Medical management of Cushing’s disease: what is the future?
- First Online:
- Cite this article as:
- Fleseriu, M. & Petersenn, S. Pituitary (2012) 15: 330. doi:10.1007/s11102-012-0397-5
- 1.7k Downloads
Cushing’s disease (CD) is caused by a corticotroph, adrenocorticotropic-hormone (ACTH)—secreting pituitary adenoma resulting in significant morbidity and mortality. Transsphenoidal surgery is the initial treatment of choice in almost all cases. Remission rates for microadenomas are good at 65–90 % (with an experienced neurosurgeon) but remission rates are much lower for macroadenomas. However, even after postoperative remission, recurrence rates are high and can be seen up to decades after an initial diagnosis. Repeat surgery or radiation can be useful in these cases, although both have clear limitations with respect to efficacy and/or side effects. Hence, there is a clear unmet need for an effective medical treatment. Currently, most drugs act by inhibiting steroidogenesis in the adrenal glands. Most is known about the effects of ketoconazole and metyrapone. While effective, access to ketoconazole and metyrapone is limited in many countries, experience with long-term use is limited, and side effects can be significant. Recent studies have suggested a role for a pituitary-directed therapy with new multireceptor ligand somatostatin analogs (e.g., pasireotide, recently approved in Europe for treatment of CD), second-generation dopamine agonists, or a combination of both. Mifepristone (a glucocorticoid receptor antagonist) is another promising drug, recently approved by the FDA for treatment of hyperglycemia associated with Cushing’s syndrome. We review available medical treatments for CD with a focus on the two most recent compounds referenced above. Our aim is to expand awareness of current research, and the possibilities afforded by available medical treatments for this mesmerizing, but often frightful disease.
KeywordsCushing’s diseaseCushing’s syndromePasireotideMifepristoneSomatostatin analogPituitary adenoma
Cushing’s disease (CD) is considered an aggressive pituitary endocrine disorder because of the devastating long-term consequences of untreated hypercortisolemia. Transsphenoidal surgery is clearly the first line therapy of choice; with remission rates of between 65 and 90 % for microadenomas, but somewhat lower rates <65 % for macroadenomas . This variability relates to the size and location of tumors, neurosurgical expertise, length of follow-up, and the definition of remission. Indeed, this last point is a controversial aspect in the management of CD, as there is no consensus on the most accurate postoperative assessment of remission and its predictive value for long-term prognosis . Furthermore, despite initial cure after surgery, up to a quarter of patients will experience recurrence within 10 years . Due to the significance of morbidity and mortality if the disease is not cured , patients require additional effective treatment if initial surgery does fail. A second surgery can be a desirable choice only in selected cases [5, 6]. Radiotherapy can also represent an alternative in patients with large remnant or recurrent disease, but can take several months to years to reach biochemical normalization and has inherent adverse effects beyond the risks of panhypopituitarism .
Therefore, there is an urgent need for a medical therapy that can effectively reverse clinical features of CD by normalizing CD-associated biochemical parameters acutely—and more importantly during long-term treatment. Although minimally invasive laparoscopic bilateral adrenalectomy is an effective and rapid-onset alternative performed with few short-term complications , corticotroph adenoma progression is seen in about 30 % of cases, and may eventually develop to Nelson’s syndrome . Life-long dependence on replacement glucocorticoids and mineralocorticoids may itself be associated with significant morbidity due to recurrent adrenal crises.
Interpretation of available data on the efficacy and safety of most drugs currently used in the treatment of CD is difficult. A review of published reports indicates that study design has varied considerably with a few small prospective, controlled, randomized studies available. Furthermore, it is unclear whether some of those studies could potentially represent publication bias. There is also significant variation in biochemical parameters used for the primary endpoint (e.g., urine free cortisol [UFC], serum and salivary cortisol, and plasma ACTH) and, with few exceptions [13, 14], reference values derived from a sufficiently large population are largely lacking, especially for some of the more recently developed assays. Unfortunately, the criteria for defining a clear and effective response to treatment, and for disease control, are insufficient .
The aim of pituitary-targeted therapy is to directly act on the underlying pituitary tumor/ACTH hypersecretion source. Although, for many years, various drugs have been tried with limited efficacy, recent approaches may eventually allow for medical treatment of selected patients with CD. Currently, three classes of drugs are under investigation in humans: (1) somatostatin analogs, (2) dopamine agonists, and (3) PPARγ ligands.
Expression of somatostatin receptors in corticotropic pituitary adenomas
Author’s data (SP)
25/39 (64 %)
29/39 (74 %)
8/38 (21 %)
12/37 (32 %)
31/37 (84 %)
In a recent case report, a dramatic drop of elevated UFC levels in a patient treated with pasireotide 600 μg bid was observed, leading to symptomatic and biochemical hypocortisolism . Prior to pasireotide, the patient demonstrated UFC levels approximately 1.5-2-fold above normal, despite three surgical attempts and radiotherapy. After 10 months of reduced pasireotide therapy (300 μg bid) the patient continued to demonstrate improved mood, stable weight and normal UFC levels. Bode et al.  reported a female patient with a giant adenoma that was initially treated by transcranial resection. Due to a symptomatic relapse and tumor recurrence, the patient received stereotactic radiosurgery, followed by bilateral adrenalectomy. Subsequently, the patient was diagnosed with a pituitary carcinoma due to the development of several intracranial and spinal metastases. She was started on temozolomide, with another dramatic rise in ACTH levels (from 882 to 3,868 ng/mL) early after initiation of therapy. As a salvage therapy, she received pasireotide 900 μg subcutaneously bid. ACTH levels decreased to 1,738 ng/mL after 2 weeks. ACTH levels than remained at 900–1,000 ng/mL, and her clinical status improved. The combination of temozolomide and pasireotide led to sustained tumor control for 12 months, after which temozolomide was stopped. The patient then remained stable on monotherapy with pasireotide for more than nine months. The authors themselves state that they cannot dissect the efficacy of both therapies. However, the combination of temozolomide and pasireotide may be a promising option for the treatment of ACTH-secreting pituitary carcinomas.
Dopamine subtype 2 receptors are expressed in corticotroph pituitary adenomas . Bromocriptine has been shown to inhibit ACTH secretion  and apoptosis in an ACTH-secreting mouse cell line . In a combined review of 23 patients with CD treated with 1.25–30 mg/day of bromocriptine for 3–180 weeks, a response rate of 48 %  was observed. Of note, primary endpoints for individual studies have varied (e.g., plasma cortisol levels, cortisol secretion rates, and 24-h urinary glucocorticoid assessments). Subsequent series have demonstrated significantly lower response rates (4–23 %) [43, 44]. Newer dopamine agonists may be more effective. In a single-center study of ten patients treated with cabergoline (1–3 mg/week for 3 months), a normalization rate of 40 % for UFC was observed . Moreover, the same authors confirmed the long-term efficacy of cabergoline in CD in a subsequent study, in which normalized UFC was maintained in 40 % of 20 patients for at least 12 months of follow-up . A dose of 1–7 mg/week (median 3.5 mg) was well tolerated, except by two patients with hypotension and severe asthenia, who stopped treatment after 12 and 18 months, respectively. This data was recently confirmed by Godbout et al. , who treated 30 CD patients with cabergoline, up to 6 mg/week. Thirty percent of patients demonstrated long-term normalization of UFC for 12–60 months, with a mean dose of 2.1 mg/week. Whether or not recently documented cardiac valve changes in patients with Parkinson’s disease receiving much high doses of cabergoline [48, 49] equate to the lower doses applied in this cohort is not clear. Of note, Pivonello et al.  documented no development of cardiac valve insufficiency and one case of worsening of a previously diagnosed tricuspid regurgitation. The results of this single-center experience need to be confirmed in a larger multi-center study that examines both efficacy and safety.
PPARγ (peroxisome proliferator-activated receptor-gamma)
PPARγ (peroxisome proliferator-activated receptor-gamma) is a member of the nuclear receptor superfamily, and functions as a transcription factor. PPARγ ligands have been shown to inhibit the growth of many tumors, including breast, colon, and prostate cancer cells . Furthermore, PPARγ ligands have demonstrated antiproliferative and apoptotic effects in a murine cell model of corticotropic adenoma cells, and inhibited proopiomelanocortin (POMC) transcription . However, very little PPARγ protein was found in pituitary adenomas. Additionally, the antiproliferative effects of PPARγ ligands were only observed with very high doses of rosiglitazone, without significant reversal by a PPARγ antagonist . In CD, suppression of steroidogenesis in the adrenal glands via inhibition of P450c17 and 3ßHSD could also be beneficial , however clinical experience is limited so far. In a study of rosiglitazone (8–16 mg) in 14 patients treated for 1–7 months, normalization of UFC was observed in 42.9 % of patients, with mild improvement in clinical features . A subsequent study in ten patients treated for 1–8 months with 4–16 mg rosiglitazone demonstrated normalization of UFC in 30 % of patients . Side effects, however, were significant and included edema, hypertension, weight gain, somnolence, increased hirsutism, and bruisability. Thus, patients were unwilling to re-enter the treatment protocol . Morcos et al. , in a report on long-term treatment of central Cushing’s syndrome with rosiglitazone observed an initial decrease in ACTH, but with a rebound after 28 weeks of therapy, despite dose increases of up to 28 mg. Suri and Weiss published a report on the effect of pioglitazone on ACTH and cortisol secretion in CD. They showed that pioglitazone at 45 mg for one month had no noteworthy effects on either ACTH or cortisol secretion. This study was, however, undertaken in a small study cohort . At a consensus conference, in the investigator panel’s expert opinion, it was concluded that current study results did not sufficiently supported routine clinical use of PPARγ ligands to treat CD . Additionally, there are concerns related to PPARγ ligands and increased cardiovascular disease risks , which means that PPARγ ligands are not the ideal drug of choice for patients with Cushing’s who are already at high risk of heart problems.
Glucocorticoid receptor antagonist therapy
Use of a glucocorticoid receptor blocker represents another medical therapy approach to treat CD . Hypercortisolemia persists, but the devastating effects are ameliorated by the antagonist binding to the glucocorticoid receptor—a concept that is similar to the use of pegvisomant when treating acromegaly . Mifepristone is currently the only available glucocorticoid receptor antagonist [61–63]. Mifepristone directly blocks the cortisol glucocorticoid receptor (GR-II) and the progesterone receptor (PR). There are 51 published case reports detailing the use of mifepristone in treating hypercortisolism (five in CD patients) [64, 65]. Chu et al.  presented a case report on a patient with an ACTH-secreting pituitary macroadenoma who had failed multiple therapies and was too ill to undergo bilateral adrenalectomy. A dramatic improvement in clinical symptoms was reported with mifepristone (starting dose 6 mg/kg/day, up to a maximum of 25 mg/kg/day) over 18 months. Severe hypokalemia (which was responsive to spironolactone administration) was attributed to excessive cortisol activation by the mineralocorticoid receptor. In a retrospective report (from seven European centers) on the use of mifepristone in patients with hypercortisolism, four additional patients with CD were presented . Clinical symptoms improved in three during treatment over 3–24 months. The fourth patient was treated for 0.5 months, during which rapid improvements in psychiatric symptoms were demonstrated. The patient later underwent surgery. Severe hypertension and hypokalemia developed in one patient.
The lack of a biochemical marker correlating with the treatment response dictates that efficacy measurements should be based on symptoms, clinical features, and metabolic improvements. Of note, direct translation of clinical study data to day-to-day use in a busy clinical practice can be challenging. Monitoring side effects like hypokalemia and hypertension, as well as early recognition of clinical AI, is essential. Long-term effects of mifepristone on tumor size and endometrial effects are as yet undetermined, and further long-term studies are ongoing . The development of a selective glucocorticoid receptor without antiprogestin effects  could also represent an important step in the long-term treatment of women with CD.
Adrenal-directed therapy: steroidogenesis inhibitors
Steroidogenic drugs have been particularly useful in treating patients with severe hypercortisolism who are waiting for surgery, after noncurative surgery, or as bridge until radiation therapy becomes effective.
Adrenal steroidogenesis inhibitors
Maximal dose in CD
200 mg bid
400 mg tid
250 mg qd
1,500 mg qid
500 mg tid
1,000 mg tid
Bolus of 0.03 μg/kg iv, followed by infusion 0.1 mg/kg/h
Metyrapone was studied in humans soon after the discovery of its steroidogenesis blocking effects in animals . It has been used as monotherapy or in combination. Metyrapone mainly acts via inhibition of 11-β-hydroxylase and to a lesser extent 17-α-, 18-, and 19-hydroxylase activities  and therefore, aldosterone biosynthesis is more severely affected than that of cortisol . Initial doses are usually 250 mg qid, which can be adjusted during treatment to a total daily dose of 500–6,000 mg . In patients with CD, a significant drop of cortisol levels is observed 2 h after a first dose . Short-term metyrapone therapy induces clinical improvements in the majority of patients, with biochemical control in 75 % (median dose of 2,250 mg/day; range 750–6,000 mg/day). Long-term use (median 27 months, range 3–140 months) induced even higher rates of biochemical remission (83 %), most likely due to an additive effect of pituitary irradiation. Jeffcoate et al. , reported biochemical control of 13 patients with CD for up to 66 months with metyrapone (nine had received pituitary irradiation in addition to metyrapone. ACTH secretion may override steroidogenic blockade in some patients with CD . However, as reported in one of the largest studies on metyrapone, this effect predominantly occurred in the first 4–6 weeks following initiation of therapy, after which no further increase was observed .
Adrenal insufficiency has been observed in 13 % of patients . Acne and hirsutism (19–70 %) due to increased androgens are relatively frequent, though mostly mild in nature , thus making metyrapone a clear second choice in women. As a result of increased 11-deoxycorticosterone levels, hypokalemia (6 %), edema (8 %) and hypertension are also observed . Other rare adverse effects included cutaneous rash (4 %), which may be transient, effects on the central nervous system (e.g., lethargy and dizziness, 15 %), and gastrointestinal complaints (e.g., nausea, 5 %). Metyrapone has limited availability in many countries (e.g., Germany, United States), but can be ordered through an international drugstore or by contacting the manufacturer (Novartis) directly.
Ketoconazole, initially licensed as an antifungal agent, was observed to lower cortisol  and testosterone  levels via inhibition of a variety of cytochrome P450 enzymes (side-chain cleavage complex, 17,20-lyase, 11-β-hydroxylase, and 17-α-hydroxylase) [80, 81]. Inhibition of ACTH secretion or action as a glucocorticoid receptor antagonist have been described in cell models, but there is a lack of clear evidence in humans [1, 42]. Ketoconazole treatment is usually started at 200 mg bid , and biochemical effect achieved at 600–800 mg/24 h . A meta-analysis of 12 studies treating 85 CD patients reported 81 % normalization of urinary steroids . A recent retrospective analysis found normalization of UFC in 45 % of patients (mean follow-up of 23 months; range 6–72 months) . Escape from pharmacological control has been reported . Rapid improvement in clinical symptoms, and regression of diabetes mellitus, hypokalemia, hypertension, hirsutism, and depression have been reported . Ketoconazole may reduce cholesterol  and vitamin D levels . Side effects are, in general, dose-dependent and include: gynecomastia (13 % of males), gastrointestinal symptoms (8 %), edema (6 %), and skin rash (2 %) . Ketoconazole represents a second-line in men due to the possible development of hypogonadism. Liver function should be monitored carefully, but mild and transient elevations in liver enzymes of up to threefold of normal are not a contraindication for further treatment. Typically, onset of ketoconazole-induced hepatitis is reported to occur within 60 days after initiation of treatment and resolves within 3 months of discontinuation . More severe abnormalities such as severe hepatic injury (seen in 1/15,000 cases)  require discontinuation .
The availability of ketoconazole is also limited in many countries. However, fluconazole appears to have similar effects , with normalized UFC levels (in the author’s [SP] limited experience with two patients treated with fluconazole 100 mg bid).
Mitotane inhibits steroidogenesis at the steps of side chain cleavage, 11- and 18-hydroxylase and 3-β-hydroxysteroid dehydrogenase. Its main use is in treatment of adrenocortical carcinoma, but has been proven effective in CD . In a study of 46 patients treated for 3–34 months at doses of 8–12 g/24 h (different formulation than currently available preparations, which require lower doses), 83 % achieved normalization of urinary markers . Sixty percent of patients relapsed after withdrawal of therapy, suggesting adrenolytic action is limited [90, 91]. Action is slow and saturation can be expected 2–3 months after initiation of therapy. Due to an accelerated metabolism of exogenous steroids, especially hydrocortisone, replacement doses must be increased to avoid adrenal crises [92, 93]. In contrast to its use in adrenal carcinoma, specific target ranges for drug concentrations in CD have not been established. Significant and frequent side effects include disturbances of the gastrointestinal tract (72 %) and the central nervous system (impaired mentation and dizziness, 45 %), as well as gynecomastia, rash, increases in liver enzymes, and hypercholesterolemia . Therefore, its use should be limited to centers with special experience. Recently, a defined preparation was licensed in Europe (Lysodren®, Bristol-Myers Squibb).
Aminoglutethimide inhibits the side-chain cleavage of cortisol biosynthesis and a variety of steroidogenic enzymes, e.g. side-chain cleavage complex, 21-hydroxylase, 17-α-hydroxylase, 11-β-hydroxylase, aromatase, 17–20 lyase, and 18-hydroxylase [42, 75]. Aminoglutethimide appears to be somehow more effective in patients with autonomous adrenal hyperfunction or ectopic ACTH production compared to patients with CD, possibly due to ACTH stimulation overriding the adrenal blockade in the latter group. As of this writing, the compound has only limited availability worldwide.
The non-opioid anesthetic etomidate induces adrenocortical suppression by dose-dependent inhibition of 11-β-hydroxylase and desmolase. It is the only such compound available for intravenous administration, thus being useful in situations where rapid control of hypercortisolism is required, or oral therapy is contraindicated. An initial bolus of 0.03 mg/kg is followed by an infusion of 0.03–0.3 mg/kg/h. Significant suppression of serum cortisol levels is observed quickly (after ~5 h), with a maximum effect at ~11 h [94–96].
LCI 699 is a potent inhibitor of 11-β-hydroxylase and 18-hydroxylase (aldosterone synthase) [97, 98] currently under investigation in a proof-of-concept study in patients with CD. Preliminary results  from 10 weeks treatment in 12 patients with mild to severe CD (UFC > 1.5xULN) showed UFC normalization in 11/12 patients, and all patients achieved the study’s primary endpoint (UFC ≤ ULN or ≥50 % reduction in UFC from baseline). Median LCI699 dose associated with UFC normalization was between 5 and 10 mg bid. LCI699 was generally well tolerated, and no serious drug-related adverse events were reported. The most common adverse events were fatigue (7/12), nausea (5/12) and headache (3/12). Four patients experienced study-drug–related hypokalemia (K+ < 3.5 mmol/L; min 3.1 mmol/L). Based on these promising results, evaluation of LCI699 in patients with CD is ongoing.
Interaction between somatostatin analogs and dopamine receptor agonists may allow for synergistic suppression of ACTH. Feelders et al.  studied 17 patients with pasireotide alone or with cabergoline and ketoconazole in CD (mean age, 45.7 years; 13 women) in an 80-day trial with normalization of urinary free cortisol (UFC) levels as the main outcome measure. Pasireotide monotherapy induced sustained normalization of UFC levels in five patients, and addition of cabergoline succeeded in normalizing an additional four patients. Despite a ≥ 50 % decrease in UFC, 8/17 patients (47 %) still had elevated UFC levels after two months on the combined treatment. The addition of low-dose ketoconazole increased the number of patients with a complete response to 88 % after a further two months.
Another study reported on the effectiveness of cabergoline (at doses of up 3 mg/week) combined with relatively low doses of ketoconazole (up to 400 mg/day) in patients with CD who were unsuccessfully treated by transsphenoidal surgery . Whereas treatment with cabergoline alone for six months allowed normalization of UFC in 3/12 patients, the addition of ketoconazole to the nine patients with insufficient response normalized UFC in an additional six patients.
Combination therapy with mitotane (3–5 g/24 h), metyrapone (3–4.5 g/24 h) and ketoconazole (400–1,200 mg/24 h) as an alternative to urgent adrenalectomy has been recently reported in 11 patients with severe CD . In all patients, high doses of these drugs were simultaneously started, and UFC decreased within 24–48 h. After this treatment, five patients were able to undergo surgery directed at ACTH excess. Side effects were tolerable and, interestingly, no greater than would be expected for each medication alone.
Cushing’s disease is a debilitating condition, with increased morbidity and mortality if biochemical remission/cure not achieved. Transsphenoidal surgery is the first-line treatment of choice, but does not result in a cure for many patients. Medical therapy represents a potentially useful addition to the range of available treatment options for many patients with CD for whom surgery has failed or is not possible. While adrenal steroidogenesis inhibitors are clearly effective, there have been few controlled trials, and little experience with their long-term use. Clinically important side effects have also limited the therapeutic potential of available therapies, and there is a possibility of publication bias for this class of therapies. Further, drugs of this type do not target the underlying pituitary tumor, and escape of ACTH secretion may require dose adjustments to maintain efficacy. Pituitary-targeted therapies may provide both an anti-secretory and an anti-proliferative treatment. Dopamine agonists have demonstrated some efficacy in small proof-of concept studies. A new somatostatin receptor ligand, pasireotide, has shown clear biochemical and clinical response. Similarly, treatment with the glucocorticoid receptor antagonist, mifepristone demonstrated clinical improvement. Our present knowledge of combination therapy is limited to case reports and a few small proof-of-concept studies; current data indicate, however, that a multimodal pharmacologic treatment approach may offer additive or synergistic clinical benefit with acceptable tolerability. Medical therapy for CD poses unique challenges and, unlike the biochemical control and tumor shrinkage achieved in patients with prolactinomas and acromegaly, comparable results have not yet been achieved in patients with CD. However, research is progressing rapidly, and we are getting closer to a more ideal treatment for this implacable disease.
The authors would like to thank Shirley McCartney, Ph.D. and David Wolff, M.S., for editorial assistance with this manuscript.
Conflict of interest
Dr. Fleseriu has received consultant fees from Novartis Pharmaceuticals and Ipsen, and is a principal investigator in clinical trials sponsored by Novartis Pharmaceuticals, Corcept Therapeutics, and Ipsen, with research support to OHSU. Dr. Petersenn has received consultant fees from Novartis Pharmaceuticals and Ipsen, and is a principal investigator in clinical trials sponsored by Novartis Pharmaceuticals and Ipsen. Partial financial support (David Wolff) for administrative assistance in the submission of this manuscript, and to allow open access, was provided by Novartis Pharmaceuticals Corporation. The manuscript was prepared by the authors and they are solely responsible for all content.
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.