Background
Ritonavir (RTV), a strong CYP3A inhibitor, is widely used as a pharmacokinetic (PK) enhancer to increase exposure to HIV-1 protease inhibitors (PIs) and more recently the SARS CoV-2 PI nirmatrelvir. Pregnancy-related changes in RTV disposition and boosting capacity have not been systematically assessed which may inform RTV dosing in pregnancy for future emerging infectious diseases. We performed a model-based meta-analysis for the population PK (popPK) of RTV using historical data from 11 separate arms of IMPAACT P1026s, a phase IV study evaluating the PK of selected antiretroviral drugs in pregnant and postpartum women with HIV.
Methods
Pregnant participants receiving RTV as a PK enhancer for either lopinavir, darunavir, or atazanavir±TDF and who were ≥20 weeks gestational age were included in the analysis. Intensive PK samples were collected pre-dose and through 24 hours post-dose in the 2nd and 3rd trimesters of pregnancy and 2-12 weeks postpartum. A popPK model was developed using non-linear mixed effects modeling (NONMEM v7.5).
Results
A total of 279 participants contributing 3798 RTV plasma concentrations (565 2nd trimester, 1632 3rd trimester, and 1601 postpartum) were included. RTV concentrations were well fit with a one-compartment structural model. Pregnancy was an independent predictor of both apparent clearance (CL/F) and apparent volume of distribution (Vd/F). No other covariates significantly impacted RTV disposition. Fixed allometric scaling significantly improved the model. The following equations described the final popPK model:
CL/F (L/h) = 12.0 × 1.84 (if Pregnant) × (WT/66.4)0.75
V/F (L) = 21.1×2.27 (if Pregnant) ×(WT/66.4)
Monte Carlo simulations of 1000 virtual subjects receiving a 100 mg daily dose predicted a median AUC0-24, ss of 3.85 and 7.09 μg*hr/mL during pregnancy and postpartum, respectively (Figure 1). Based on established in vivo concentration-effect relationships of CYP3A inhibition by RTV1, at a RTV dose of 100 mg daily, CYP3A metabolic clearance is expected to be reduced to 13% of baseline in non-pregnant adults compared to 24% of baseline during pregnancy.
Conclusions
RTV PK was well described by a one compartment model with first-order elimination and pregnancy significantly increased RTV CL/F and V/F. Decreased RTV exposures during pregnancy are predicted to lead to a 1.8-fold reduction in RTV-mediated CYP3A inhibition. Dosing requirements of RTV and/or boosted CYP3A substrates may be altered during pregnancy.
Table or Figure
