Commentary

In this issue of the Journal, Muirhead et al. [1] report on the pharmacokinetic interaction between sildenafil (Viagra®) and the HIV protease inhibitors ritonavir and saquinavir. The single dose pharmacokinetic profile of sildenafil was markedly affected by the protease inhibitors. Ritonavir showed a greater effect than saquinavir with increases in sildenafil AUC and C_max of 11-fold and 3.9-fold compared with 3.1-fold and 2.4-fold for coadministration with saquinavir. There was no effect of sildenafil on the steady state pharmacokinetics of saquinavir or ritonavir. These data are consistent with a previous study [2] which demonstrated that the AUC of sildenafil when given with the protease inhibitor indinavir was 4.4-fold greater than data from historical controls.

Not all drug interactions reported in the literature are clinically relevant; this one has direct relevance to HIV-infected men. The prevalence of erectile dysfunction in this group has been reported as 33%, increasing with more advanced HIV disease. Many patients wish to avail themselves of sildenafil therapy but given that they routinely receive multiple medications – antiretroviral drug combinations and other drugs – there is justifiably concern about potential interactions. Protease inhibitors, particularly ritonavir, are among the most potent inhibitors of CYP3A4 and other cytochrome P450 isoenzymes [3, 4]. With sildenafil undergoing oxidative metabolism to an N-demethylated metabolite, catalysed by CYP3A4 (major route) and CYP2C9 (minor route), it is not surprising that the protease inhibitors have such a major impact on sildenafil metabolism in vivo. What is possibly surprising from the report of Muirhead et al. [1] in this healthy volunteer study, is that the increases in systemic bioavailability of sildenafil were not associated with any change in blood pressure, heart rate or ECG parameters. In contrast, Merry et al. [2] in HIV-infected patients noted a mean maximal decrease in blood pressure of 14/10 mmHg with 25 mg sildenafil plus indinavir (which was greater than the maximum decrease noted following 100 mg sildenafil in other clinical studies). In both studies the most frequently reported adverse events were headache, facial flushing, dyspepsia and rhinitis. Despite this difference between the two reported studies (healthy volunteers vs patients?) the authors are agreed that a low starting dose of sildenafil (25 mg [1], 12.5 mg [2]) is advisable in patients concomitantly receiving protease inhibitors. And, given the pharmacokinetic interaction with prolongation of the pharmacodynamic effect [2], there is reason to recommend not exceeding a maximum dose in a given time period.

Altered pharmacokinetics can be expected when sildenafil is coadministered with any drug which inhibits CYP3A4 (and possibly CYP2C9). There is a threefold increase in plasma sildenafil concentrations in the presence of erythromycin and a 50% increase with cimetidine [5]. There are therefore grounds to be cautious when sildenafil is used with a whole range of drugs. We probably need to know more about the role of the active metabolite (UK 103 320; 50% potency of parent compound) and whether inhibition of its further breakdown is important for pharmacological activity. Then there is the whole area of enzyme inducers which will reduce the effect of sildenafil by lowering plasma concentrations but will increase the active metabolite. Given the recent interest in interaction with St John's Wort [6] and other herbal medications, it is clear that pharmacokinetic and drug interactions studies live on. Maybe they are not at the cutting edge of innovative clinical science but who can argue that such studies are not important for patient management. That is surely a good enough reason for this journal to maintain a pro-active role in publishing key drug interactions.