A model-based comparison of dose reduction strategies for fixed-dose dolutegravir-containing regimens.

BACKGROUND: Countries faced with an impending shortage of antiretrovirals need to conserve supplies and maximize treatment effectiveness while limiting harms. Potential strategies to achieve this could include reduced doses of fixed-dose tenofovir disoproxil-lamivudine-dolutegravir (TLD). METHODS: Nonlinear mixed effects was applied to dolutegravir concentration-time data from three healthy volunteer studies (two including dolutegravir 'washout' over 10 and 15 days post-cessation), adapting a previously published model. The model was used to simulate (n = 1000) pharmacokinetic profiles of standard dolutegravir (one pill TLD daily) and four dose reduction strategies: half a pill TLD daily; one pill TLD every other day; 4 days on, 3 days off (4:3) and 5 days on, 2 days off (5:2) therapy. Proportions above and time below the EC90 (320 ng/mL) were determined and compared. RESULTS: Seventy-three volunteers (60% female, n = 825 concentrations) were included in the model. Marked differences between simulated dose reduction regimens were observed, with alternate day, 4:3 and 5:2 dosing, associated with a lower proportion of dolutegravir trough above the EC90 (41.6%-62.1%, 2.0%-5.2%, 14.1%-20.7%, respectively) across 50, 70 and 90 kg compared with daily dosing at full or half doses (88.7%-99.8%). Of the four dose reduction strategies, 4:3 dosing exhibited the longest predicted time below target before the next due dose (median ∼2 days). CONCLUSIONS: Safety and effectiveness of TLD dose reduction strategies can only be established in clinical studies and cannot be extrapolated from models. However, these models can inform regimen design to limit the window of opportunity for emergence of resistance, i.e. where viral replication occurs in the presence of appreciable drug concentrations.