LAUSR

Objectives: Atovaquone is one of the alternatives to trimethoprim/sulfamethoxazole for prophylaxis of Pneumocystis jirovecii pneumonia (PCP) in immunocompromised patients. In volunteers, there was wide inter‐individual variability in atovaquone bioavailability. The aim of this study was to assess the plasma concentrations of atovaquone in immunocompromised patients under PCP prophylaxis. Methods: Adult haematology or HIV‐positive patients receiving atovaquone (750 mg oral suspension twice a day) for PCP prophylaxis were included. Plasma concentrations were assessed using UV‐HPLC, around 12 h after the evening dose (Cmin) and 1–5 h after the morning dose (Cmax). Results: A total of 82 measurements were performed in 33 patients. This included 19 HSCT recipients, 7 haematology non‐transplant patients and 7 HIV‐positive patients. The median Cmin (IQR) was 11.3 &mgr;g/mL (6.2–27.8) and the median Cmax was 13.4 &mgr;g/mL (6.0–28.3). The Cmin and Cmax of atovaquone were not different between HIV‐negative and HIV‐positive patients, or between HSCT and non‐HSCT patients. Atovaquone concentrations were not influenced by the co‐administration of valaciclovir (n = 20) or ciclosporin (n = 11), by gut graft‐versus‐host disease (n = 7) or by the intake of atovaquone with food. Nineteen of the 33 (58%) patients had Cmin <15 &mgr;g/mL, a threshold associated with a low rate of clinical response in PCP treatment. Conclusions: Atovaquone is poorly absorbed in more than half of immunocompromised patients and its bioavailability varies between individuals. These unpredictable variations raise the question of therapeutic drug monitoring, in order to identify patients with low concentrations and those who could benefit from regimen adaptation or from alternatives.