In a recent study, Van Luin et al demonstrated that geometric mean ratios (95% confidence interval) of the area under the curve (AUC) of atovaquone in patients on efavirenz, lopinavir/ritonavir and atazanavir/ritonavir compared with healthy volunteers were 0.25 (0.16-0.38), 0.26 (0.17-0.41) and 0.54 (0.35-0.83), respectively. Proguanil plasma concentrations were also significantly lower (38-43%). No data on anti-retroviral pharmacokinetics during atovaquone/proguanil prophylaxis were provided. The authors concluded that physicians should be alert to atovaquone/proguanil prophylaxis failures in patients taking efavirenz, lopinavir/ritonavir or atazanavir/ritonavir . The anti-retrovirals' pharmacokinetic during atovaquone/proguanil prophylaxis were analysed. A marked increase in etravirine and saquinavir AUC0-12h (+55% and +274%, respectively) and a slight decrease in raltegravir and maraviroc AUC0-12h (-23% and -9%, respectively) were documented. Although AUCs between pre- and post-prophylaxis are not very different, in the PK profiles of maraviroc and raltegravir the Tmax and the Cmax in post- prophylaxis appear to be earlier than in pre- prophylaxis, suggesting a more fast absorption of both drugs. Unfortunately, it was not possible to determine atovaquone and proguanil plasma concentrations. However, the absence of established minimum effective atovaquone plasma concentrations in the setting of malaria prophylaxis makes it difficult to assess the clinical relevance of dosing atovaquone and proguanil plasma concentrations, except for the evaluation of potential interactions with anti-retroviral drugs . The evidence that atovaquone/proguanil significantly interacts with etravirine and saquinavir, but not with raltegravir and maraviroc, suggests that the mechanism of interaction is related to cytochrome P450. Etravirine is metabolized primarily by hydroxylation through several cytochrome P-450 isoenzymes (mainly CYP3A, CYP2C9 and CYP2C19), followed by glucuronidation of the metabolites. It is also an inducer of CYP3A and a mild inhibitor of CYP2C9, 2C19 and P-glycoprotein . Saquinavir is metabolized by CYP3A4. It is neither an inducer nor an inhibitor of P450 system, but it is substrate and inhibitor of P-glycoprotein and MRP1 [5, 6]. Raltegravir is metabolized via UGT1A1-mediated glucuronidation and is neither an inducer nor an inhibitor of the CYP oxidation system or P-glycoprotein . Maraviroc is extensively metabolized by CYP3A4 and it is a substrate for P-glycoprotein. At higher doses, maraviroc is a potential inhibitor of CYP2D6 and perhaps of P-glycoprotein . In vitro, maraviroc was not found to inhibit any major CYP isoenzyme, including CYPIA2, CYP2B6, CYP2C8, CYP2C19, CYP2D6, and CYP3A4 . Basing on the pharmacokinetic characteristics of these anti-retrovirals, it is supposable that atovaquone, or more likely proguanil, interferes with etravirine and saquinavir for CYP450 mediated metabolism (perhaps on the 2C isoenzymes) with a consequent marked increase in plasma concentrations of both anti-retrovirals. Despite of the clear effect in plasma concentrations of etravirine and saquinavir, in this case the patient noticed no side effects, and blood examinations during prophylaxis documented any lab toxicities. It is possible that, in this case, the absence of drug-related toxicity was due to the use of unboosted protease inhibitor; while for etravirine there are no clear data on the correlation between PK and toxicity .
This case should alert the physicians to be cautious and vigilant when atovaquone/proguanil prophylaxis is prescribed in patients treated with etravirine and/or saquinavir (especially if ritonavir-boosted).