Study DesignStudy 1: Participants with Hepatic Impairment

Study 1 was a phase I, nonrandomized, open-label, multiple-dose, parallel-cohort, two-part study to investigate the effects of hepatic impairment on the plasma PK, safety, and tolerability of ritlecitinib. Part 1 planned to enroll ≈16 participants: eight with moderate hepatic impairment and eight HPs with normal hepatic function. Participants were aged 18‒70 years, body mass index (BMI) of  ≥17.5 to ≤40 kg/m2, and body weight of  >50 kg. Dependent on outcomes in part 1, part 2 planned to enroll ≈8 participants with mild hepatic impairment and would be conducted if the point estimate of ritlecitinib area under the plasma concentration–time curve (AUC) for 24-h dosing interval (AUC0-24) geometric mean ratio (GMR) for the moderate hepatic impairment cohort versus the HP cohort was  ≥2.0. Part 2 was not conducted as the criterion to progress was not met (the point estimate of the GMR was 1.19; see the “Results”).

Participants in the hepatic impairment cohort met the criteria for class B of the Child–Pugh classification (6) (moderate: 7–9 points), within 28 days of ritlecitinib administration. They had stable hepatic impairment, defined as no clinically significant change in disease status within 30 days prior to screening, and had a diagnosis of hepatic dysfunction due to hepatocellular disease.

HPs had no known or suspected hepatic disease and no evidence or history of clinically significant hematologic, renal, endocrine, pulmonary, gastrointestinal, cardiovascular, hepatic, psychiatric, neurologic, dermatologic, or allergic disease. HPs were enrolled following participants with moderate hepatic impairment and matched for mean age (± 10 years), weight (± 15 kg), race, and sex (± 2 participants per sex) of the moderate impairment cohort.

Participants received oral doses of ritlecitinib 30 mg QD under nonfasting conditions for 9 days. On day 10, participants received ritlecitinib 30 mg following an 8-h fast. Blood samples were collected predose on days 7, 8, 9, and 10 for assessment of ritlecitinib minimum plasma concentration, and at 0.5, 1, 2, 3, 4, 6, 8, 12, 14, and 24 h postdose on day 10 for PK assessments (AUC0-24 and maximum observed plasma concentration [Cmax]). Multiple doses were used to allow ritlecitinib concentrations to achieve steady state. PK parameters were calculated for each participant and each treatment using noncompartmental analysis (NCA) of concentration–time data. Samples below the lower limit of quantification were set to zero. The study design was consistent with guidance for PK studies in patients with impaired hepatic function (7).

All-causality and treatment-related treatment-emergent adverse events (TEAEs) were monitored throughout the study and via telephone follow-up 28 + 3 days after last dose administration. Clinical laboratory tests, vital sign assessments, and physical examinations were performed during screening and days specified in the protocol.

Study 2: Participants with Renal Impairment

Study 2 was a phase I, nonrandomized, open-label, multiple-dose, parallel-cohort, multisite, two-part study to investigate the effect of renal impairment on the plasma PK, safety, and tolerability of ritlecitinib. Part 1 planned to enroll ≈8 participants with severe renal impairment and 8 HPs with normal renal function. Approximately 8 participants with moderate renal impairment and 8 with mild renal impairment were to be enrolled in part 2. Participants in the severe renal impairment group were recruited first: HPs were to be recruited later so that mean age (± 10 years) and weight (± 15 kg) could be matched. Following part 1, part 2 would be conducted if the point estimate of ritlecitinib AUC0-24 GMR for the severe renal impairment cohort versus the HP cohort was  >2.0.

Participants were aged 18‒75 years, with BMI of  ≥17.5 to ≤40 kg/m2 and body weight of  >50 kg. Participants in the severe renal impairment cohort met criteria for severe renal impairment based on the Modification of Diet in Renal Disease equation (5) (estimated glomerular filtration rate [eGFR]  <30 mL/min but not requiring hemodialysis) but were otherwise in good general health commensurate with the population with renal impairment. Participants in the HP cohort had no clinically relevant abnormalities and eGFR  ≥90 mL/min.

Participants received oral doses of ritlecitinib 50 mg QD under nonfasting conditions for 9 days. On day 10, participants received ritlecitinib 50 mg following a fast of  ≥10 h. Blood samples were collected to assess ritlecitinib concentration predose on days 8, 9, and 10, and 0.25, 0.5, 1, 2, 4, 6, 8, 12, 16, and 24 h postdose on day 10. Steady-state (day 10) AUC0-24 and Cmax were calculated using the NCA approach. The study design was consistent with guidance for PK studies in patients with impaired renal function (8).

TEAEs were monitored throughout the study and by telephone follow-up 28‒35 days after last dose. Physical examinations and vital signs assessments were performed during screening and on days specified in the protocol. Clinical laboratory tests and eGFR measurements were conducted during two screening visits, on day 1, and on days 2, 5, 8, and 11 (at the second screening visit, and on days 2 and 8, only serum creatinine for eGFR measurement was assessed).

Estimation of Effect of Renal Impairment on Ritlecitinib Exposure

The severe renal impairment cohort in study 2 was completed; however, the study was paused due to the COVID-19 pandemic, and the HP cohort for part 1 was not enrolled. HP data were obtained from available data of completed phase I studies using two different approaches: (1) HP data from completed study 1 (shared cohort approach) and (2) in silico HP cohorts derived from a population PK (POPPK) model.

Analysis of Variance Approach

A statistical approach using analysis of variance (ANOVA; SAS v9.4) was applied to compare the natural log-transformed ritlecitinib AUC0-24 and Cmax of the various cohorts. For study 1, the HP (reference) and the moderate hepatic impaired function (test) cohorts were compared. For study 2, the HP cohort (shared cohort) from study 1 (reference) and the severe impaired renal function cohort from study 2 (test) were compared. The HP cohort from study 1 was selected as a reference group because the cohort demographics closely matched that of the severe renal impairment cohort from study 2. HPs with eGFR  <90 mL/min were excluded from the analysis to meet the study 2 inclusion/exclusion criteria. This assessment was consistent with traditional methods as described in US Food and Drug Administration guidance for studies in patients with impaired renal function (8). Since systemic exposures of ritlecitinib increase linearly from 30- to 50 -mg doses, the PK parameters (AUC0-24 and Cmax) for HP cohort from study 1 were normalized to the 50-mg dose administered in study 2 for analyses.

Estimates of adjusted mean differences (test − reference) and corresponding 90% confidence intervals (CIs) were obtained from the model. The mean differences and 90% CIs for the differences were exponentiated to provide estimates of the ratio of geometric means (test/reference) and 90% CIs for the ratios.

Clinical Trial Simulation Approach Population Pharmacokinetic Model

To simulate in silico HP cohorts, a POPPK model for ritlecitinib was developed using pooled data from two phase I studies in HPs (NCT02309827 and NCT02684760) and one phase II study in patients with RA (NCT02969044) (Fig. 1). A POPPK model had been previously developed using NONMEM® and standard nonlinear mixed-effect modeling procedures and goodness-of-fit diagnostics (Text S1 and Figs. 1–3 of Supplementary Material).

Fig. 1

Simulation and analysis of an in silico cohort. AUC24, area under the plasma concentration–time curve for 24-h dosing interval; QD, once daily

The final model was a two-compartment model with first-order oral absorption, interindividual variance on apparent clearance (CL/F) and apparent central volume of distribution (Vc/F), and a proportional residual error model. Covariates incorporated into the model included allometric scaling on CL/F and Vc/F referenced to a 75-kg individual with exponents of 0.75 and 1, respectively; effect of patients with RA on CL/F; effect of food on first-order absorption rate constant; and effect of total daily doses  >100 mg on CL/F. There was no apparent difference in clearance with respect to age.

The simulation was conducted in two parts independently for study 1 and study 2. The first part simulated PK parameters of demographically matched HP cohorts for the completed study 1, which served as an external validation of the simulation approach. The second part applied the simulation to study 2. The 1000 trials were simulated and for each study, the in silico HP cohort consisting of eight individuals with PK parameters was created by randomly generated weights from a uniform distribution between 69.9 and 99.9 kg (study 1) or 73 and 103 kg (study 2). Individual random-effect parameters for CL/F and Vc/F were drawn from the variance–covariance matrix described by the POPPK model. Each simulated individual was administered their respective study dose for 10 days to reflect actual study conduct. Simulated concentrations (individual predictions with residual unexplained variance applied) were sampled on day 10 following the study schedule. Those below the lower limit of quantification were set to zero. AUC0-24 was calculated for each individual based on the linear-up/log-down NCA approach using the PKNCA package (v0.9.1) in R (v3.6.1).

Comparison with Hepatic and Renal Impairment Cohort

To compare simulated HP reference data with the study 1 (moderate hepatic impairment) and study 2 (severe renal impairment) cohorts, the GMRs (90% CI) for AUC0-24 and Cmax were calculated for each in silico HP cohort relative to their respective impairment cohort using ANOVA. Results for the 1000 trials were summarized by median GMR (and nonparametric 90% prediction interval [PI] of all GMRs), mean GMR (and mean lower and upper 90% CI derived from ANOVA results), and proportion of trials with a GMR of  ≥2.0.

Ethics

All study protocols were reviewed and approved by each site’s institutional review board or ethics committee and conducted in accordance with the Declaration of Helsinki and in compliance with all International Council for Harmonisation Good Clinical Practice Guidelines. All participants provided written, informed consent.