Proton pump inhibitors do not reduce atazanavir concentrations in HIV-infected patients treated with ritonavir-boosted atazanavir

Concern was recently expressed regarding the co-administration of atazanavir with proton pump inhibitors (PPI). A European Medicines Agency public statement released in December 2004 recommended that the co-administration of atazanavir and omeprazole should be avoided, because of a marked fall in atazanavir plasma concentrations observed in healthy subjects [1,2]. Between 20 and 38% of HIV-infected patients report PPI use at some time during atazanavir therapy [3], but the risk of interaction has only been assessed in healthy volunteers. This prospective open-label study compared plasma atazanavir trough concentrations (Cmin) in HIV-infected patients receiving atazanavir plus ritonavir (300/100 mg), with or without a PPI, from January to March 2005 in two HIV clinics in Paris. Patients receiving more than two protease inhibitors or non-nucleotide reverse transcriptase inhibitors were excluded from the study. The patients receiving a stable atazanavir/ritonavir regimen for at least one week were interviewed by a physician on any antacid treatment they were taking and when they had taken their last atazanavir/ritonavir dose. Only atazanavir Cmin measured 24 ± 4 h after the last intake was taken into account. Atazanavir Cmin was determined with a validated high-performance liquid chromatography-based method with quantification cutoffs of 10 ng/ml for atazanavir [4]. Data were analysed using the Mann–Whitney test. A total of 92 patients (77 men, 15 women) were enrolled, of whom 13 (11 men, two women) stated they were taking a PPI. Among them, 10 were using omeprazole (20 mg a day in nine patients, 40 mg a day in one patient) and three were using rabeprazole (20 mg a day). All but one was taking PPI tablets in the evening. None of them had liver or kidney dysfunction. Twenty-nine patients were taking tenofovir as part of their treatment (four in the PPI group, 25 in the group without PPI). The median atazanavir Cmin was 551 ng/ml (203–1976) in the patients receiving PPI and 469 ng/ml (65–1944) in the group without PPI (Table 1). In patients treated with tenofovir, atazanavir Cmin values were 554 (203–1976) and 481 ng/ml (65–1944), respectively. Neither PPI nor tenofovir treatment affected atazanavir Cmin values (P = 0.86). In three PPI-treated patients, the atazanavir Cmin was also measured before and between 2 and 14 days after PPI discontinuation. The results did not differ between the two timepoints (1944 and 1976, 520 and 551, 1332 and 1259 ng/ml in patients 1, 2 and 3, respectively).Table 1: Atazanavir concentrations (median and range).In healthy individuals, a pharmacokinetic study of the potential interaction between tenofovir and atazanavir/ritonavir (300/100 mg) showed a 20% reduction in the atazanavir Cmin[5]. In keeping with previous studies of HIV-infected patients, we found no effect of tenofovir on the atazanavir Cmin[6–8]. Data on patients treated with tenofovir were thus taken into account. We found a significantly (2.5-fold) lower median atazanavir Cmin than that reported in healthy individuals (482 versus 1206 ng/ml, P = 0.01). In contrast, our results are in keeping with values previously found in HIV-infected patients (496–862 ng/ml) [6–9]. We found that the PPI intake had no significant impact on atazanavir Cmin during ritonavir-boosted atazanavir therapy, both when we compared PPI-treated and PPI-untreated patients, and when we compared values during and after PPI treatment. Because of its chemical structure, atazanavir absorption is highly dependent on an appropriate gastric pH, and potential interaction with PPI is expected to reduce atazanavir exposure. During PPI treatment, the time during which the intragastric pH is above 4.0 increases with the acid-suppressive dose, and pH can vary widely during the day in a given individual [10]. The study by Agarwala et al. [2] of healthy volunteers involved a relatively high PPI dose (40 mg/day omeprazole), and it is possible that lower doses may not significantly affect atazanavir absorption. The time interval between atazanavir and PPI administration may also play a role, and it is possible that our patients may not have accurately stated the time of the last drug intake. Our results may also be explained by gastric hypochlorhydria, which has been reported in HIV-infected patients [11]. However, a recent case report with pharmacokinetic results showed no interaction between boosted atazanavir and lansoprazole, and another showed no clinical or immunovirological failures after 6 months’ follow-up in 14 patients treated with atazanavir and PPI [12,13]. Our findings are in keeping with these results. In clinical practice, PPI appear to be compatible with boosted atazanavir therapy. It may nonetheless be prudent to monitor atazanavir drug concentrations. Finally, our results show that the results of pharmacological studies in healthy subjects must be systematically confirmed in the target patient population. Sponsorship: This study was supported solely by the Department of Infectious and Tropical Diseases of Hôpital Tenon.