SAGE-217, A Novel GABA_A Receptor Positive Allosteric Modulator: Clinical Pharmacology and Tolerability in Randomized Phase I Dose-Finding Studies

The SAD and MAD studies were approved by the Institutional Review Boards at each study site and were performed in accordance with ethical standards as described in the 1964 Declaration of Helsinki and its later amendments. Written informed consent was provided at screening and was required for enrollment in each study. Participants were of both sexes, ranging in age from 22 to 55 years. No formal sample size calculations were undertaken for these PK, safety, and tolerability studies. The number of subjects in each cohort and at each dose level was thought to be sufficient to assess PK, safety, and tolerability for determining escalating doses of SAGE-217 according to predetermined criteria.

The SAD trial was a multicenter, double-blind, placebo-controlled trial conducted at two sites in the US from 29 September 2015 (first subject enrolled) to 7 June 2016 (last subject, last visit), while the MAD trial was a single center, double-blind, placebo-controlled study conducted in the US from 6 January 2016 (first subject enrolled) to 28 April 2016 (last subject, last visit). The staggered start of the trials allowed the SAD trial to inform MAD dosing, although the trials ran in parallel after MAD trial initiation. All authors vouch for the accuracy and completeness of the data, data analyses, and the fidelity of this publication to the study protocol.

The SAD and MAD studies were conducted in a double-blind manner. The active and placebo treatments were identical in appearance within each study. In order to facilitate dose selection for future cohorts, Sage Therapeutics, Inc. may have unblinded individual or group data from a cohort once all subjects in that cohort completed their follow-up visits and any AEs were resolved.

Subjects were randomly assigned to either placebo or active treatment with SAGE-217 according to a randomization schedule prepared by an independent statistician. In each of the SAD cohorts, subjects were randomly assigned to receive either SAGE-217 (six subjects) or placebo (two subjects) in a blinded manner, and in each of the MAD cohorts, subjects were randomly assigned to receive either SAGE-217 (nine subjects) or placebo (three subjects) in a blinded manner.

SAGE-217 doses were prepared in an aqueous solution containing hydroxypropyl β-cyclodextrin and administered as single oral doses in each study. Sentinel dosing was employed for the first SAD cohort, with one subject randomized to receive SAGE-217 and the other subject randomized to receive placebo on the first day. The other six subjects in the first cohort were dosed approximately 24 h later. Escalation to the next dose cohort was undertaken only after safety and PK data were reviewed by the Safety Review Committee (SRC) and agreement reached that it was safe to increase the dose. Each SAD cohort was dosed at approximately weekly intervals to allow adequate time for collection and review of safety and PK data. Six subjects in each of the nine SAD cohorts received a single SAGE-217 dose of 0.25, 0.75, 2, 5.5, 11, 22, 44, 55, or 66 mg. Doses were selected to interrogate a predefined exposure range bounded by a maximum recommended starting dose determined from non-clinical studies and doses triggering predefined stopping criterion, informed by emerging pharmacodynamic and tolerability data. The utilization of a solution formulation allowed flexibility in dose selection. The starting dose of 0.25 mg was selected based on established guidelines, and provided a > 50-fold margin below the no adverse effect dose observed in non-clinical toxicology studies [21].

The total daily dose for each MAD cohort was based on information obtained from the SAD study. Nine subjects each received single SAGE-217 doses of 15 mg (Cohort 1), 35 mg (Cohort 2), and 30 mg (Cohort 3) in the morning for 7 days; nine subjects received placebo for 7 days. Subjects participating in Cohort 3 were dosed in the morning for 7 days (Part 1), and after a suitable washout period of at least 7 days, they returned for a second period of 7 days of dosing in the evening (Cohort 3, Part 2) to determine if there was a differential tolerability profile compared with morning dosing. For both the SAD and MAD studies, dosing was escalated until the maximum tolerated dose (MTD) was achieved based on prespecified stopping criteria (see the outcomes below; MTD guidelines are shown in the electronic supplementary material).

Bioanalysis of plasma samples for the determination of SAGE-217 levels was conducted utilizing a validated liquid chromatography-tandem mass spectrometry method at Agilux Laboratories (Worcester, MA, USA). Plasma sample (50 µL) was spiked with stable isotopic-labeled internal standard (d3-SAGE-217), and the sample was extracted using protein precipitation. The analyte was separated on a C18 column using a gradient, and ionized using electrospray in a positive ion mode. Multiple reaction monitoring (MRM) was employed to monitor SAGE-217. The peak area ratio of SAGE-217 to internal standard was used to determine the concentration. The calibration curve was validated using linear regression with 1/x² weighting. The concentrations of the quality controls and PK samples were interpolated from the calibration standard curve. The method was validated with a dynamic range of 1.00–1000 ng/mL. The intra-assay accuracy and precision ranged from − 5.7 to 7.0 (% bias) and from 2.3 to 15.9 (percentage coefficient of variation [%CV]). The interassay accuracy and precision ranged from − 4.8 to 3.0 (% bias) and from 4.1 to 14.3 (%CV). Stability was established for four freeze/thaw cycles when stored at − 20 or − 80 °C, and for 2 h as whole blood at room temperature. SAGE-217 long-term storage stability in frozen plasma was established for 367 days at − 20 and − 80 °C.

Determination of the MTD was an endpoint of both the SAD and MAD studies. The MTD in both trials was determined by prespecified stopping criteria (see the electronic supplementary material for details). Briefly, dosing was escalated in both studies until an ascending dose triggered one or more of the defined stopping criteria. At this point, a lower dose was tested, and if this dose did not trigger any of the defined stopping criteria, it was considered the MTD. Overall safety and tolerability of SAGE-217 were assessed by the frequency and severity of adverse events (AEs), the Stanford Sleepiness Scale (SSS) [22], the Modified Observer’s Assessment of Alertness/Sedation Scale (MOAA/S) [23], vital signs, changes in clinical laboratory measures, physical examinations, electrocardiograms (ECGs), and suicide ideation using the Columbia-Suicide Severity Rating Scale (C-SSRS) [24].

Plasma samples were taken from all cohorts in both studies for PK analyses. Plasma samples for PK analysis were collected according to the sampling collection times specified. For both the SAD and MAD studies, blood samples were taken for analysis at 0, 0.25, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6, 7, 8, 10, 12, 16, 24, 28, 32, 36, and 48 h. Time zero samples were collected just prior to study drug administration. An additional PK sample may have been collected at any time if clinically indicated and at the discretion of the Investigator (e.g. for unusual or severe AEs).

Derived PK parameters included area under the plasma concentration–time curve from the time of dosing extrapolated to infinity (AUC_∞), the terminal-phase half-life (t_½), C_max, time to reach maximum concentration (t_max), and apparent clearance (CL/F). PK parameters were calculated from the individual plasma concentrations, and summarized using appropriate descriptive statistics.

The safety population was defined as all subjects who were administered study drug, while the PK population was defined as all subjects who were administered SAGE-217 and had at least one quantifiable plasma concentration. For each measured parameter, the change from baseline value was calculated at each time point and was summarized. Out-of-range safety endpoints were categorized as low or high, where applicable. Plasma concentrations that were below the limit of quantification were substituted with zero for the calculation of descriptive statistics by time point. SAGE-217 concentrations in plasma were measured utilizing a validated method with liquid chromatography and tandem mass spectrometry (see the Bioanalysis, above).

Descriptive summaries were presented by treatment and time point, where applicable. For categorical endpoints, summaries included counts and percentages, and for continuous endpoints, summaries included number of observations, mean, median, standard deviation, coefficient of variation, minimum, and maximum. Dose proportionality was analyzed using regression analysis [25].

In the SAD study, a total of 72 subjects with a mean age of 35.9 years were randomized (Table 1), 54 subjects to SAGE-217 and 18 subjects to placebo. Seventy-one subjects completed the study; however, one subject in the 44 mg cohort was lost to follow-up.

Thirty-six subjects were enrolled and treated in the MAD study (12 subjects in each of the MAD cohorts), with a mean age of 38.3 years (Table 1). All subjects who received SAGE-217 15 mg, SAGE-217 35 mg, and placebo completed the study. Of the nine subjects who received SAGE-217 30 mg in the morning for 7 days, seven (77.8%) completed a second period of 7 days dosing in the evening after a washout period. Two of the nine subjects (22.2%) discontinued from the study, one due to pre-existing vagotonia discovered after the first dose of study drug, and one due to the investigator’s discretion regarding a low hemoglobin on period 1, day 14. The total SAGE-217 group included 27 subjects, while the placebo group included 9 subjects. Additional demographic information is provided in Table 1.

Following a single dose of SAGE-217 oral solution, plasma concentrations rapidly increased and reached maximum concentrations approximately 1 h postdose. Concentrations then declined in a biphasic manner and exhibited a terminal-phase half-life ranging from 16 to 23 h (Fig. 1a, b; Tables 2a, b). Terminal-phase half-life estimates were determined from the higher doses of the SAD cohort (≥ 11 mg), where there were sufficient time points above the assay limit of quantitation (1 ng/mL). For both single and repeated dose administration, t_max and terminal-phase half-life showed dose independence. The dose proportionality of SAGE-217 was examined using data from both the SAD and MAD studies, encompassing a 264-fold range of doses. SAGE-217 exhibited dose proportionality as indicated by the analysis of C_max, area under the concentration–time curve from time zero to the last measurable concentration (AUC_last), or AUC_∞, with point estimates for the slope near 1 in the SAD study and 95% confidence intervals containing 1 in the MAD study over the dose range studied (Fig. 2, Table 3). Accumulation of SAGE-217 after 7 days of dosing was less than twofold and was consistent with steady-state predictions based on terminal-phase half-life (Table 2a, b). Negligible SAGE-217 was detected in urine, indicating that renal excretion of unchanged drug does not contribute to CL (data not shown).

The MTD for both the SAD and MAD studies was determined by application of predefined stopping criteria (see the electronic supplementary material). These stopping criteria were assembled to assess AEs anticipated based on the primary pharmacology of SAGE-217 (i.e. sedation due to GABA_A receptor PAM activity). The maximum dose in each study (66 mg for SAD; 35 mg for MAD) triggered the MOAA/S stopping criterion of a recorded and confirmed MOAA/S score of ≤ 2 during normal waking hours. Two subjects in the 66 mg SAD cohort recorded MOAA/S scores ≤ 2 within the first 2 h postdose, and two subjects in the 35 mg MAD cohort recorded MOAA/S scores ≤ 2 within the first hour postdose, resulting in cessation of dose escalation in the respective studies. MOAA/S scores for these subjects returned to normal by 7 h postdose. In each study, dosing was reduced to the MTD in the subsequent cohort, which did not trigger any stopping criteria. The SAD MTD was established at 55 mg, and the MAD MTD was 30 mg/day.

In both the SAD and MAD studies, there were no serious treatment-emergent AEs or AEs leading to withdrawal from the study (electronic supplementary Tables S1a, b). No subjects reported severe AEs in the MAD study, while four subjects reported severe AEs in the SAD study, all at the highest dose of 66 mg of SAGE-217 (change in mental status, n = 2; unresponsive to stimuli, n = 2; and somnolence, n = 1). Most AEs were reported to be mild. Sedation was the most common AE overall, occurring in 23 of 72 subjects (31.9%; including three placebo subjects) in the SAD study, and 25 of 43 (58.1%; including two placebo subjects and three SAGE-217 30 mg evening-dose subjects) in the MAD study. In both studies, sedation increased with increasing dose of SAGE-217, consistent with the pharmacology of GABA_A receptor PAMs (electronic supplementary Tables S1a,b). There were no reported AEs of loss of consciousness in either the SAD or MAD study at any dose. Fewer AEs related to the primary pharmacology of the drug (such as sedation, somnolence, dizziness, and fatigue) were reported with evening dosing compared with daytime dosing.

As dose-dependent changes in sedation was a predicted effect based on both the proposed mechanism of action for SAGE-217 and preclinical animal studies, the SSS [22] and MOAA/S scale [23] were used in concert with AE reporting to investigate any relationship between SAGE-217 concentrations and reports of sedation or somnolence. The SSS is a self-reported measure of sleepiness, where a score of 1 is ‘subject feeling active, vital, alert, or wide awake’ and the highest score of 7 indicates ‘no longer fighting sleep, sleep onset soon; having dream-like thoughts’. The MOAA/S scale is a clinician-rated scale of sedation where 0 indicates ‘no response after a painful trapezius squeeze’ and 5 indicates ‘readily responds to name spoken in a normal tone’. The maximum mean SSS and minimum mean MOAA/S scores for each dosing period occurred at 1 h postdosing for all doses, with no apparent change in the temporal relationship between dosing and maximum score over 7 days of dosing. The mean values for the 30 mg morning MTD group are shown in Fig. 3. In this dosing condition, the highest recorded SSS score for any subject was 6, and all such values were recorded between 1 and 3 h after dosing. Similarly, the lowest MOAA/S score recorded in a subject in the 30 mg dose group was 4 at 2 h after dosing. In contrast, the 35 mg dosing condition triggered the MOAA/S stopping criterion (see above).

The 30 mg evening dosing condition also resulted in a maximum individual SSS score of 6 between 1 and 3 h after dosing, and the lowest MOAA/S score of 3 was recorded in one subject at 3 h after dosing in a single dosing period. At the next measured time point (12 h postdose, the morning following dosing), no subject reported an SSS score higher than 2 in any dosing period, and all subjects reported the maximum MOAA/S score (i.e. the least sedated) of 5 at the 12-h time point in every dosing period.