Drug interactions responsible for suboptimal drug exposure can lead to anti-human immunodeficiency virus (anti-HIV) treatment failure. Amprenavir (APV) is an HIV protease inhibitor with 90% inhibitory concentrations of 40 and 250 ng/ml without and with 50% human serum, respectively; its mean trough plasma level (daily dose of 1,200 mg twice a day [b.i.d.]) is 250 ± 200 ng/ml (1, 3–5; S. Piscitelli, S. Vogel, B. Sadler, W. Fiske, L. Metcalf, H. Masur, and J. Falloon, Program Abstr. V Conf. Retroviruses and Opportunistic Infect., abstr. 346, p. 144, 1998). Efavirenz (EFV), a nonnucleoside reverse transcriptase inhibitor induces cytochromes P450-3A (CYP3A) and decreases dramatically plasma APV trough levels in half of the patients (2; Piscitelli et al., Program Abstr. V Conf. Retroviruses Opportunistic Infect; B. Sadler, C. Gillotin, G. E. Chittick, and W. T. Symonds, Program Abstr. XII World AIDS Conf., abstr. 12389, p. 91, 1998).
We analyzed sequentially the APV-EFV interaction in seven HIV-infected patients. Plasma drug trough levels were determined by high-performance liquid chromatography (HPLC) at days 7 and 14, and subsequently as needed, after the initiation of APV (1,200 mg b.i.d.)–EFV (600 mg once a day) therapy. One patient had the expected and stable (365 ± 104 ng/ml) trough levels of APV (days 7, 14, 18, and 90). In three patients, APV levels were low as early as day 7 (53, 67, and 33 ng/ml) and below 20% of the expected mean APV value. Two patients had the expected APV levels at day 7 (510 and 170 ng/ml), which decreased to 65 and 62 ng/ml at day 14. The last patient had APV level determinations only at days 45 and 60, and the levels were low (62 and 48 ng/ml). At day 14, the median plasma APV trough level was 64 (range, 33 to 260) ng/ml.
To improve the pharmacokinetic profile of APV and avoid suboptimal exposure, ritonavir (RTV) (100 mg b.i.d.), a known strong inhibitor of cytochromes P450-3A (CYP3A) was administered concomitantly with reduction of APV dosages (900 mg b.i.d. to 450 mg b.i.d.) to the six patients with low APV trough levels. However, in one patient, EFV was stopped because of a rash, with APV levels going up from 67 to 175 ng/ml, and another patient did not accept RTV. The other four patients received RTV, which led to the increase of APV trough levels of 15, 40, 40-, and 90-fold. One of those patient had sequential determinations of APV levels at days 7, 14, 21, 28, 60, 90, 120, and 180 showing high and stable levels (1,400 and 2,300 ng/ml). In all patients, the successive determinations of EFV levels were within the therapeutic range.
Our data confirm the low plasma APV levels at day 7 in patients receiving concomitant EFV therapy, as previously reported by Piscitelli et al. (Program Abstr. V Conf. Retroviruses Opportunistic Infect.). Sadler et al. (Program Abstr. XII World AIDS Conf.) reported that the magnitude of the decrease was bimodal (15% ± 3% and 67% ± 8%), with equal distributions of patients in these two subgroups. However, the timing of the APV level determination was not mentioned. In our study, the low level of APV was in fact observed in half of the patients at day 7 (3 of 6) but in almost all patients at day 14 (5 of 6), reflecting the time required to approach the steady state of induction of cytochromes P450-3A (CYP3A).
The addition of low-dose RTV therapy reversed the effect of the induction of APV metabolism by EFV. Prior to the addition of RTV therapy at day 15, APV levels were suboptimal and could have induced the emergence of resistance to antiretrovirals. To avoid this, RTV should be added as soon as EFV-APV therapy is initiated. However, the safety of this little-documented antiretroviral combination has to be studied.
We thank Glaxo-Wellcome and Dupont Pharma for generously providing APV and EFV and for HPLC determinations.