The effect of fluconazole on ritonavir and saquinavir pharmacokinetics in HIV-1-infected individuals

Abstract

Aims

To study the effect of fluconazole on the steady-state pharmacokinetics of the protease inhibitors ritonavir and saquinavir in HIV-1-infected patients.

Methods

Five subjects treated with saquinavir and three with ritonavir received the protease inhibitor alone (saquinavir 1200 mg three times daily, ritonavir 600 mg twice daily) on day 1, and the same protease inhibitor in combination with fluconazole (400 mg on day 2 and 200 mg on days 3 to 8). Pharmacokinetic parameters were determined on days 1 and 8.

Results

In the saquinavir group, the median increase in the area under the plasma concentration vs time curve was 50% from 1800 µg l⁻¹ h to 2700 µg l⁻¹ h (P = 0.04, median increase: 900 µg l⁻¹ h; 2.5 and 97.5 percentile: 500–1300), and 56% for the peak concentration in plasma (from 550 to 870 µg l⁻¹, P = 0.04; median increase: 320 µg l⁻¹ h, 2.5 and 97.5 percentile: 60–450 µg l⁻¹). In the ritonavir group, there were no detectable changes in the pharmacokinetic parameters on addition of fluconazole.

Conclusions

Because of the favourable safety profile of saquinavir, dose adjustments are probably not necessary with concomitant use of fluconazole, as is the case for ritonavir.

Introduction

Fluconazole and itraconazole are antifungal drugs that belong to the class of the triazoles. Their mechanism of action is the inhibition of the fungal cytochrome P450-mediated C-14 demethylase, which involves the conversion of lanosterol to ergosterol, eventually leading to leaky cell membranes. Azole antifungals are also substrates for human cytochrome P450 isoenzymes [1, 2]. Fluconazole has high specificity for fungal cytochrome P450, whereas it possesses only weak affinity for human cytochrome P450. Some findings, however, indicate that fluconazole can inhibit cytochrome P450 3A4-mediated metabolism [1, 3, 4]. By inhibiting cytochrome P450 3A4, fluconazole decreases the clearance of several concomitantly administered drugs [3, 4]. Effects of other cytochrome P450-inhibiting or -inducing drugs on the metabolism of fluconazole is, however, not expected, since it is relatively stable to metabolic conversion and is largely excreted unchanged in the urine [1].

Infections with fungi and yeast frequently occur in patients infected with the human immunodeficiency virus type 1 (HIV-1), and antifungal drugs such as fluconazole and itraconazole are often used in combination with antiretroviral drugs. Concomitant use of these classes of drugs could lead to drug–drug interactions, especially with protease inhibitors, which are extensively metabolized by cytochrome P450 enzymes. Protease inhibitors are used to treat HIV-1 infection, mostly in combination with two nucleoside analogue reverse transcriptase inhibitors [5, 6].

In our hospital, we observed a median five-fold increase in the saquinavir area under the plasma concentration vs time curve (AUC) when itraconazole was added to saquinavir-containing regimens [7]. This increase in the exposure to saquinavir is probably due to an increased uptake from the gut as a result of the inhibition of intestinal P-glycoprotein activity by itraconazole and/or the inhibition of cytochrome P450 by itraconazole in the gut wall and liver. As a consequence of this observed drug–drug interaction between itraconazole and saquinavir, the question was raised whether fluconazole could also alter the pharmacokinetics of protease inhibitors.

The metabolism of saquinavir is mainly hepatic and mediated by cytochrome P450 3A4, but intestinal metabolism by the same enzyme has also been reported [6]. Ritonavir is primarily metabolized by cytochrome P450 3A4 and, to a lesser extent, by cytochrome P450 2D6 [5, 8]. Both protease inhibitors inhibit cytochrome P450 3A4, although the inhibition by saquinavir is much weaker than that by ritonavir [6, 8, 9].

In light of the reported inhibition of cytochrome P450 3A4 by fluconazole, the metabolism of protease inhibitors via this pathway, and the regularly occurring combination of protease inhibitors and fluconazole in HIV-1-infected patients, this putative drug–drug interaction may be clinically relevant.

Thus, on theoretical grounds, the addition of fluconazole to the drug schedule of HIV-1-infected patients, treated with ritonavir or saquinavir, could lead to higher plasma concentrations of these protease inhibitors. This may be beneficial for antiretroviral efficacy, as was seen for saquinavir in combination with ritonavir, but could also lead to an increased frequency of side-effects [7, 10, 11]. This study was performed to investigate the effect of the addition of fluconazole on the steady-state pharmacokinetics of saquinavir and ritonavir in HIV-1-infected individuals.

Methods

Patients

Ambulatory patients were recruited from the outpatient clinics of the Slotervaart Hospital, Amsterdam, and the University Hospital Nijmegen, Nijmegen, both in the Netherlands. Patients were eligible for inclusion if they had documented HIV-1 infection, which was established by a positive ELISA and confirmed by Western blot. They had to be taking the same dose of ritonavir (600 mg twice daily) or saquinavir (1200 mg three times daily) for at least 4 weeks plus two nucleoside analogue reverse transcriptase inhibitors. The patients were all over the age of 18 years.

Patients with a history of hypersensitivity to fluconazole, ritonavir or saquinavir, a haemoglobin concentration in blood < 6.0 mmol l⁻¹, or had used fluconazole within 1 month before the first study day were excluded. Other exclusion criteria were the use of inducers of cytochrome P450 such as rifampicin, rifabutin, phenobarbitone, carbamazepine or phenytoin, or the use of inhibitors of cytochrome P450 such as cimetidine or ketoconazole within 1 month before the study. Physical impediments were diarrhoea on the study days, wasting syndrome or active/acute opportunistic infections, except oropharyngeal or oesophageal Candida infection.

Patients were asked to state the occurrence and severity of any side-effects on both study days. Withdrawal from the study was possible at all times without consequences for further treatment. The study was approved by the Institutional Review Boards of the participating hospitals and all patients gave informed consent.

Trial design and treatment

Pharmacokinetic parameters of the protease inhibitors (ritonavir or saquinavir) were determined on 2 days (day 1 and 8) with subjects as in-patients. On day 1, the steady-state pharmacokinetic parameters of the protease inhibitors were determined without the use of fluconazole. Patients came to the hospital before having breakfast and their usual morning medication. At the hospital all prescribed drugs, including the protease inhibitor (saquinavir 1200 mg hard-gelatin capsules, Invirase®, or ritonavir 600 mg former capsule formulation, Norvir®), were ingested during a standard breakfast. If used, didanosine was ingested 1 h before the breakfast. The use of comedication was allowed and all drugs and doses were recorded. Medication was not changed during the study. Serial 10 ml volumes of blood were collected immediately before the protease inhibitor was ingested, and after 30, 60, 90, 120, 150, 180, 240, 300, 360, 420, and 480 min. Plasma was separated by centrifugation at 3000 g for 10 min and stored at −30 °C until assayed. No additional daily dosages of the drugs were ingested before the last blood sample was collected.

In order to achieve concentrations comparable with those at a steady-state situation after daily intake of 200 mg fluconazole, the patients were instructed to ingest 400 mg fluconazole on day 2 with their usual morning medication, followed by 200 mg on days 3–7 [1]. On day 8, the pharmacokinetic parameters of the protease inhibitors were determined again, following concomitant use of 200 mg fluconazole ingested at the same time as the protease inhibitor. Blood samples were collected according to the same schedule as performed on day 1.

Bioanalysis

Concentrations of ritonavir and saquinavir in plasma were determined at the Department of Pharmacy and Pharmacology of the Slotervaart Hospital, Amsterdam, the Netherlands. The plasma concentrations of both drugs were quantified using a sensitive and validated isocratic, reversed-phase ion-pair high-performance liquid chromatographic assay as previously described [12, 13]. For the saquinavir assay, between-day and within-day precisions ranged from 3.4 to 5.2%; the mean accuracy was 105%, the lower limit of quantification 2.5 ng ml⁻¹, and the assay was linear up to concentrations of at least 4000 ng ml⁻¹. For the analysis of ritonavir, the assay was linear over the range of 50 ng ml⁻¹; the within-day and between-day precisions ranged from 0.7% to 6.8%, with a mean accuracy of 98%.

Pharmacokinetic analysis

Plasma concentration–time data were analysed by noncompartmental methods [14]. Pharmacokinetic parameters calculated were: C_max (peak concentration in plasma), t_max (time to reach C_max), AUC (area under the plasma concentration – time curve, calculated by the linear trapezoidal method), CL/F (apparent clearance; F =oral bioavailability), t_1/2 (elimination half-life), V/F (apparent volume of distribution) and C_min (plasma trough concentration). The C_min for ritonavir was extrapolated to 12 h post dose by means of the formula: C_t = C₀·e^–k·t (C_t =concentration at time t, C₀ =concentration at time zero, k =elimination rate constant).

Statistics

The difference in pharmacokinetic parameters on both study days was tested by means of a paired Wilcoxon test. A P value ≤0.05 was considered statistically significant. Statistical analysis was performed using SPSS for Windows®, version 6.1 (SPSS Inc., Chicago, IL, USA).

Results

Eight patients were enrolled in the study, five on saquinavir and three on ritonavir. All patients were male. Their mean age was 45.2 years in the saquinavir group and 42.7 years in the ritonavir group. Co-medication in the saquinavir group could be divided into antiretroviral drugs (zidovudine four patients, zalcitabine one patient, lamivudine three patients, stavudine one patient, didanosine one patient) and other drugs (cetirizine one patient, cotrimoxazole four patients, and budesonide one patient). In the ritonavir group, all patients concomitantly used stavudine. Furthermore, lamivudine (two patients) or didanosine (one patient) were also used. All patients completed the study.

Fluconazole influenced the pharmacokinetics of saquinavir. Although there was large interindividual variability in pharmacokinetic parameters, significant differences were shown in (a) the median value of the AUC that increased by 50% from 1800 µg l⁻¹ h to 2700 µg l⁻¹ h (P = 0.04, median increase: 900 µg l⁻¹ h; 2.5 and 97.5 percentile for the difference: 500–1300), and (b) the median peak plasma concentration (56% increase from 550 to 870 µg l⁻¹, P = 0.04; median increase: 320 µg l⁻¹, 2.5 and 97.5 percentile: 60–450 µg l⁻¹) after the addition of fluconazole (Figure 1a,b). The apparent clearance decreased by 50% (P = 0.04) from a median value of 590 l h⁻¹ to 380 l h⁻¹ (median decrease: 210 l h⁻¹, 2.5 and 97.5 percentile: 90–4170 l h⁻¹). No change occurred in the other pharmacokinetic parameters (median values before fluconazole and 2.5 and 97.5 percentiles for the difference between day 1 and 8 were: t_max =2.5 h (−1.4–2.2 h), t_1/2 = 1.4 h (−1.3–1.9 h), V/F = 2100 l (−17,300–600 l), and C_min =61 µg l⁻¹ (−6–116)). The median saquinavir plasma concentration vs time curves with and without concomitant use of fluconazole are shown in Figure 2. No difference in the incidence, severity, or nature of side-effects with or without concomitant use of fluconazole was reported.

Figure 1.

Individual values for the area under the plasma concentration – time curve (a) and the peak concentration in plasma (b) of saquinavir, with (W) and without (W/O) concomitant use of fluconazole in a dosage of 200 mg daily in five HIV-1-infected patients.

Figure 2.

The median saquinavir plasma concentration vs time curves (1200 mg three times daily) with (□) and without (•) concomitant use of fluconazole in a dosage of 200 mg daily in five HIV-1-infected patients.

For ritonavir, no significant differences in pharmacokinetic parameters were observed between days 1 (control) and 8 (with fluconazole). The median values of the main pharmacokinetic parameters on day 1 and the 2.5 and 97.5 percentiles of the differences between day 1 and 8 were: AUC = 46 000 µg l⁻¹ h (−37,500–7600 µg l⁻¹ h), CL/F = 10 l h⁻¹ (−2.9–0.6 l h⁻¹), t_max =2.5 h (−0.9–0.6 h), and C_max =9400 µg l⁻¹ (−10,800–500 µg l⁻¹).

Discussion

Protease inhibitors show large interindividual variation in pharmacokinetic parameters. For saquinavir, approximately 12-fold variability in the AUC and maximum plasma concentration has been observed [6, 15]. In our study, even larger variability (AUC ± 15-fold, C_max ±27-fold) was evident. The mean oral bioavailability of a single saquinavir dose in the original hard-gel capsule formulation, which was used in our study, taken with food is 4%. The bioavailability of the new soft-gelatin capsule formulation is three to four times higher [10]. This limited oral bioavailability is generally attributed to the combined effects of limited absorption and extensive first-pass metabolism by cytochrome P450 3A4 [6, 15]. Recently, it has become clear that intestinal P-glycoprotein is also likely to contribute to this phenomenon [7, 16]. Protease inhibitors are substrates for this protein that is responsible for the active efflux of drugs from epithelial cells back into the intestinal lumen. Nowadays saquinavir is mainly administered in combination with ritonavir, which causes an approximately 20-fold increase in the AUC of the former, leading to the possibility that the drug should be given at a smaller dose and less frequently. This effect on pharmacokinetics is considered to be due to significant inhibition of cytochrome P450 activity by ritonavir [7, 10, 11].

In the present study, a significant increase was observed for the AUC and C_max of saquinavir when fluconazole was added to the regimen. We did not observe a significant difference in elimination half-life, which may suggest that the interaction takes place mainly in the intestine as a result of cytochrome P450 inhibition and/or P-glycoprotein. Whether a similar effect on pharmacokinetics occurs using saquinavir soft gelatin capsules remains to be established.

We previously observed an increase in plasma concentrations and AUC of saquinavir when itraconazole was administered to patients who also took saquinavir [7]. However this increase (400%) was substantially larger than that caused by fluconazole (50%). These in vivo data confirm the in vitro results that showed itraconazole to be a more potent inhibitor of cytochrome P450 3A4 activity than fluconazole [3]. While the interaction between itraconazole and saquinavir might be used in clinical practice to increase saquinavir concentrations, addition of fluconazole to a saquinavir-containing regimen may not be effective, as the median increase in plasma concentrations does not exceed the intrinsic interindividual variability of these parameters.

The metabolism of ritonavir is generally little affected by cytochrome P450 3A4 inhibitors because of its high affinity for cytochrome P450 3A4 [3]. In this study no significant influence of fluconazole on ritonavir pharmacokinetics was observed. These findings are at variance with data that show a modest, but significant, increase in ritonovir AUC and C_max (12–15%) during fluconazole administration [5, 8, 17]. Given the large interpatient variability, our small number of patients is unlikely to have sufficient power to exclude any differences. However, the clinical significance of this increase was judged minor and should not require dose adjustments.

The use of antifungal drugs is common in HIV-1-infected patients because of the high incidence of infections with fungi and yeast in this population. Concomitant use of antifungals and protease inhibitors can potentially lead to altered pharmacokinetic parameters due to drug–drug interactions. The results of this study have shown that fluconazole, a moderately potent inhibitor of cytochrome P450 3A4, increases saquinavir exposure by 50%, while exposure to ritonavir is largely unchanged. It is unlikely that dose adjustments for saquinavir or ritonavir need to be made if fluconazole is used concomitantly. This is also likely to be applicable when fluconazole is added to a regimen in which ritonavir and saquinavir are combined.