Marjorie Imperial1, Emma Hughes1, Ravindre Panchia2, Onyema Ogbuagu3, Susan Buchbinder4,5, Lillian Brown1, Alexander Kintu1, Renu Singh1, Ana Ruiz-Garcia1
1Gilead Sciences, Inc., 2Perinatal HIV Research Unit, University of the Witwatersrand, 3School of Medicine, Yale University, 4San Francisco Department of Public Health, 5University of California
Introduction: Lenacapavir (LEN) is a first-in-class HIV-1 capsid inhibitor administered as a twice-yearly subcutaneous (SC) injection for HIV-1 prevention. In two Phase 3 trials, twice-yearly SC LEN administered with an initial oral loading regimen was highly effective, with a favorable safety profile for pre-exposure prophylaxis (PrEP) in diverse populations [1,2]. Whereas oral LEN is rapidly absorbed (apparent half-life of 10-12 days), SC LEN absorption occurs over several months (apparent half-life of 8-12 weeks) due to flip-flop kinetics [3,4]. The absorption of LEN is complex following SC administration due to its prolonged release kinetics. This complexity is further compounded when combined with oral loading doses, posing challenges for pharmacokinetic (PK) characterization and identification of covariates that describe interindividual variability. Objectives: •Utilize a stepwise population PK (popPK) modeling approach to characterize the absorption and disposition of LEN following intravenous (IV), oral, and SC administration in healthy volunteers and people who want or need PrEP •Identify potential covariates of LEN PK parameters and quantify their impact on LEN exposures Methods: LEN plasma concentrations from 13 studies (N = 1337 participants; N = 14,648 samples) in healthy volunteers, people with HIV-1 (PWH), or people who want or need PrEP were included. To characterize the oral and SC absorption, as well as the disposition of LEN, data from healthy volunteers were incrementally incorporated in a stepwise manner during model development: IV data for disposition, IV and oral data for oral absorption, and IV and SC data for SC absorption [5]. Covariates including, but not limited to, body weight, age, sex assigned at birth, race, fed status, and LEN dose were evaluated on model parameters, as appropriate. Once the disposition and absorption models were optimized, data from PWH and people who want or need PrEP were incorporated to evaluate potential population-specific covariates. These covariates included, but were not limited to, age (eg, adolescents), gender identity, use of long-acting contraceptives (etonogestrel, medroxyprogesterone, and norethisterone enanthate), use of gender-affirming hormone therapy (estradiol or testosterone), and gestational age of pregnancy. Covariate selection was based on visual inspection of the data, improvement in model fit, and clinical relevance. At each step of model development, random effect models were evaluated. NONMEM version 7.5 was used to develop the popPK model. Simulations were conducted to evaluate the impact of covariates on LEN exposures (AUC and Cmax between Weeks 0 and 26, and Ctrough at Week 26), relative to a reference group (72 kg, fasted, male sex at birth), for the twice-yearly SC LEN simplified regimen (600 mg oral LEN on Day 1 and Day 2, 927 mg SC LEN on Day 1 and every 26 weeks thereafter). Results: LEN PK data were described by a three-compartment model with linear disposition, a transit model with an absorption lag and dose-dependent bioavailability for oral absorption, and three parallel absorption phases with increasing delays for SC absorption. Here, covariates for people who want or need PrEP are presented. With fixed allometric scaling, the impact of baseline body weight in people who want or need PrEP on LEN exposures (AUC, Cmax, and Ctrough) ranged from -29% to +35% for people with extreme covariate values (ie, 95th and 5th weight percentiles, respectively), relative to median exposures. Additionally, sex assigned at birth and fed status were included as covariates in the model but both had minimal impact on LEN exposures (< 10%). No other covariates were identified in graphical evaluations with PK parameters or as statistically significant (P value < 0.01) in the model. Conclusions: A stepwise modeling approach was implemented to characterize the complex PK of LEN based on a large, robust dataset. Body weight, sex assigned at birth, and fed status were identified as statistically significant covariates in the popPK model; however, their effects on LEN exposures were not considered clinically relevant. Therefore, no dose adjustments in people who want or need PrEP for LEN are required and LEN can be administered without regard to food. Age, gestational age of pregnancy, use of long-acting contraceptives, use of gender-affirming hormone therapy, and gender identity had no impact on LEN exposures in people who want or need PrEP.
[1] Bekker L-G, et al, N Engl J Med. 2024;391:1179-92. [2] Kelley CF, et al, N Engl J Med. Published online November 27, 2024. doi:10.1056/NEJMoa2411858. [3] Sunlenca US Prescribing Information, Gilead Sciences, Inc., 2022 https://www.accessdata.fda.gov/drugsatfda_docs/label/.2022/215973s000lbl.pdf. [4] Sunlenca EU Summary of Product Characteristics, Gilead Sciences, Inc., 2022 https://www.ema.europa.eu/en/medicines/human/EPAR/sunlenca. [5] Imperial MZ, et al. Abstract 11073 presented at: PAGE 32; June 26-28, 2024; Rome, Italy www.page-meeting.org/?abstract=11073.
Reference: PAGE 33 (2025) Abstr 11420 [www.page-meeting.org/?abstract=11420]
Poster: Drug/Disease Modelling - Infection