Osilodrostat, a potent oral 11β-hydroxylase inhibitor: 22-week, prospective, Phase II study in Cushing’s disease

Cushing’s disease is caused by an adrenocorticotropic hormone (ACTH)-secreting pituitary tumor and is the most common cause of excess endogenous cortisol secretion [1‐3]. Hypercortisolism can lead to substantial morbidity and premature death compared with the general population [4]. The primary treatment goals for Cushing’s disease are to normalize cortisol levels and reverse the signs and symptoms of hypercortisolism [2, 3]. First-line treatment is transsphenoidal surgery [2], although this is not always successful [5] and patients may relapse many years after apparent surgical success [6]. A number of medical therapies are currently used in clinical practice for the treatment of Cushing’s disease. These include pasireotide (multireceptor-targeted somatostatin analogue), cabergoline (dopamine receptor agonist), metyrapone and ketoconazole (adrenal steroidogenesis inhibitors), mitotane (adrenolytic agent) and mifepristone (glucocorticoid receptor antagonist) [3, 5, 7‐18]. Since not all patients with Cushing’s disease achieve sufficient benefit with available therapies, there is a continuing need for new medical therapies.

Osilodrostat (LCI699) is an oral inhibitor of 11β-hydroxylase, which catalyzes the final step of cortisol synthesis. Although this mechanism of action is similar to that of metyrapone, osilodrostat has a longer plasma half-life (4–5 versus ~2 h), allowing twice-daily dosing (instead of 3–4 times daily), and is more potent against 11β-hydroxylase (in vitro IC₅₀ of 2.5 versus ~7.5 nM for metyrapone), allowing for an overall more convenient dosing schedule. A proof-of-concept study (LINC 1; LCI IN Cushing’s) demonstrated that osilodrostat normalized urinary free cortisol (UFC) levels in 11/12 patients with Cushing’s disease after 10 weeks [19]. The current Phase II study (LINC 2; clinicaltrials.gov NCT01331239) further assessed osilodrostat in patients with Cushing’s disease over a longer period of time and in more patients.

The LINC 2 study demonstrated that osilodrostat, a potent oral 11β-hydroxylase inhibitor, decreased UFC levels in all patients with Cushing’s disease and maintained normal UFC in 79 % of patients (n/N = 15/19) after 22 weeks of treatment. Of the remaining four patients, two had ~50 % decreases in UFC—though it is important to note that normalization of cortisol levels is the goal of treatment—and two discontinued the study (one because of an AE, the other as a result of an administrative issue). Two patients were responders at week 10, but not at week 22; there is a possibility that one of these patients ‘escaped’ from response (see Supplementary Appendix). Notably, all enrolled patients had normalized UFC levels at least once during the study. Overall, these data confirm, over a longer follow-up period, the results from the LINC 1 study in which 11/12 patients (92 %) had normalized UFC after 10 weeks [19]. Taken together, these data demonstrate that osilodrostat effectively controls UFC in patients with Cushing’s disease. Changes in late-night salivary cortisol levels were variable and remained above normal throughout treatment; at this time, we do not know whether the exact timing of the evening osilodrostat dose might have impacted on the changes in late-night salivary cortisol.

Osilodrostat treatment led to expected increases above normal in 11-deoxycortisol and 11-deoxycorticosterone; these increases are in line with observations in LINC 1 [19]. It is worth noting that in LINC 1, levels of 11-deoxycortisol and 11-deoxycorticosterone decreased towards baseline following 14 days of osilodrostat washout; there was no such washout period in LINC 2. ACTH levels also increased in LINC 2, possibly as a compensatory reaction to the reduction in serum cortisol levels. The large increase in ACTH levels in the expansion cohort was primarily driven by two patients who had dramatic increases. Renin levels changed in opposite directions in the two cohorts. This is difficult to explain, but it may be a chance finding related to the small patient numbers, particularly in the follow-up cohort, and the large variability in the observed data. In addition, an effect of concomitant medications (e.g. diuretics such as spironolactone or eplerenone) on renin levels cannot be excluded.

Mean body weight was relatively unchanged throughout osilodrostat treatment in LINC 2; a possible effect of the mineralocorticoid precursors on body weight through fluid retention cannot be excluded. By contrast, in LINC 1, there was an increase in mean weight of 3.5 kg in the overall population [19]. Although the change in LINC 1 was primarily due to a single patient who experienced a 19 kg increase, there remained an overall mean increase of 2.4 kg when this patient was not included. There was no change in blood pressure in the current study, whereas there was a trend towards an improvement in LINC 1. However, baseline blood pressure levels in LINC 2 were generally lower than in LINC 1, which may explain the difference between the studies. Another possibility is that the increase in mineralocorticoid precursors may have offset the blood-pressure-lowering effect of inhibiting aldosterone synthesis (see Fig. 4). In the overall study population, there were modest improvements in glucose and HbA_1c levels during osilodrostat treatment; the improvements were more substantial in patients with a history of diabetes mellitus at baseline. There were also modest decreases in cholesterol (total, HDL, and LDL) and triglyceride levels; the reduction in HDL-cholesterol is an unfavorable effect and, in women, may be related to increased testosterone levels [26]. The possibility that the effects of concomitant medications (e.g. for hypertension, diabetes, or dyslipidemia) may have impacted changes in these clinical parameters cannot be excluded. An analysis of the relationship between such medications and associated clinical outcomes is planned in future studies. There were no significant changes in mean potassium levels, although nine patients developed mild hypokalemia and one patient developed borderline hyperkalemia (2 weeks after initiating osilodrostat therapy) based on laboratory assessments. Only one case of hypokalemia was reported as an AE. The mild abnormalities in serum potassium levels were transient and managed effectively with potassium supplements or resolved without intervention. In the overall population, mean baseline levels of fasting plasma glucose, HbA_1c, cholesterol and triglycerides were within the normal range. Therefore, it is perhaps not surprising that no substantial changes were observed during osilodrostat treatment, despite the improvements in UFC levels; longer-term studies evaluating the effect of osilodrostat on changes in clinical and laboratory parameters in a larger number of patients are required. In addition, patients who had diabetes mellitus and dyslipidemia at baseline appeared to be well controlled, since they also had fasting plasma glucose/HbA_1c and cholesterol/triglyceride levels, respectively, within the normal range at week 22.

Osilodrostat treatment was generally well tolerated; AEs were consistent with those observed in LINC 1. Only one patient discontinued because of an AE, and 16/17 patients who completed LINC 2 elected to continue in the optional extension, which is ongoing; this suggests a high level of patient acceptance of the medication in this small patient population. Adrenal insufficiency was reported as an AE in six patients; the symptoms may alternatively have been a result of glucocorticoid withdrawal syndrome in some patients. The observation of adrenal insufficiency and/or steroid withdrawal in approximately one-third of patients highlights the potency of osilodrostat. It may therefore be prudent in future studies to titrate the dose of osilodrostat more slowly than was done in this dose-escalation study, and to target UFC levels in the mid-normal range.

Serial pituitary imaging was performed to screen for the theoretical risk of corticotroph tumor progression, analogous to Nelson’s syndrome in patients with bilateral adrenalectomy [21]. Although there was minimal change (<2.0 mm in maximal diameter) in tumor diameter during the period of observation in this study in the six patients with measurable tumors at baseline and week 22, it is important to note that longer follow-up is needed to explore more fully changes in tumor volume during osilodrostat treatment.

One potential effect of increased testosterone levels is hirsutism in women. No hirsutism was reported in LINC 1 despite significantly increased testosterone levels, although the authors speculated that this was because the short study duration (10 weeks) may have been insufficient to observe such effects [19]. In this 22-week study, testosterone levels increased in the female population, and new or worsening hirsutism (n = 2) and/or acne (n = 3) were reported among four female patients, all of whom had testosterone levels >ULN at week 22. Most increases in testosterone in female patients were only moderate, except for one patient who had a >10-fold increase (to ~16 nmol/L). Interestingly, this patient was not one of the four to report acne or hirsutism as an AE. Testosterone levels in male patients increased from sub- or low-normal levels to the mid-normal range, suggesting that osilodrostat may have a potential therapeutic effect. In general, data on gonadotroph function and possible clinical effects of androgen increases are limited in this Phase II study, but may be evaluated in greater detail in future studies of osilodrostat.

It should be noted that these data are somewhat limited by the fact that LINC 2 is an expanded, open-label, uncontrolled study, conducted in a small number of patients, some of whom (4/19) had previously received osilodrostat. A confirmatory Phase III study (LINC 3) is ongoing to evaluate the effect of osilodrostat in a larger patient population. This and other future studies will determine the place of osilodrostat in the medical treatment of Cushing’s disease.

In conclusion, osilodrostat demonstrated good efficacy with a satisfactory safety profile in this Phase II study, showing promise for the future treatment of patients with Cushing’s disease.