healthcare center. Other
drugs were not shown to have a major superiority com-
pared with ketoconazole. Despite the retrospective design
of our study, our results suggest that ketoconazole com-
pares favorably with pasireotide with regard to the per-
centage of patients with normal UFC and with side effects,
particularly diabetes, which was slightly improved by ke-
toconazole use. Mifepristone is probably at least as effec-
tive as ketoconazole for improving metabolic parameters,
but monitoring of the drug might be a major issue for
long-term treatments. Finally, the paucity of reports on
metyrapone and cabergoline does not allow accurate com-
parison of the benefit to risk ratio. To conclude, keto-
conazole is a highly effective drug lowering circulating
cortisol levels. In this study, hepatotoxicity was a rela-
tively frequent event but remained moderate in most the
patients. In the absence of other hepatotoxic factors,
weekly monitoring of liver enzymes for 1 month at keto-
conazole initiation and dose changes could potentially al-
low early detection and therefore prevention of severe hep-
atotoxicity. Under these conditions, we believe the benefit/
risk balance should remain in favor of ketoconazole use in
patients requiring medical treatment for CD.
Acknowledgments
Address all correspondence and requests for reprints to: Frederic
Castinetti, MD, PhD, Department of Endocrinology, La Timone
Hospital, Rue Saint Pierre, 13005 Marseille, France. E-mail:
Frederic.castinetti@ap-hm.fr.
Disclosure Summary: The authors have nothing to disclose.
References
1. Angeli A, Frairia R. Ketoconazole therapy in Cushing’s disease. Lan-
cet. 1985;1:821.
2. Bradbrook ID, Gillies HC, Morrison PJ, Robinson J, Rogers HJ,
Spector RG. Effects of single and multiple doses of ketoconazole on
adrenal function in normal subjects. Br J Clin Pharmacol. 1985;20:
163–165.
3. Couch RM, Muller J, Perry YS, Winter JS. Kinetic analysis of inhi-
bition of human adrenal steroidogenesis by ketoconazole. J Clin
Endocrinol Metab. 1987;65:551–554.
4. Dandona P, Mohiuddin J, Prentice HG. Ketoconazole and adreno-
cortical secretion. Lancet 1985;1:227.
5. Pont A, Williams PL, Loose DS, et al. Ketoconazole blocks adrenal
steroid synthesis. Ann Intern Med. 1982;97:370 –372.
6. Feelders RA, Hofland LJ, de Herder WW. Medical treatment of
Cushing’s syndrome: adrenal-blocking drugs and ketaconazole.
Neuroendocrinology. 2010;92(Suppl 1):111–115.
7. Bertagna X, Guignat L. Approach to the Cushing’s disease patient
with persistent/recurrent hypercortisolism after pituitary surgery.
J Clin Endocrinol Metab. 2013;98:1307–1318.
8. Biller BM, Grossman AB, Stewart PM, et al. Treatment of adreno-
jcem.endojournals.org 1629
corticotropin-dependent Cushing’s syndrome: a consensus state-
ment. J Clin Endocrinol Metab. 2008;93:2454 –2462.
9. Patil CG, Prevedello DM, Lad SP, et al. Late recurrences of Cushing’s
disease after initial successful transsphenoidal surgery. J Clin En-
docrinol Metab. 2008;93:358 –362.
10. Cannavò S, Almoto B, Dall’Asta C, et al. Long-term results of treat-
ment in patients with ACTH-secreting pituitary macroadenomas.
Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
Eur J Endocrinol. 2003;149:195–200.
11. Castinetti F, Morange I, Jaquet P, Conte-Devolx B, Brue T. Keto-
conazole revisited: a preoperative or postoperative treatment in
Cushing’s disease. Eur J Endocrinol. 2008;158:91–99.
12. Cerdas S, Billaud L, Guilhaume B, Laudat MH, Bertagna X, Luton
JP. Short term effects of ketoconazole in Cushing’s syndrome [in
French]. Ann Endocrinol (Paris). 1989;50:489 – 496.
13. Chou SC, Lin JD. Long-term effects of ketoconazole in the treatment
of residual or recurrent Cushing’s disease. Endocr J. 2000;47:401–
406.
14. Leal-Cerro A, Garcia-Luna PP, Villar J, et al. Arterial hypertension
as a complication of prolonged ketoconazole treatment. J Hypertens
Suppl. 1989;7:S212–S213.
15. Loli P, Berselli ME, Tagliaferri M. Use of ketoconazole in the treat-
ment of Cushing’s syndrome. J Clin Endocrinol Metab. 1986;63:
1365–1371.
16. McCance DR, Hadden DR, Kennedy L, Sheridan B, Atkinson AB.
Clinical experience with ketoconazole as a therapy for patients with
Cushing’s syndrome. Clin Endocrinol (Oxf). 1987;27:593–599.
17. Sonino N. The use of ketoconazole as an inhibitor of steroid pro-
duction. N Engl J Med. 1987;317:812– 818.
18. Sonino N, Boscaro M, Paoletta A, Mantero F, Ziliotto D. Keto-
conazole treatment in Cushing’s syndrome: experience in 34 pa-
tients. Clin Endocrinol (Oxf). 1991;35:347–352.
19. Tabarin A, Navarranne A, Guerin J, Corcuff JB, Parneix M, Roger
P. Use of ketoconazole in the treatment of Cushing’s disease and
ectopic ACTH syndrome. Clin Endocrinol (Oxf). 1991;34:63– 69.
20. Arnaldi G, Angeli A, Atkinson AB, et al. Diagnosis and complica-
tions of Cushing’s syndrome: a consensus statement. J Clin Endo-
crinol Metab. 2003;88:5593–5602.
21. Castinetti F, Morange I, Dufour H, et al. Desmopressin test during
petrosal sinus sampling: a valuable tool to discriminate pituitary or
ectopic ACTH-dependent Cushing’s syndrome. Eur J Endocrinol.
2007;157:271–277.
22. Feelders RA, de Bruin C, Pereira AM, et al. Pasireotide alone or with
cabergoline and ketoconazole in Cushing’s disease. N Engl J Med.
2010;362:1846 –1848.
23. Kamenicky P, Droumaguet C, Salenave S, et al. Mitotane, metyrap-
one, and ketoconazole combination therapy as an alternative to
rescue adrenalectomy for severe ACTH-dependent Cushing’s syn-
drome. J Clin Endocrinol Metab. 2011;96:2796 –2804.
24. Colao A, Petersenn S, Newell-Price J, et al. A 12-month phase 3
study of pasireotide in Cushing’s disease. N Engl J Med. 2012;366:
914 –924.
25. Pivonello R, De Martino MC, Cappabianca P, et al. The medical
treatment of Cushing’s disease: effectiveness of chronic treatment
with the dopamine agonist cabergoline in patients unsuccessfully
treated by surgery. J Clin Endocrinol Metab. 2009;94:223–230.
26. Steen RE, Kapelrud H, Haug E, Frey H. In vivo and in vitro inhi-
bition by ketoconazole of ACTH secretion from a human thymic
carcinoid tumour. Acta Endocrinol (Copenh). 1991;125:331–334.
27. Terzolo M, Panarelli M, Piovesan A, Torta M, Paccotti P, Angeli A.
Ketoconazole treatment in Cushing’s disease. Effect on the circadian
profile of plasma ACTH and cortisol. J Endocrinol Invest. 1988;
11:717–721.
28. Boscaro M, Sonino N, Rampazzo A, Menegus AM, Mantero F.
ACTH response to corticotropin releasing hormone in Cushing’s
disease before and after ketoconazole: in vivo and in vitro studies.
Horm Metab Res Suppl. 1987;16:60 – 62.
29. Verhelst JA, Trainer PJ, Howlett TA, et al. Short and long-term
responses to metyrapone in the medical management of 91 patients
1630 Castinetti et al Ketoconazole in Cushing’s Disease
with Cushing’s syndrome. Clin Endocrinol (Oxf). 1991;35:169 –
178.
30. Valassi E, Crespo I, Gich I, Rodriguez J, Webb SM. A reappraisal of
the medical therapy with steroidogenesis inhibitors in Cushing’s
syndrome. Clin Endocrinol (Oxf). 2012;77:735–742.
31. van den Bosch OF, Stades AM, Zelissen PM. Increased long-term
remission after adequate medical cortisol suppression therapy as
presurgical treatment in Cushing’s disease. Clin Endocrinol (Oxf).
2014;80:184 –190.
32. Bercoff E, Bernuau J, Degott C, et al. Ketoconazole-induced fulmi-
nant hepatitis. Gut. 1985;26:636 – 638.
33. Chien RN, Yang LJ, Lin PY, Liaw YF. Hepatic injury during keto-
conazole therapy in patients with onychomycosis: a controlled co-
hort study. Hepatology. 1997;25:103–107.
34. Duarte PA, Chow CC, Simmons F, Ruskin J. Fatal hepatitis asso-
ciated with ketoconazole therapy. Arch Intern Med. 1984;144:
1069 –1070.
You can post your CV, post an open position or
look for your next career opportunity at EndoCareers.
www.endocareers.org
J Clin Endocrinol Metab, May 2014, 99(5):1623–1630
35. Lake-Bakaar G, Scheuer PJ, Sherlock S. Hepatic reactions associated
with ketoconazole in the United Kingdom. Br Med J (Clin Res Ed).
1987;294:419 – 422.
36. Lewis JH, Zimmerman HJ, Benson GD, Ishak KG. Hepatic injury
associated with ketoconazole therapy. Analysis of 33 cases. Gas-
troenterology. 1984;86:503–513.
37. Meiring PD, Whitelaw DA. Ketoconazole induced hepatitis. S Afr
Downloaded from https://academic.oup.com/jcem/article/99/5/1623/2537349 by Carol Davila Univ of Medicine and Pharmacy user on 13 March 2024
Med J. 1987;72:577–578.
38. Okumura H, Aramaki T, Satomura K, et al. Severe hepatitis during
ketoconazole therapy. Gastroenterol Jpn. 1983;18:142–147.
39. Svedhem A. Toxic hepatitis following ketoconazole treatment.
Scand J Infect Dis. 1984;16:123–125.
40. Tkach JR, Rinaldi MG. Severe hepatitis associated with ketocona-
zole therapy for chronic mucocutaneous candidiasis. Cutis. 1982;
29:482– 484.
41. Khosla S, Wolfson JS, Demerjian Z, Godine JE. Adrenal crisis in the
setting of high-dose ketoconazole therapy. Arch Intern Med.
1989;149:802– 804.