Patient characteristics

Overall, 77.4% of patients enrolled in the study were female. Mean baseline mUFC was 7.3 × ULN (Table 1). All patients had at least one relevant medical history/current medical condition, which included hypertension, diabetes and dyslipidaemia (Table 1). At baseline, the most frequently reported physical manifestation rated as severe was central obesity (16.1% mild, 29.9% moderate, 25.5% severe). Additional physical manifestations were rated mostly as mild or moderate: supraclavicular fat pad (24.8% mild, 35.0% moderate, 8.8% severe); dorsal fat pad (32.1% mild, 34.3% moderate, 7.3% severe); facial rubor (34.3% mild, 23.4% moderate, 5.8% severe); hirsutism (females; 37.7% mild, 17.0% moderate, 3.8% severe); proximal muscle atrophy (30.7% mild, 15.3% moderate, 5.8% severe); striae (24.1% mild, 18.2% moderate, 6.6% severe); ecchymoses (22.6% mild, 11.7% moderate, 4.4% severe).

Table 1 Baseline clinical characteristicsEffect of osilodrostat on mUFC

Median duration of osilodrostat exposure from baseline to core study data cut-off was 75 weeks (range 1–165, IQR 48–117), and average median daily osilodrostat dose was 7.1 mg/day (range 1.1–53.9, IQR 3.8–14.0). The majority (71.5%) of patients achieved control of mUFC (≤ ULN) by week 12, with a further 13.9% achieving partial control. The proportion of patients with mUFC ≤ ULN remained high throughout the core study (Fig. 1). Overall, 132 (96.4%) patients achieved control of mUFC at least once during the 48 weeks. Descriptive analyses of the proportion of patients with mUFC control by age, sex, race and time since diagnosis suggest that these factors do not affect response to treatment at weeks 24 and 48 (Fig. 2; Supplementary Table 1).

Fig. 1

Proportion of patients with controlled, partially controlled and uncontrolled mUFC over time. Based on patients in the full analysis set (N = 137). Patients who had discontinued or otherwise had missing mUFC values at a given visit were counted as uncontrolled. Controlled, mUFC ≤ ULN; partially controlled, mUFC > ULN but ≥ 50% reduction from baseline; uncontrolled, mUFC > ULN and < 50% reduction from baseline

Fig. 2

Overview of mUFC control by age, sex, race and time since diagnosis at A) 24 and B) 48 weeks. mUFC control defined as mUFC ≤ ULN; shaded area indicates the ‘all patients’ data

Changes in blood pressure

In total, 86.9% of patients were classified as hypertensive at baseline (Table 1). Of patients with baseline SBP > 130 mmHg (n = 79), 58.2, 50.6 and 49.4% had SBP ≤ 130 mmHg after 12, 24 and 48 weeks of osilodrostat treatment, respectively; of patients with baseline DBP > 90 mmHg (n = 50), 72.0, 62.0 and 66.0% had DBP ≤ 90 mmHg at weeks 12, 24 and 48, respectively. In patients without hypertension at baseline, both SBP and DBP generally remained stable during osilodrostat treatment. At week 48, five patients with baseline SBP ≤ 130 mmHg had SBP > 130 mmHg, and four patients with baseline DBP ≤ 90 mmHg had DBP > 90 mmHg. Of all patients classified as hypertensive at baseline, at week 48, mean (95% CI) SBP and DBP improved both in patients who did and in those who did not receive antihypertensive medications during the study (SBP − 8.7 [− 12.6, − 4.8] and − 14.5 [− 20.1, − 9.0] mmHg, respectively; DBP − 5.1 [− 7.7, − 2.5] and − 10.2 [− 14.5, − 6.0] mmHg, respectively). Overall, at week 48, 40% (n = 34/85) of patients taking antihypertensive medication at baseline had either stopped or reduced the dose; 40% (n = 34/85) had an increase in dose of antihypertensive medication or number of medications. Of patients with controlled, partially controlled and uncontrolled mUFC at week 48, 43.1% (n = 25/58), 33.3% (n = 3/9) and 33.3% (n = 6/18), respectively, had either stopped or reduced the dose of antihypertensive medication; 39.7% (n = 23/58), 44.5% (n = 4/9) and 38.9% (n = 7/18), respectively, had an increase in dose or number of medications, and 17.2% (n = 10/58), 22.2% (n = 2/9) and 27.8% (n = 5/18), respectively, had no change. Clinically relevant improvements from baseline in mean (95% CI) SBP and DBP were noted by week 24 in patients with controlled (− 7.1 [− 10.3, − 3.9] and − 4.9 [− 7.1, − 2.7] mmHg, respectively) and partially controlled mUFC (− 7.2 [− 14.0, − 0.4] and − 5.2 [− 9.8, − 0.6] mmHg), but not in patients with uncontrolled mUFC (2.0 [− 9.7, 13.7] and 5.3 [− 3.3, 13.9] mmHg). By week 48, improvements in both SBP and DBP were seen irrespective of mUFC control (Fig. 3A). There was no correlation between change from baseline in mUFC and change from baseline in SBP or DBP at week 24 (r = − 0.02, P = 0.8326 and r = 0.02, P = 0.8160, respectively) and a weak correlation at week 48 (r = 0.20, P = 0.0433 and r = 0.18, P = 0.0715, respectively).

Fig. 3

Mean (95% CI) change from baseline to week 48 in A) clinical signs, B) lipid profile, and C) glycaemic parameters by degree of mUFC control at week 48

Changes in body weight, waist circumference and BMI

Most patients had physical evidence of central obesity at baseline; mean weight was 80.8 kg and mean BMI was 30.3 kg/m2 (Table 1). Clinically relevant improvements in mean (95% CI) weight, waist circumference and BMI were noted by week 24 in patients with controlled mUFC (− 2.7 [− 3.6, − 1.8] kg, − 3.4 [− 4.9, − 1.8] cm and − 1.1 [− 1.4, − 0.7] kg/m2, respectively); improvements were not clinically relevant in patients with partially controlled mUFC (− 1.7 [− 3.8, 0.3] kg, − 2.4 [− 5.0, 0.3] cm and − 0.6 [− 1.4, 0.2] kg/m2) or uncontrolled mUFC (− 1.6 [− 3.5, 0.3] kg, 0.1 [− 2.4, 2.6] cm and − 0.6 [− 1.4, 0.1] kg/m2). By week 48, improvements from baseline were seen irrespective of mUFC control (Fig. 3A). No correlation was observed between change in mUFC and change in mean body weight, waist circumference or BMI at either week 24 or week 48 (Supplementary Table 2).

Changes in lipid profile

In total, 83.9% of patients had dyslipidaemia at baseline (Table 1). Mean changes in cholesterol (including LDL-c and HDL-c) and triglycerides were similar across mUFC response subgroups at both week 24 and week 48 (Fig. 3B). There were no strong correlations between change in mUFC and change in total cholesterol, LDL-c, HDL-c or triglycerides at either week 24 or week 48 (Supplementary Table 2).

Changes in glycaemic status

At baseline, 61 (44.5%) patients were classified as diabetic. Based on FPG levels only, most (67.9%) patients had normoglycaemia (FPG < 100 mg/dL) at baseline; 26.3% of patients had either impaired fasting glucose or diabetes (FPG ≥ 100 mg/dL). Of patients with baseline FPG ≥ 100 mg/dL (n = 36), 58.3, 63.8 and 44.4% had FPG < 100 mg/dL by weeks 12, 24 and 48, respectively. In total, at week 48, 48.8% (n = 21/43) of patients taking antidiabetic medication at baseline had stopped or reduced the dose, and 23.3% (n = 10/43) increased the dose or number of medications. Of patients with controlled, partially controlled and uncontrolled mUFC at week 48, 55.6% (n = 15/27), 75% (n = 3/4) and 25% (n = 3/12), respectively, had either stopped or reduced the dose of antidiabetic medication; 18.5% (n = 5/27), 25% (n = 1/4) and 33.3% (n = 4/12), respectively, had an increase in dose or number of medications, and 25.9% (n = 7/27), 0% (n = 0/4) and 41.7% (n = 5/12), respectively, had no change. Improvements in mean (95% CI) FPG and HbA1c at week 24 occurred irrespective of mUFC control, although they were more evident in patients with controlled mUFC (− 13.3 [− 18.9, − 7.7] mg/dL and − 0.3% [− 0.5, − 0.2], respectively) or partially controlled mUFC (− 21.2 [− 37.0, − 5.4] mg/dL and − 0.4% [− 0.9, 0.0]) than in those with uncontrolled mUFC (− 7.4 [− 17.1, 2.4] mg/dL and − 0.0% [− 0.3, 0.3]). A similar pattern of mean changes in FPG and HbA1c was observed at week 48 (Fig. 3C). There was a weak correlation between change from baseline in mUFC and change from baseline in FPG (r = 0.25, P = 0.008) and HbA1c (r = 0.23, P = 0.012) at week 24. At week 48, the correlation was stronger for FPG (r = 0.33, P = 0.001) but absent for HbA1c (r = 0.14, P = 0.161).

Changes in BMD

Mean baseline L1‒L4 lumbar spine and total hip BMD in all patients was 1.0 and 0.9 g/cm2, respectively. Mean change from baseline to week 48 in BMD at both the lumbar spine and total hip was 0.0 g/cm2 in both male and female patients; as a mean percentage change from baseline (95% CI), the increase in BMD was more pronounced in males (lumbar spine: + 3.9% [− 0.2, 8.0]; total hip: + 1.7% [− 0.5, 4.0]) than females (lumbar spine: + 2.7% [1.3, 4.1]; total hip: − 0.1% [− 1.6, 1.3]). The degree of mUFC response had no effect on outcomes (Supplementary Table 3).

Changes in physical manifestations of hypercortisolaemia

Physical manifestations of hypercortisolaemia were prevalent at baseline. At weeks 24 and 48, improvements in rated severity score from baseline occurred across all physical manifestations, with few patients rated with worsening scores. Improvements in physical manifestation severity scores were seen irrespective of the degree of mUFC response, with the exception of proximal muscle atrophy at week 24, whereby a numerically larger proportion of patients with controlled (26.5%) and partially controlled (43.8%) mUFC had improved severity scores compared with uncontrolled patients (14.3%; Fig. 4).

Fig. 4

Proportion of patients with improvements in rated severity score from baseline for physical manifestations of hypercortisolism by degree of mUFC control at week 24 and week 48. An improvement was defined as the symptom score being lower (ie less severe) than at baseline. The denominator for the percentage is the number of patients in the full analysis set (all enrolled patients who received at least one dose of osilodrostat) with data available at both baseline and the given visit. *Females only

Changes in HRQoL

Clinically meaningful improvements in mean CushingQoL scores for the overall study population were reached at weeks 26, 30, 32, 34 and 48, and for mean change in BDI-II scores at weeks 24, 26, 28, 30 and 48 [12, 14, 15]. At week 24, numerical improvements in both mean (95% CI) CushingQoL and BDI-II scores occurred irrespective of whether patients had controlled mUFC (9.2 [6.1, 12.4] and − 3.6 [− 5.4, − 1.9], respectively), partially controlled mUFC (9.4 [0.9, 17.8] and − 6.8 [− 11.2, − 2.4]) or uncontrolled mUFC (10.1 [2.0, 18.2] and − 4.6 [− 11.2, 2.0]). A similar finding was observed at week 48, although the 95% CIs were notably wider for patients with uncontrolled mUFC (Fig. 5). There was a weak correlation between change in mUFC and change in CushingQoL score at week 24 (r = − 0.15, P = 0.0875), which was not evident at week 48 (r = 0.02, P = 0.8028). A weak correlation between change in mUFC and change in BDI-II score was observed at both week 24 and week 48 (r = 0.24, P = 0.0081 and r = 0.30, P = 0.0016, respectively).

Fig. 5

Mean (95% CI) change from baseline to week 48 in A) CushingQoL total score and B) BDI-II score by degree of mUFC control at week 48. CushingQoL scores range from 12 (worst) to 60 (best), and BDI-II scores range from 0 (best) to 63 (worst)