Background: There are potential benefits to individualizing dosage in patients treated with efavirenz (EFV). We tested a simplified algorithm based on a Bayesian pharmacokinetic approach for guiding dose reduction in patients with EFV concentrations above the 75th percentile (P75) with documented virological efficacy.

Methods: We designed a prospective, open-label, multicentre study. All consenting participants with EFV concentrations above P75 on standard dosage were included in a dose-reduction cycle. Primary end point was the number of patients who reached plasma concentrations within target (1,000-4,000 ng/ml) after, at most, two cycles of dose reduction at 3 and 6 months. CYP2B6 genetic characterization was performed.

Results: Seventy-two patients were screened and 13 fulfilled selection criteria. These patients, with undetectable viraemia on a stable 600 mg EFV-based regimen, had a median (interquartile range) EFV plasma level of 8,112 ng/ml (5,993-10,278) at baseline; 38% (between P75 and P95) qualified for a 400 mg EFV dose, and 62% (above P95) qualified for a 200 mg EFV dose. After one to two dose-reduction cycles, all patients reached targets for EFV plasma concentration at 24 weeks. The predictive dose reduction based on genetic profile differed from dose reduction according to therapeutic drug monitoring (TDM) in three patients. All patients maintained viral suppression at 6 months.

Conclusions: A standardized TDM-guided EFV dose-reduction strategy over a 24-week period was successful, safe and yielded EFV plasma concentrations within the recommended therapeutic range. In addition to improving neuropsychiatric tolerability, EFV dose reduction has the potential to substantially decrease treatment cost.