Sharon Sawe1, Lufina Tsirizani1, Allan Kengo1, Letisha Najjemba2, Lubbe Wiesner1, Henry Mugerwa3, Catherine Orrell4, Catriona Waitt2,5, Paolo Denti1
1Division of Clinical Pharmacology, Department of Medicine, University of Cape Town, 2Infectious Diseases Institute, Makerere University College of Health Sciences, 3Joint Clinical Research Centre, Research Department, 4Desmond Tutu HIV Foundation, University of Cape Town, 5Department of Women’s and Children’s Health, University of Liverpool
Objectives: Access to life-saving antiretroviral therapy (ART) in hard-to-study populations, such as pregnant women, children, adolescents and individuals with extreme body mass index (BMI), is often delayed due to their exclusion from clinical trials [1], [2]. Even when these drugs become available, there is uncertainty about whether these groups receive appropriate dosing, as pharmacokinetic (PK) and safety data are frequently lacking. This is true for atazanavir boosted with ritonavir (ATV/r), a protease inhibitor [3], commonly used in combination ART in low-income settings. While ATV/r PK is well described in standard populations, data remains sparse in these understudied groups. We aimed to assess, using limited available data, whether the PK of ATV/r in pregnant and postpartum women, individuals with obesity or malnutrition, and children/adolescents, was comparable to that in non-complicated adults with HIV. Methods: This observational PK study (NCT03923231) was nested within the VirTUAL consortium and conducted at three sites (two in Uganda, one in South Africa). A larger, intensive PK study in non-complicated adults with HIV (NCT04121195) served as the main study [4], while these additional cohorts were conducted simultaneously using the same study procedures. However, the sample sizes for these cohorts were smaller and opportunistic, with a much sparser sampling schedule. Participants were individuals with HIV on ATV/r-based second-line ART for at least six months. Pregnant women were enrolled at >20 weeks gestation, obese individuals had a BMI >30 kg/m², malnourished individuals had a BMI <18.5 kg/m², and children/adolescents were <18 years old. All participants received ATV/r 300/100 mg once daily, given orally. Sparse PK sampling was performed across three visits, with each individual contributing at least two blood samples per visit. Drug concentrations were measured using HPLC-MS/MS (lower limits of quantification: 0.03 mg/L for atazanavir, 0.05 mg/L for ritonavir). Data were analyzed in NONMEM v7.5.1. PK models developed in standard HIV-infected adults on ATV/r were used to fit data from each sub-population without re-estimating parameters [5]. Model performance was evaluated using visual predictive checks (VPCs), to assess whether the median observed concentrations fell within the expected confidence intervals predicted by the model. Results: PK data were available from 8 pregnant and postpartum women, 7 obese individuals, 7 malnourished individuals, and 11 children/adolescents, contributing a total of 405 ATV/r concentrations. The median (range) age of adults (pregnant women, obese, and malnourished individuals) was 29 (22–49) years, while children/adolescents were 15 (11–17) years. Pregnant women had a weight of 71.0 (54.0–99.4) kg, and children/adolescents had a weight of 46.5 (32.7–85.2) kg. Obese participants had a BMI of 37.1 (31.2–54.3) kg/m², while malnourished participants had a BMI of 16.5 (13.1–19.4) kg/m². The original two-compartment model with first-order elimination and transit absorption compartments developed in non-complicated adults adequately described atazanavir and ritonavir concentrations across all sub-populations, except for overpredicting exposures in malnourished individuals. Conclusions: We found no major deviations in ATV/r PK for pregnant/postpartum women, children/adolescents, and obese individuals compared to non-complicated HIV-infected adults. However, malnourished individuals had lower-than-expected exposures, likely due to reduced bioavailability. Using a PK model instead of raw concentrations allowed for a meaningful comparison of exposures, even with sparse data, and assessed whether the standard adult model could be used across subpopulations. By conducting all cohorts within a single study under standardized procedures, we minimized variability, enhancing comparability. This approach is faster and more cost-effective than separate studies. We advocate for integrated study designs that incorporate multiple subpopulations under a unified protocol, leveraging modeling to expedite knowledge acquisition and ensure timely, appropriate drug dosing in understudied groups.
[1] A. Gupta, M. D. Hughes, A. J. Garcia-Prats, K. McIntire, and A. C. Hesseling, “Inclusion of key populations in clinical trials of new antituberculosis treatments: Current barriers and recommendations for pregnant and lactating women, children, and HIV-infected persons,” PLoS Med., vol. 16, no. 8, pp. 1–26, 2019, doi: 10.1371/journal.pmed.1002882. [2] P. Dania, “Visual Abstract of ‘How Should We Approach Body Size Diversity in Clinical Trials?,’” AMA J. ethics, vol. 25, no. 12, pp. E914–E915, 2023, doi: 10.1001/amajethics.2023.914. [3] R. Wood, “Atazanavir: Its role in HIV treatment,” Expert Rev. Anti. Infect. Ther., vol. 6, no. 6, pp. 785–796, 2008, doi: 10.1586/14787210.6.6.785. [4] K. Gausi et al., “Pharmacokinetics and Safety of Twice-daily Ritonavir-boosted Atazanavir with Rifampicin,” Clin. Infect. Dis., vol. 78, no. 5, pp. 1246–1255, 2024, doi: 10.1093/cid/ciad700. [5] A. Kengo et al., “Population pharmacokinetic analysis of the drug interaction between ritonavir-boosted atazanavir and rifampicin .,” Popul. Approach Gr. Eur., 2024, [Online]. Available: https://www.page-meeting.org/default.asp?abstract=11045
Reference: PAGE 33 (2025) Abstr 11669 [www.page-meeting.org/?abstract=11669]
Poster: Drug/Disease Modelling - Infection