A Phase 2, Multicenter Study of Nevanimibe for the Treatment of Congenital Adrenal Hyperplasia

Abstract

Context

Patients with classic congenital adrenal hyperplasia (CAH) often require supraphysiologic glucocorticoid doses to suppress adrenocorticotropic hormone (ACTH) and control androgen excess. Nevanimibe hydrochloride (ATR-101), which selectively inhibits adrenal cortex function, might reduce androgen excess independent of ACTH and thus allow for lower glucocorticoid dosing in CAH. 17-hydroxyprogesterone (17-OHP) and androstenedione are CAH biomarkers used to monitor androgen excess.

Objective

Evaluate the efficacy and safety of nevanimibe in subjects with uncontrolled classic CAH.

Design

This was a multicenter, single-blind, dose-titration study. CAH subjects with baseline 17-OHP ≥4× the upper limit of normal (ULN) received the lowest dose of nevanimibe for 2 weeks followed by a single-blind 2-week placebo washout. Nevanimibe was gradually titrated up if the primary outcome measure (17-OHP ≤2× ULN) was not met. A total of 5 nevanimibe dose levels were possible (125, 250, 500, 750, 1000 mg twice daily).

Results

The study enrolled 10 adults: 9 completed the study, and 1 discontinued early due to a related serious adverse event. At baseline, the mean age was 30.3 ± 13.8 years, and the maintenance glucocorticoid dose, expressed as hydrocortisone equivalents, was 24.7 ± 10.4 mg/day. Two subjects met the primary endpoint, and 5 others experienced 17-OHP decreases ranging from 27% to 72% during nevanimibe treatment. The most common side effects were gastrointestinal (30%). There were no dose-related trends in adverse events.

Conclusions

Nevanimibe decreased 17-OHP levels within 2 weeks of treatment. Larger studies of longer duration are needed to further evaluate its efficacy as add-on therapy for CAH.

Congenital adrenal hyperplasia (CAH) is a group of autosomal recessive disorders characterized by varying degrees of enzyme deficiencies affecting adrenal steroidogenesis. The most common form of CAH is 21-hydroxylase deficiency (21OHD), which results in androgen excess with or without cortisol (and mineralocorticoid) deficiency (1). Classic 21OHD is a severe form of CAH, where patients are dependent on lifelong replacement of mineralocorticoid and glucocorticoid. Patients with classic CAH often require supraphysiologic doses of glucocorticoids to overcome androgen excess by suppressing adrenocorticotropic hormone (ACTH) (2). Symptoms of androgen excess in patients with classic 21OHD range from hirsutism and acne to infertility (irregular menses in women, testicular adrenal rest tumors and gonadotropin suppression in men), and adrenal tumor formation in women and men. In the treatment of classic 21OHD, glucocorticoids must be carefully titrated to prevent hyperandrogenism, while avoiding overtreatment, with resultant iatrogenic Cushing syndrome—including weight gain, impaired glucose tolerance, adverse bone effects, and stunted growth in children (2-4). Thus, better treatment options for classic 21OHD are needed.

Nevanimibe HCl (ATR-101) is the first of a novel therapeutic molecule class that inhibits cholesterol homeostasis and function of the adrenal cortex. Nevanimibe potently inhibits acyl-coenzyme A:cholesterol O-acyltransferase 1 (ACAT1, or sterol O-acyltransferase 1 [SOAT1]), the principal enzyme that catalyzes the esterification of free cholesterol to cholesteryl esters for storage in adrenal cortex cells. At lower concentrations, nevanimibe reduces adrenal steroidogenesis across all 3 adrenocortical steroid pathways (i.e., mineralocorticoid, glucocorticoid, and androgens), while at higher concentrations, the increase in free cholesterol caused by ACAT1 inhibition results in dysregulation of calcium stores in the endoplasmic reticulum, the unfolded protein response, and ultimately apoptosis in adrenocortical cells (5). Nevanimibe thus presents a novel mechanism of action, with the unique potential for dose-dependent utility, ranging from inhibition of steroid synthesis to apoptosis in the treatment of adrenal diseases. Nevanimibe was shown to decrease ACTH-stimulated cortisol production in dogs with Cushing syndrome (6) and to induce apoptosis in adrenocortical cells in dogs and in vitro (5). Nevanimibe appears to also directly inhibit adrenal steroid production at doses lower than required to cause cellular toxicity (6), thereby potentially decreasing adrenal androgen production and reducing the need for supraphysiologic glucocorticoids in 21OHD. The goal of this study was to evaluate the safety and efficacy of nevanimibe in a phase 2, single-blind, placebo-controlled study of adults with classic 21OHD.

Methods

Study design and subjects

This was a phase 2, single-blind, multiple-dose, multicenter study of nevanimibe with placebo-washout components. The study was conducted in 4 centers in the US: University of Michigan, the National Institutes of Health Clinical Center, Mayo Clinic, and Children’s Hospital of Philadelphia. The study included adult subjects (18-80 years of age) with classic CAH due to 21OHD, with elevated 17-hydroxyprogesterone (17-OHP) levels (≥4× the upper limit of normal [ULN]), who maintained a stable glucocorticoid and mineralocorticoid regimen for at least 1 month before study entry. Exclusion criteria included nonclassic CAH, adrenal insufficiency due to other disorders (including other forms of CAH), liver abnormalities, recent surgery, active cancer, or other significant comorbidities. Comprehensive history, physical examination, laboratory assessment, and electrocardiogram were performed at screening and subsequent visits. Routine laboratory assessment included complete blood count and comprehensive chemistry panel with liver function tests. Hormones were obtained in the morning prior to dosing with steroids and study drug. Hormone panel included 17-OHP, androstenedione, free testosterone, dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEAS), 11-deoxycorticosterone, direct renin, aldosterone, progesterone, and ACTH through a central laboratory (MedPace, Cincinnati, OH). ACTH was assessed using enzyme immunoassay; aldosterone was measured using chemiluminescence immunoassay; DHEAS was assessed using electrochemiluminescence immunoassay; free testosterone was calculated based on total testosterone and sex hormone binding globulin levels obtained using electrochemiluminescence immunoassay and albumin levels obtained using photometry; 11-deoxycorticosterone, 17-OHP, androstenedione, DHEA, and progesterone were quantified by tandem mass spectrometry; and direct renin was assessed using quantitative immunoradiometry. A standardized glucocorticoid equivalent was calculated for each subject by multiplying the current glucocorticoid dose by 1 for hydrocortisone, by 5 for prednisone or prednisolone, and by 80 for dexamethasone (1). Premenopausal women were assumed to be in the luteal phase of the menstrual cycle during the 14 days prior to the start of menses, and were otherwise generally regarded as being in the follicular phase.

Treatment

After a 14-day run-in period, treatment consisted of 5 dose levels of nevanimibe at 125, 250, 500, 750, and 1000 mg administered orally, twice daily for 14 days at each dose level, followed by 14 days of placebo washout after each dose level (Fig. 1). The study drug was administered approximately every 12 hours immediately after a meal. The study drug efficacy and safety were evaluated at the end of each treatment period. The primary endpoint was met if the 17-OHP fell to ≤2× ULN. All subjects started with dose level 1 and the dose was progressively escalated if the primary endpoint was not met by the end of the current treatment period, and if no safety concerns were identified. Once the primary endpoint was met, the treatment was stopped. The study drug was supplied as tablets containing 125 mg, 250 mg, or 500 mg of nevanimibe HCl or matching placebo.

Figure 1.

Study design: phase 2 study of nevanimibe in the treatment of congenital adrenal hyperplasia. The treatment period consisted of 5 dose levels of nevanimibe starting at 125 mg followed by 250, 500, 750, and 1000 mg twice daily for 14 days at each dose level, followed by 14 days of placebo washout after each dose level. Subjects proceeded to the next dose level if the primary endpoint was not met. †Nevanimibe dosing commenced following a 2-week placebo lead-in period. *Subject proceeded if primary endpoint was not yet met.

Safety was evaluated by monitoring adverse events, including changes in vital signs, physical examination, electrocardiogram, and laboratory parameters. Following screening and baseline lead-in visits, subjects were seen every 2 weeks, at the time of each new dose level or placebo period.

Objectives

The primary objective of this study was to evaluate the safety and efficacy of nevanimibe in subjects with classic 21OHD. The secondary objectives were to evaluate the changes in adrenal cortical steroids and steroid intermediates, changes in ACTH, and the pharmacokinetics of nevanimibe.

Statistical analysis

The safety population and the pharmacokinetics population included all subjects who received at least 1 dose of nevanimibe.

Efficacy analysis.

The percentage of responders was tabulated for the safety population overall. A responder was defined as a subject who met the primary endpoint of 17-OHP ≤2× ULN (the 17-OHP target for premenopausal women was based on the normal range for the menstrual cycle phase) at any dose level. The percentage of subjects showing “drug effect” was tabulated for the safety population overall. A subject showing “drug effect” was defined as a subject who was a responder or who had a negative mean percentage change in 17-OHP during the nevanimibe treatment periods. The percentage of responders and percentage of subjects showing drug effect were also tabulated for the per-protocol population overall as a sensitivity analysis to test the robustness of the primary result. The change and percentage change in 17-OHP were also calculated during the placebo washout periods (i.e., from the end of each nevanimibe treatment period to the start of the next nevanimibe treatment period for each dose level to evaluate the changes in 17-OHP during placebo washout. Similar assessments were performed on other hormones (e.g., androstenedione).

The percentage of “placebo responders” was tabulated for the safety population and the per protocol population overall. A “placebo responder” was defined as a subject who met the primary endpoint at the end of any placebo washout period. The percentage of subjects showing “placebo effect” was tabulated for the safety population and the per-protocol population overall. A subject showing “placebo effect” was defined as a subject who was a placebo responder or who had a negative mean percentage change in 17-OHP during the placebo washout periods.

Pharmacokinetic analysis.

Actual collection times were used in pharmacokinetic parameter calculations. The Linear Trapezoidal Linear Interpolation method was used in the computation of all area under the curve (AUC) values. For the calculation of AUC_0-4, if the nominal (planned) time of the last observed measurable concentration was at 4 hours, but the actual time of the last observed measurable concentration was <4 hours (such as actual time = 3.9 hours), the concentration at 4 hours was imputed. The imputed concentration was calculated using the linear extrapolation of the previous 2 records.

Safety analysis.

Adverse events were classified using the Medical Dictionary for Regulatory Activities (MedDRA) and summarized by dose level and treatment (nevanimibe or placebo) at adverse event onset. Other safety parameters were summarized using descriptive statistics by time point.

Sample size determination.

Sample size considerations were based on a Simon 2-stage Minimax design with 1-sided alpha of 5% and 80% power.

Stage 1 consisted of 9 completed subjects. The study was designed to stop at Stage 1 if either of the following conditions were met: (1) futility criterion: defined as ≤2 of 9 (22%) subjects having a Day-15 17-OHP ≤2× ULN or a negative mean percentage change (i.e., decrease) in 17-OHP during nevanimibe treatment periods; or (2) success criterion: defined as ≥7 of 9 (78%) subjects having a Day-15 17-OHP ≤2× ULN or a negative mean percentage change (i.e., decrease) in 17-OHP during nevanimibe treatment periods.

Stage 2 was to have commenced if 3, 4, 5, or 6 of the 9 subjects met the primary endpoint. In this scenario, accrual would continue with up to 8 additional subjects for a total of no more than 17 subjects. However, accrual was stopped after Stage 1 as the success criterion was met.

Results

Of 10 enrolled participants, 9 completed the study, and 1 withdrew due to a serious adverse event of gastroenteritis (subject 2) during dose level 5. This subject was excluded from efficacy analyses due to receiving stress doses of glucocorticoids. The cohort was 50% male, with the majority (90%) being Caucasian, and had a mean age of 30.3 years at screening. The baseline 17-OHP level ranged from 7× ULN to as high as 187× ULN, with a mean value of 53× ULN. All subjects continued their baseline mineralocorticoid and glucocorticoid regimen unchanged throughout the study, except subject 2 required stress-dose steroids. At baseline, 1 subject was treated with dexamethasone, 6 with hydrocortisone, 2 with prednisolone, and 1 with prednisone. All but one (subject 5) received supraphysiologic replacement doses of glucocorticoid (≥12 mg hydrocortisone/m²), ranging from 10.3 to 23.4 mg/m² hydrocortisone equivalencies (Table 1).

Table 1.

Subject Demographics and Baseline Hormonal Status in Phase 2 Study of Nevanimibe in Congenital Adrenal Hyperplasia

		Baseline 17-OHP
Subject	Sex/Age (yrs)	Level (ng/dL)	Multiple of ULN	GC and MC Daily Doses†	Body Surface Area (m2)	MC Replacement	GC Replacement (equivalency in mg hydrocortisone/m2 body surface area)§
1	M/52	26 000	187	HC 10/10/10 mg P 5 mg FC 0.1/0.1 mg	2.136	Yes	23.4
2	M/18	11 400	82	P 3/3 mg^ FC 0.1/0.1 mg	1.919	Yes	12.5
3	M/25	10 500	76	HC 10/10/5 mg FC 0.1 mg	1.875	Yes	13.3
4	M/29	8,420	61	HC 10/10/10 mg FC 0.1 mg	2.266	Yes	13.2
5	M/26	7,370	53	P 2.5/2.5 mg FC 0.15/0.1 mg	1.935	Yes	10.3
6	F/61	3,780	18	HC 5/2.5 mg P 5 mg FC 0.05 mg	1.679	Yes	16.4
7	F/25	1,210	17	HC 10/10 mg FC 0.1 mg	1.510	Yes	13.2
8	F/25*	950	14	HC 10/10 mg FC 0.1 mg	1.481	Yes	13.5
9	F/21	768	11	P 2.5/2.5 mg FC 0.1 mg	1.553	Yes	12.9
10	F/24*	481	7	D 375 mcg FC 0.15 mg	1.626	Yes	18.5

Calculation of Multiples of ULN was performed using the normal upper limit for the individual subject’s age, sex, menstrual cycle phase, and/or menopausal status, as appropriate. The baseline 17-OHP values for subjects 7–10 were obtained during the follicular phase.

Abbreviations: 17-OHP, 17-hydroxyprogesterone; F, female; FC, fludrocortisone; GC, glucocorticoid; HC, hydrocortisone; M, male; m², square meters of body surface area; MC, mineralocorticoid; P, prednisone/prednisolone; ULN, upper limit of normal range

†Divided doses are indicated by a ‘/’ between doses. For all subjects except subject 2, glucocorticoid and mineralocorticoid doses were stable during the treatment period of the study

§The standardized glucocorticoid equivalent was calculated for each subject by multiplying the current glucocorticoid dose by 1 for hydrocortisone, by 5 for prednisone or prednisolone, and by 80 for dexamethasone (1)

^Subject 2 also received stress-dose steroids for serious adverse events of viral gastroenteritis and enteritis

*Met the primary endpoint of 17-hydroxyprogesterone (17-OHP) ≤ 2x upper limit of normal (ULN)

Normal range for serum 17-hydroxyprogesterone:

•Adult males: <139 ng/dL

•Post-menopausal females: <207 ng/dL

•Pre-menopausal adult females, follicular phase: <70 ng/dL

Seventy percent of subjects demonstrated ≥50% decrease in 17-OHP during at least one of the nevanimibe treatment periods. Two subjects (20%) met the primary endpoint (17-OHP ≤2× ULN): one after dose level 2, and one after dose level 5; and 5 other subjects experienced maximal decreases in 17-OHP ranging from 27% to 72% during nevanimibe treatment. A potential drug effect was observed in 80% of subjects. However, 1 subject who met the technical definition for having a drug effect did not appear to have a consistent pattern of changes in 17-OHP that reflected a response to nevanimibe.

Overall, 17-OHP levels for the group decreased during treatment periods for all 5 dose levels of nevanimibe, whereas an increase 17-OHP levels for the group was seen during all placebo washout periods (Fig. 2A).

Figure 2A.

17-Hydroxyprogesterone changes (%) during treatment and placebo in phase 2 study of nevanimibe in congenital adrenal hyperplasia. The white boxes represent changes during nevanimibe treatment. The grey boxes represent changes during treatment with matching placebo for corresponding dose level. Box and whisker plot representations: horizontal lines within the boxes represent the median; top and bottom edges of each box represent the interquartile range([i.e., the 75th percentile [upper quartile] and 25th percentile [lower quartile], respectively); and the high and low whiskers represent the inner fence (i.e., 1.5× the interquartile range above and below the median). Open circles represent outliers. Note: 1 extreme outlier value for the placebo washout period of Dose Level 2 (7850.98% in subject 10) is not shown in the figure. Abbreviation: BID, twice daily.

Figure 2B. Androstenedione changes (%) during treatment and placebo in phase 2 study of nevanimibe in congenital adrenal hyperplasia. The white boxes represent changes during nevanimibe treatment. The grey boxes represent changes during treatment with matching placebo for corresponding dose level. Box and whisker plot representations: Horizontal lines within the boxes represent the median; top and bottom edges of each box represent the interquartile range (i.e., the 75th percentile [upper quartile] and 25th percentile [lower quartile], respectively); and the high and low whiskers represent 1.5× the interquartile range above and below the median. Open circles represent outliers. Abbreviation: BID, twice daily.

Androstenedione levels showed less consistent changes with a decline in group androstenedione levels only during nevanimibe dose levels 4 and 5 (Fig. 2B). However, 67% of subjects achieved decreases in androstenedione of at least 30% during nevanimibe treatment, compared with 22% of subjects during placebo. In addition, androstenedione levels tracked with changes in 17-OHP in 70% of subjects. Figure 3 presents 17-OHP and androstenedione values for one of the subjects who met the primary endpoint. No observable changes were seen in other hormones: free testosterone (stratified by sex), DHEA, DHEAS, 11-deoxycorticosterone, direct renin, aldosterone, progesterone, cortisol, or ACTH, during nevanimibe treatment at any dose level.

Figure 3.

Androstenedione and 17-hydroxyprogesterone values during the treatment period in a selected subject. Androstenedione (upper limit of normal range [ULN] for adult females <39 years of age, 2.14 ng/mL) and 17-hydroxyprogesterone (17-OHP, solid line; ULN for follicular phase of menstrual cycle, 70 ng/dL) levels in subject 8, a 25-year-old woman with classic congenital adrenal hyperplasia who participated in the study and met the primary endpoint. Androstenedione levels mirrored changes in 17-OHP starting with the nevanimibe 500 mg twice daily dose. Study subjects received each dose level of nevanimibe for 2 weeks, followed by a 2-week placebo washout, for each of 5 dose levels. Abbreviation: BID, twice daily.

As expected, the mean area under the nevanimibe concentration-time curve from time 0 to 4 hours increased appropriately with each dose level increase (Fig. 4). The mean maximal observed concentration of nevanimibe also increased in a dose-dependent manner and reached 147, 290, 683, 946, and 1262 ng/mL, at dose levels 1, 2, 3, 4, and 5 respectively (data not shown). The half-life (t_1/2) of nevanimibe in pharmacokinetic modeling studies was found to be approximately 10 hours.

Figure 4.

Pharmacokinetics of nevanimibe: mean AUC_0-4 in phase 2 study of nevanimibe in congenital adrenal hyperplasia. Area under the concentration-time curve for nevanimibe from time 0 to 4 hours (AUC_0-4) increased appropriately with each dose level increase from dose level 1 (125 mg bid) to dose level 5 (1000 mg bid). Values are mean ± SEM. AUC_0-4 = area under the curve from time of dose to 4 hours post-dose.

Overall, across all dose levels including placebo washout periods, 9 (90%) subjects experienced a total of 58 treatment-emergent adverse events (TEAEs). TEAEs by MedDRA system organ class are shown in Table 2. A total of 31 TEAEs were observed during nevanimibe treatment periods, and a total of 27 TEAEs were observed during placebo washout periods. The majority of TEAEs were mild to moderate in severity. One serious adverse event (enteritis) occurred during nevanimibe treatment and resulted in study withdrawal. TEAEs considered to be drug-related occurred similarly during nevanimibe treatment periods (3 subjects and 10 TEAEs) and placebo (3 subjects and 5 TEAEs). The most common cause of drug-related TEAEs was gastrointestinal, and the frequency was similar during nevanimibe treatment (30%) and placebo washout (20%). Other less commonly reported related TEAEs included edema, somnolence/dizziness, and pruritis. There were no TEAEs of diminished libido or erectile dysfunction. A total of 2 serious adverse events (SAEs), viral gastroenteritis and enteritis, occurred during the study; both SAEs occurred for the same subject (subject 2) during nevanimibe treatment. This subject experienced a severe SAE of viral gastroenteritis at dose level 2, which was considered unrelated to study drug, and another SAE at dose level 5, which was considered related to study drug; study drug was then discontinued. This was the only event that led to study discontinuation. There were no deaths during the study. Following gastrointestinal symptoms, the most common adverse events were headache in 40% overall and migraine in 30% overall; these events occurred during both nevanimibe and placebo treatment periods.

Table 2.

Treatment-Emergent Adverse Events in Phase 2 Study of Nevanimibe in Congenital Adrenal Hyperplasia

MedDRA System Organ Class	Overall Nevanimibe (N = 10) n (%)	Overall Placebo (N = 10) n (%)	Overall (N = 10) n (%)
Patients with any TEAE	8 (80%)	9 (90%)	9 (90%)
Cardiac disorders	0	1 (10%)	1 (10%)
Gastrointestinal disorders	5 (50%)	4 (40%)	7 (70%)
General disorders and administration site conditions	1 (10%)	1 (10%)	2 (20%)
Infections and infestations	4 (40%)	4 (40%)	6 (60%)
Injury, poisoning, and procedural complications	1 (10%)	0	1 (10%)
Investigations	0	1 (10%)	1 (10%)
Metabolism and nutrition disorders	2 (20%)	1 (10%)	3 (30%)
Musculoskeletal and connective tissue disorders	0	1 (10%)	1 (10%)
Nervous system disorders	5 (50%)	4 (40%)	8 (80%)
Psychiatric disorders	1 (10%)	1 (10%)	2 (20%)
Respiratory, thoracic, and mediastinal disorders	1 (10%)	3 (30%)	4 (40%)
Skin and subcutaneous tissue disorders	3 (30%)	1 (10%)	3 (30%)

Treatment-emergent adverse events (TEAEs) were defined as adverse events that happened for the first time on or after the first dose date (Dose Level 1, Day 1) of nevanimibe. Events starting during nevanimibe treatment were attributed to nevanimibe, and events starting during placebo treatment were attributed to placebo. Events were coded using the Medical Dictionary for Regulatory Activities (MedDRA) (version 18.0). Counts are of number of subjects. System organ classes with events occurring in at least 1 subject are shown. Percentages were calculated using the number of subjects in the column header as the denominator. At each level of summarization, a subject was counted only once if the subject reported 1 or more TEAEs within that level.

Discussion

ACAT1 catalyzes the esterification of free cholesterol to cholesteryl ester, the storage pool of cholesterol substrate for adrenal steroid biosynthesis. The use of nevanimibe in the treatment of CAH is supported by a detailed understanding of the mechanism of action (5-8). At lower doses, nevanimibe treatment results in a reduction of the cholesteryl ester reservoir and thus reduced adrenal steroidogenesis (5). ACAT1-derived inhibition of adrenal steroidogenesis could potentially reduce the need to suppress ACTH with supraphysiologic doses of glucocorticoid while still maintaining control of excess androgens. We present here results from a first study of nevanimibe HCl (ATR-101), a novel ACAT1 inhibitor, in classic 21OHD.

The intent of this proof-of-mechanism study was to demonstrate that nevanimibe could decrease 17-OHP and potentially other relevant biomarkers/hormones that are elevated in classic 21OHD. Although variability in 17-OHP was observed at baseline (which ranged from 7× to 187× ULN) and during the study, no overall response was observed during placebo administration, as mean 17-OHP levels increased during all placebo washout periods while mean 17-OHP decreased during nevanimibe treatment periods. The study design (i.e., 2 weeks of nevanimibe followed immediately by a 2-week placebo washout period, for 5 consecutive dose levels) resulted in a distinct pattern of response to nevanimibe that was deemed to be consistent with a drug effect. The increase in 17-OHP during the placebo washout periods indicates that the apparent effect of nevanimibe persists for less than 2 weeks after discontinuation.

Two of the 10 (20%) subjects met the primary endpoint (17-OHP ≤2× ULN) with both having relatively mild elevations of 17-OHP at baseline (<1000 ng/dL), suggesting that the doses and/or the duration of treatment may not have been sufficient for subjects with higher baseline elevations in 17-OHP to meet the primary endpoint. It is probable that a 2-week treatment period was insufficient to deplete the presumably larger cholesteryl ester stores in subjects who may have had longer-standing, poorly-controlled disease (mean 17-OHP at baseline was 53× ULN). However, 70% of subjects had at least a 50% decrease in 17-OHP associated with nevanimibe treatment, consistent with the anticipated drug activity. The 2-week placebo washout periods between the nevanimibe treatment periods might have allowed replenishment of cholesteryl ester stores, which could have counteracted the nevanimibe mechanism of action (i.e., preventing formation of cholesteryl esters and depleting this substrate) and most likely impacted overall efficacy assessments.

Unlike with 17-OHP, consistent reductions in androstenedione were not observed following each dose level of nevanimibe. This discrepancy could reflect the short duration of treatment intervals (2 weeks) or suggest that higher doses are needed to see a more robust effect, since larger mean changes in androstenedione occurred only with the higher doses. However, 67% of subjects achieved at least a 30% decrease in androstenedione during nevanimibe treatment, compared with 22% of subjects during placebo; and changes in androstenedione did track with 17-OHP in the majority of subjects.

Patients with CAH have significant comorbidities, including impaired metabolic profile, tumor formation, infertility, and poor quality-of-life (3, 9-12). Various studies attribute these adverse outcomes to inadequate treatment regimens and the practice of using supraphysiologic glucocorticoid therapy to suppress androgen production (12, 13). New treatments aimed at minimizing glucocorticoid exposure are being developed including improved glucocorticoid delivery methods (14, 15), sex steroid antagonists and synthesis inhibitors (16, 17), and corticotropin-releasing hormone receptor type 1 antagonists (18). Nevanimibe, by inhibiting steroidogenesis, theoretically has the potential to be used as an add-on therapy allowing for physiologic glucocorticoid replacement by removing hyperandrogenism risk. Maintaining near physiological hormone balance would likely reduce the comorbidities present in many patients with CAH.

Nevanimibe was generally well tolerated, with overall adverse event rates similar to placebo. The most commonly reported side effects for nevanimibe were gastrointestinal (30%) and occurred at the higher doses (mainly dose level 5; 1000 mg twice daily). Other less commonly reported side effects included edema, somnolence/dizziness, and pruritis. These adverse events were consistent with those observed in a completed phase 1 maximum tolerated dose study in adrenocortical carcinoma (ACC) patients (NCT01898715) (19). In the ACC study, nevanimibe doses were escalated for an apoptotic effect with the highest dose achieved of 158 mg/kg per day (approximately 11 000 mg per day for a 70-kg individual). If tolerated at high doses, nevanimibe could potentially be used to treat adrenal and adrenal rest tumors in CAH patients. The current study evaluated nevanimibe at lower doses, focusing the efficacy outcome on steroidogenesis inhibition rather than cellular toxicity. In preclinical studies, nevanimibe was concentrated primarily in the adrenal glands, and no gonadal toxicity was observed. Furthermore, the SOAT1 (ACAT1) enzyme is highly expressed in the adrenal cortex (20), which limits the potential for nevanimibe to impair gonadal steroidogenesis in human beings. Nevertheless, we cannot exclude an effect of chronic nevanimibe exposure on human gonadal function beyond the dosing scheme of this clinical trial.

In conclusion, in this early phase 2 study of 2-week interval treatments of nevanimibe, reductions in 17-OHP were observed in the majority of subjects, and 2 of 10 subjects met the primary endpoint. Because of the short duration of treatment, it was not possible to evaluate the effect of nevanimibe on clinical endpoints, such as menstrual status in women or testicular adrenal rest tumors in men. Future studies with longer duration and possibly higher doses of nevanimibe are needed to further evaluate its safety, efficacy and tolerability in patients with CAH, especially in patients with poor disease control.

Additional Information

Disclosure Summary: D.P.M. received research funds from Diurnal Limited and Millendo Therapeutics US, Inc. through the National Institutes of Health Cooperative Research and Development Agreement and is a commissioned officer in the US Public Health Service. A.Y.C. is a contracted researcher for Millendo Therapeutics US, Inc. V.H.L, L.A.W., M.P., and P.M. are current or former employees of Millendo Therapeutics US, Inc. R.J.A. is a contracted researcher for Novartis Pharmaceuticals, Strongbridge Biopharma, Millendo Therapeutics US, Inc., Spruce Biosciences, Corcept Therapeutics, and Neurocrine Biosciences; and serves as consultant for Quest Diagnostics, Corcept Therapeutics, Janssen Pharmaceuticals, Novartis Pharmaceuticals, Adrenas Therapeutics, Selenity Therapeutics, and Strongbridge Biopharma.

Data Availability: The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.