Population pharmacokinetics of tecovirimat in persons with mpox: Initial results from ACTG A5418

1 Background: The Study of Tecovirimat for Human Mpox Virus (STOMP) (ACTG A5418) assessed the safety, efficacy and pharmacokinetics (PK) of tecovirimat in clade II mpox infection. We previously showed lower tecovirimat exposures among participants in the open-label arm in comparison to healthy volunteers (HVs), though predicted concentrations at the end of the dosing interval (Ctau) were comparable or above the therapeutic target in non-human primates (NHPs). In clinical trials, tecovirimat did not reduce time to clinical resolution vs. placebo, highlighting the need to examine factors affecting tecovirimat PK in persons with mpox and the adequacy of current dosing strategies. Here, we report the initial population PK modelling of tecovirimat in persons with mpox and comparisons with historical data in HVs. Methods: Participants in the randomized and open-label arms received oral tecovirimat 600 mg either twice daily (weight 40–<120 kg) or three times daily (weight ≥120 kg) for 14 days. Intensive PK sampling occurred on treatment Day 8 at time 0 (pre-dose) and 1, 2, 3, 4, 6, 8 and 10 h post-dose with sparse sampling on Days 3, 8 and/or 15. Plasma tecovirimat concentrations were quantified by a validated LC–MS/MS assay. The population PK model was developed in Phoenix NLME. One- and two- compartment models with additive, multiplicative, or combined error models were investigated. Fixed allometric scaling of weight on apparent volume of distribution (V/F) and apparent oral clearance (CL/F), inclusion of an absorption lag time (Tlag) and the effect of meals on relative bioavailability (F) and absorption rate (ka) were assessed in base model development in alignment with prior modelling in HVs. Simulations with 100 replicates per participant were performed with the final base model, and predicted AUC0-24h, Cmax and Ctau estimates were compared to those previously simulated in HVs (US FDA Clinical Pharmacology Review) and the NHP therapeutic target (169 ng/mL). Results: PK data were available in 75 participants (41 participants with 315 intensive+30 sparse observations; 34 participants with 59 sparse observations; 404 observations total). A one-compartment model with combined error, Tlag and allometric scaling for weight (centred to 70 kg) fixed at 1 and 0.75 on V/F and CL/F, respectively, best fit the available data. Population estimates (95% CI) for ka, Tlag, V/F and CL/F were 0.56 1/h (0.20–0.92), 0.99 h (0.91–1.07), 367 L (222–510) and 51.6 L/h (44.7–58.5), respectively. Interindividual variability for ka, Tlag, V/F and CL/F were 124%, 57.6%, 65.9% and 42.7%, respectively, and generally indicated more variability than HVs. Median simulated AUC0-24h, Cmax and Ctau were 38.8%, 38.5% and 36.4% lower in A5418 participants than HVs. Simulated median Ctau (5th, 95th percentiles) was 599 ng/mL (112, 1523) in A5418 participants, with 8.5% of predicted Ctau values from A5418 falling below the NHP therapeutic target. Conclusions: Lower tecovirimat exposures and greater PK variability were observed in A5418 participants vs. HVs, but simulated Ctau estimates exceeded the NHP therapeutic target in ~91.5% of cases. Investigation into the mechanism of these differences and their relationship to clinical/virologic outcomes is warranted.