Fixed dosing versus weight-based dosing of HIV-1 prophylactic monoclonal antibodies in adults: a post-hoc, cross-protocol pharmacokinetics modelling study

Abstract

Background: Pharmacokinetic (PK) modelling and simulations have been used to support label changes of dosing levels or strategies for multiple marketed therapeutic monoclonal antibodies (mAbs). Using data from early-phase clinical trials in adults without HIV-1, we compared fixed and weight-based dosing strategies for three HIV-1 broadly neutralising mAbs planned for prevention efficacy evaluation: PGDM1400LS, PGT121.414.LS, and VRC07-523LS.

Methods: We used a two-compartment population PK model to describe overall trends and inter-individual variability in post-administration serum concentrations over time from individuals administered PGDM1400LS (n = 95), PGT121.414.LS (n = 113), or VRC07-523LS (n = 251) subcutaneously or intravenously. We evaluated the effect of body weight on various PK parameters, including clearance rate, and simulated mAb concentrations after fixed and weight-based dosing administrations using sex-specific weights observed in participants from two recent HIV-1 mAb efficacy trials. We compared magnitudes and inter-individual variabilities of concentrations at specific post-administration timepoints, areas under the time-concentration curves (AUC), and predicted neutralisation titres against representative HIV-1 virus strains.

Findings: For all three mAbs, we observed a modest effect of body weight on clearance rate and volumes of the central and peripheral compartments. The population-level magnitude and variability in time-specific concentrations, AUC, and predicted neutralisation titres were comparable between the two dosing strategies for both sexes. The relationship between body weight and concentrations differed between the two dosing strategies with a positive correlation for weight-based dosing and a negative correlation for fixed dosing. For individuals with body weight below the 15th or above the 85th percentiles, fixed dosing resulted in <3% difference in median AUC compared to the overall population. For lower weight individuals, fixed dosing improved AUC, potentially correcting the underdosing seen in the previous weight-based mAb efficacy trials. For higher weight individuals (e.g., >100 kg), body weight-based dosing or a higher fixed dose may be preferred.

Interpretation: For HIV-1 prophylactic mAbs, a fixed-dose approach, possibly banded by weight categories may be advantageous over weight-based dosing, as it offers increased operational efficiency while maintaining comparable pharmacokinetics and inter-individual consistency.

Funding: NIAID.

Keywords: Dosing strategy; Exponent of body weight; HIV prevention; Population pharmacokinetics model; Two-compartment; Variability in exposure.

Fig. 1

Study diagram. The first column shows the sample sizes and administration routes for each mAb from each of the five trials. The second column shows the combined sample sizes for each mAb considered in the subsequent popPK modelling. The third and last column shows the sizes of the populations considered in the simulations to assess the body weight-based and fixed dosing strategies for each mAb. For weight-based dosing, the dose level was 20 mg/kg for PGDM1400LS and PGT121.414.LS, and 40 mg/kg for VRC07-523LS. For fixed dosing, the dose amount was the product of the weight-based dose level and sex-specific median body weight for each mAb. ∗Treatment groups included in this analysis do not necessarily encompass all treatment groups of contributing trials (Supplementary Table S1).

Fig. 2

Distributions of steady-state area under the time-concentration curve (AUC), Day 1, Week 12 and Week 24 concentrations by sex under the weight-based and fixed dosing strategies. The final popPK models described in Supplementary Table S1 and the body weight observed in female and male participants enrolled in the AMP trials were used to simulate mAb-specific concentrations in steady state after administrations of each mAb every 24 weeks; in steady state, concentration profiles are stable, and no accumulations are expected. The mean and standard deviation (SD) of the log-transformed values are displayed above each violin plot.

Fig. 3

Changes in steady-state area under the time-concentration curve (AUC), Day 1, Week 12 and Week 24 serum concentrations between the weight-based and fixed dosing strategies. For easier readability, shown are data from a random sample of 100 individuals. The final popPK models described in Supplementary Table S1 and the body weight observed in female and male participants enrolled in the AMP trials were used to simulate mAb-specific concentrations in steady state after administrations of each mAb every 24 weeks; in steady state, concentration profiles are stable, and no accumulations are expected.

Fig. 4

Relationship between steady-state area under the time-concentration curve (AUC) and body weight under the weight-based (blue) and fixed (red) dosing strategies for PGDM1400LS (Panel A), PGT121.414.LS (Panel B) and VRC07-523LS (Panel C). The straight line is a simple linear regression line. The final popPK models described in Supplementary Table S1 and the body weight observed in female and male participants enrolled in the AMP trials were used to simulate mAb-specific concentrations in steady state after administrations of each mAb every 24 weeks; in steady state, concentration profiles are stable, and no accumulations are expected.

Fig. 5

Predicted steady-state 80% inhibitory neutralisation titre (PT80) under the weight-based (blue) and two fixed doses based on median weight (red) or 80 kg weight (purple) for the triple combination mAb regimen PGDM1400LS + PGT121.414.LS + VRC07-523LS, by body weight range and overall. The final popPK models described in Supplementary Table S1 and the body weight observed in female and male participants enrolled in the AMP trials were used to simulate mAb-specific concentrations in steady state after administrations of each mAb every 24 weeks; in steady state, concentration profiles are stable, and no accumulations are expected. The dashed line at a PT80 value of 200 corresponds to an estimated threshold required for 90% prevention efficacy against acquisition of these viruses. Error bars indicate 95% confidence intervals.