INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Ritonavir is a highly potent human immunodeficiency virus (HIV) protease inhibitor (15) that leads to exponential decreases in plasma viral RNA within a few days after administration (9, 13, 22). The HIV reverse transcriptase inhibitor zidovudine (ZDV) also delays the progression of disease in patients infected with HIV (11). The development of viral resistance to antiretroviral drugs is more likely to occur when a single antiretroviral drug is administered alone (19, 21, 28, 29). Thus, in an attempt to reduce the incidence of the development of HIV variants with resistance and improve the safety and efficacy profile, a combination of drugs of different classes (e.g., ritonavir and ZDV) and possibly other drugs within a class (i.e., triple-drug therapy) may be administered concurrently.

Treatment of HIV infection with a multidrug regimen introduces the possibility of drug-drug interactions. Ritonavir is metabolized extensively by the cytochrome P450 (CYP) system, particularly CYP3A, and ritonavir is a potent inhibitor of CYP3A (17). In addition, there is evidence that ritonavir induces glucuronidation in rats (18) and humans (25). ZDV is excreted to a minor extent (mean of 14%) as unchanged drug in the urine, while on average, 75% of a dose is excreted in the urine as the 5' glucuronide metabolite (ZDV-G) (1). A small portion of ZDV is metabolized to 3'-amino-3'-deoxythymidine (AMT), which is more toxic than ZDV (8). Both cytochrome P450 reductase and cytochrome P450 are involved in the reduction of ZDV to AMT (5, 8, 26), and the formation of AMT is inhibited by ketoconazole (10), a potent inhibitor of CYP3A (24). Because of the potential ritonavir-induced changes in cytochrome P450 activity and glucuronidation, the pharmacokinetics of both ritonavir and ZDV were investigated when the drugs were administered alone and concurrently.

(This study was presented in part at the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, Calif., 17 to 20 September 1995.)

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Patients. The 18 patients enrolled in this study were asymptomatic, HIV-positive men. Patients were otherwise healthy based upon the results of a medical history, physical exam, ophthalmologic exam, laboratory profile, electrocardiogram, and a negative hepatitis B antigen test and had no recent history of drug or alcohol abuse. Patients were excluded from study participation if they had a CD4 lymphocyte count of <300/mm³, had received any investigational drug within the 4-week period before the initial drug administration in this study, or were using any other drug that could not be discontinued, including over-the-counter medications, at least 2 weeks before the initial drug dose was given in the present study. All patients gave written, informed consent in compliance with Food and Drug Administration regulations, and approval was obtained from the institutional review board.

Study design. This was a single-center, multiple-dose, open-label, three-period complete crossover study. Patients were assigned randomly in equal numbers to six different sequences of the following three regimens: ZDV (200 mg every 8 h [q8h]) alone for 4 days, ritonavir (300 mg q6h) alone for 4 days, and ZDV (200 mg q8h) with ritonavir (300 mg q6h) for 4 days. Patients received doses on days 1 to 5 of each period, with the morning ritonavir and ZDV doses administered at approximately 6:30 and 8:30 a.m., respectively. The final dose on day 5 was administered at approximately 12:30 a.m. A solution formulation of ritonavir was used for this study because the capsule formulation had not been developed. For ZDV, Retrovir capsules (100 mg/capsule; Glaxo Wellcome Company) were administered. Ritonavir doses were administered within 10 min after a meal or snack; ZDV was administered between meals. All doses were taken with approximately 200 ml of water. The study was conducted during three confinement periods, each lasting approximately 6 days and separated by a washout interval of at least 9 days during which neither drug was administered. Blood samples (5 ml each) were collected beginning on day 4 at approximately 6:30 a.m. (within 5 min prior to the morning ritonavir dose [time 0] and/or 2 h prior to the morning ZDV dose) and 1, 2, 2.5, 3, 4, 5, 6, 7, 8, 9, 10, 10.5, 11, 12, 13, 14, 15, 16, 18, 19, 20, 21, 22, and 24 h later. Administration of ritonavir or ZDV continued during the sample collection period (Table 1). Samples were obtained immediately prior to dosing when sampling and dosing times coincided. The plasma was harvested and stored frozen at 20°C or colder in appropriately labeled tubes until assayed for ritonavir or ZDV and ZDV-G. Additionally, AMT plasma concentrations of selected samples were measured.

Drug analyses. Ritonavir samples were analyzed by a validated high-pressure liquid chromatography (HPLC) method (23). Briefly, samples were extracted by liquid-liquid extraction with ethyl acetate:hexane (9:1 [vol/vol]), the organic phase was evaporated to dryness, and the sample was reconstituted and washed twice with hexane. An aliquot of the reconstituted extract was then analyzed by reverse-phase HPLC with UV detection at 205 nm. Calibration standards (10 to 15,000 ng/ml) and quality control samples (150, 7,500, and 12,000 ng/ml) had coefficients of variation of 14.1% or lower and ranged in accuracy from 91.7 to 104%. The lower limit of quantification (LLQ) was 10 ng/ml. Samples with ritonavir concentrations greater than 15,000 ng/ml were diluted and reanalyzed. Samples for ZDV and ZDV-G concentrations were analyzed with a validated competitive direct equilibrium radioimmunoassay with the ZDV-Trac ¹²⁵I radioimmunoassay kit (Incstar Corporation, Stillwater, Minn.). Calibration standards ranged from 15 to 3,000 ng/ml for ZDV and 25 to 2,000 ng/ml for ZDV-G. Quality control samples (15, 25, 500, and 2,000 ng/ml for ZDV; 25, 70, 500, and 2,000 ng/ml for ZDV-G) had coefficients of variation of 15.3% or lower and ranged in accuracy from 85.9 to 102%. ZDV concentrations after hydrolysis with -glucuronidase (ZDV_h) were used to estimate ZDV-G concentrations according to the following equation: where the multiplier converts the molecular weight of ZDV to that of ZDV-G. The LLQs were 15 ng/ml for ZDV and 25 ng/ml for ZDV-G. Samples for AMT concentrations were analyzed by a previously published HPLC assay with fluorescence detection (32). AMT calibration standards ranged from 3 to 75 or 3 to 100 ng/ml. Estimated concentrations of quality control samples were similar to theoretical values of 7 and 30 ng/ml, ranging from 7.1 to 7.8 and 27.1 to 29.6 ng/ml, respectively. The LLQ and limit of detection were 3.0 and 0.5 ng/ml, respectively.

Pharmacokinetic and statistical analyses. All pharmacokinetic parameters were calculated by noncompartmental methods for a 24-h period beginning at approximately 6:30 a.m. on the morning of day 4 of each period. The time to reach the observed maximum concentration (T_max) for individual dose intervals was obtained directly from the plasma concentration-time data. For the 24-h period, T_max was calculated as the mean T_max of individual dose intervals. Maximum and minimum observed concentrations (C_max and C_min, respectively) were obtained directly from the plasma concentration-time data for the 0- to 24-h interval and for each individual dose interval. Area under the plasma concentration-time curve (AUC) was calculated by the linear trapezoidal method. For ritonavir, ZDV, and ZDV-G, the AUC for the 24-h interval (AUC_0-24) was calculated beginning from time 0 (immediately prior to the morning ritonavir dose). For each dose interval completely sampled (all intervals except the last ZDV/ZDV-G interval), the concentration-time data used to calculate AUC ranged from the values obtained immediately prior to the dose to those obtained immediately prior to the subsequent dose. Assuming ZDV and ZDV-G had reached steady-state concentrations by the fourth day of dosing, the AUC of the last ZDV and ZDV-G dose interval was calculated as AUC_18-24 + AUC_0-2. The apparent oral clearance (CL/F) for individual dose intervals was estimated as the quotient of dose and AUC. The average concentration for the 24-h interval (C_avg) and the 24-h CL/F were estimated as the quotient of AUC_0-24 and 24 h and the quotient of the total daily dose and AUC_0-24, respectively. For ZDV,  (elimination phase) was estimated for each dose of the 24-h interval as the negative of the slope of the straight line obtained by regression of the logarithms of the concentrations versus time in the log-linear terminal phase of the curve. All regressions were based on the measurable concentrations from C_max to the sample collected immediately prior to the subsequent dose. For ritonavir, no log-linear terminal phase could be observed with this dosage regimen. ZDV half-life (t_1/2) was calculated as ln(2)/, and the harmonic mean and pseudo standard deviation t_1/2 values are reported.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The mean ± standard deviation (range) age and body weight of the 18 men enrolled were 37.4 ± 6.8 (28 to 49) years and 77.8 ± 11.1 (61.2 to 109.8) kg, respectively. One patient was removed from the study because of noncompliance (unable to eat several meals). Six patients were withdrawn from the study during the ritonavir alone or combination regimens due to various adverse events, most commonly nausea (n = 3) and paresthesia (n = 2). Three additional patients had documented or apparent dosage interruptions that affected the ZDV and ZDV-G concentration-time profiles during the ZDV alone or combination regimen. For ritonavir, only one patient clearly had a dosage interruption, and this patient was excluded from analysis of the ritonavir 24-h interval and the third and fourth dose intervals. The concentration-time profiles of two other patients (very low concentrations) suggested that some doses were missed during both periods that ritonavir was administered. However, because dosage interruptions were not as demonstrable for these two patients (i.e., no clear log-linear decline in concentrations over an entire dose interval), they were not excluded from analysis. Thus, ritonavir comparisons are based on the data for 10 or 11 patients, and ZDV/ZDV-G comparisons are based on the data for 9 or 11 patients (depending on the dose interval).

Ritonavir mean 24-h pharmacokinetic parameters (Table 2) and plasma concentration-time profiles of ritonavir (Fig. 1) reflect minimal differences in concentrations between regimens. Ritonavir C_max, C_min, and AUC mean values were higher for the first dose interval than for those of other intervals, although the actual difference in any parameter between dose intervals was relatively minor (Table 3). Similarly, differences in mean C_max and AUC due to the addition of ZDV to the regimen of ritonavir alone were 8% or less for all dose intervals (C_max increases, 5, 5, 8, and 6% and AUC differences, 0, 2, 8, and 5% for the four dose intervals). Differences in ritonavir T_max, C_min, C_max and AUC with coadministration compared to administration of ritonavir alone were not statistically significant (P > 0.09). Thus, ritonavir pharmacokinetics were unaffected by coadministration with ZDV.

View larger version (11K): [in this window] [in a new window]   FIG. 1.   Mean ritonavir plasma concentrations after administration of ritonavir (300 mg q6h) alone (circles) or in combination (squares) with ZDV (200 mg q8h).

For ZDV, mean 24-h pharmacokinetic parameters (Table 4) and plasma concentration-time profiles of ZDV (Fig. 2) indicate small but statistically significant differences in concentrations between regimens. For individual dose intervals, there was a trend of decreasing C_max and AUC mean values within each regimen (Table 5). However, changes in ZDV C_max and AUC due to the addition of ritonavir (decreases of 22 to 33%) compared to ZDV administered alone were not obviously related to the dose interval. Differences between regimens in ZDV 0- to 24-h C_max and AUC_0-24 were similar to those of the individual dose intervals, approximately a 26% reduction in parameter mean values with coadministration compared to ZDV alone (P < 0.03). Because of the minimal amount of AUC contained in the 6- to 8-h postdose portion of the dosage interval, the method of calculation of AUC_0-24 had no effect on the estimates; AUC_18-24 + AUC_0-2 values were similar to those calculated as AUC_18-24 + AUC_24-26, with AUC_24-26 obtained by extrapolation with . The 24-h AUC was the primary statistical parameter; thus, the AUC based on the actual data collected was reported. ZDV AUC and C_max were statistically significantly different between regimens for all dose intervals (P < 0.05). There were no statistically significant differences between regimens in ZDV mean T_max (P = 0.19) or in T_max of the first and third dose intervals (P > 0.42), but there was a statistically significant difference in T_max of the second dose interval (P = 0.047). Coadministration with ritonavir had no effect on ZDV  (P > 0.55 for each dose interval).

View larger version (19K): [in this window] [in a new window]   FIG. 2.   Mean ZDV plasma concentrations after administration of ZDV (200 mg q8h) alone (circles) or in combination (squares) with ritonavir (300 mg q6h).

Similar to ZDV C_max and AUC, there was a trend of decreasing ZDV-G C_max and AUC mean values of individual dose intervals within each regimen, and differences between regimens were not obviously related to the dose interval (Table 5). In contrast to the statistically significant changes in ZDV AUC_0-24, ZDV-G AUC_0-24 with the combination was similar to that for ZDV administration alone (P = 0.64). In fact, there were no statistically significant differences between regimens for any ZDV-G pharmacokinetic parameters (P > 0.13; Tables 4 and 5), and plasma concentration-time profiles of ZDV-G were similar between regimens (Fig. 3).

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Mean ZDV-G plasma concentrations after administration of ZDV (200 mg q8h) alone (circles) or in combination (squares) with ritonavir (300 mg q6h).

Of the samples assayed for AMT, there were no apparent differences in AMT concentrations between the two regimens (data not shown). Of the 47 samples analyzed for AMT, 39 had AMT concentrations below the LLQ, with the majority of these below the lower limit of detection, regardless of whether ZDV was dosed alone or in combination with ritonavir. The highest concentration of AMT was 7.5 ng/ml for ZDV administered alone and 7.6 ng/ml for the combination regimen. Because most of the samples had AMT concentrations below the LLQ, pharmacokinetic and statistical analyses of AMT concentrations could not be performed.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Although ZDV therapy prolongs survival, there is a higher incidence of the development of viral resistance when the drug is administered alone (19, 21, 27-30) through a variety of potential mechanisms (12). Thus, the prevention of the development of resistant viruses by the use of combination regimens of drugs of different classes (e.g., reverse transcriptase inhibitors and protease inhibitors) is paramount. However, with multidrug regimens, there is an increase in the potential for drug-drug interactions causing altered pharmacokinetics of either drug, possibly leading to an increase in adverse events or a reduction in efficacy.

Of the 18 patients enrolled in the present study, 6 did not complete at least one regimen because of adverse events, and several others had dosage interruptions during at least one sample collection period. It is possible that the adverse events that caused five patients to be unable to complete the combination regimen were due to a pharmacokinetic interaction. However, of the adverse effects experienced by the six patients who did not complete the study, those associated with the combination regimen do not seem to have differed much from those associated with the administration of ritonavir or ZDV alone. Thus, it is unlikely that these patients were unable to complete the study due to a drug-drug interaction leading to high concentrations of ZDV or ritonavir.

The addition of ZDV to the regimen of ritonavir alone had little if any effect on ritonavir pharmacokinetics. Ritonavir C_max and AUC for each dosage interval differed by 8% or less between regimens. Ritonavir T_max also was unaffected by the addition of ZDV to the regimen and was approximately 4 h for both regimens, similar to that of recent studies (2, 3). Also consistent with recent studies (2, 3), the maximum values of C_max and AUC generally occurred after the morning dose regardless of regimen, suggesting that absorption of ritonavir may proceed more efficiently during the morning dose interval (i.e., diurnal variation of absorption).

In contrast to the lack of effect of ZDV on ritonavir pharmacokinetics, ZDV C_max and AUC decreased by about 26% in the presence of ritonavir. The small differences in ZDV-G C_max and AUC were not statistically significant. In general, ZDV is rapidly and completely absorbed, but first-pass metabolism reduces the absolute bioavailability to about 60%. Maximum plasma concentrations of ZDV are achieved within about 1 h, and elimination is rapid, with a mean half-life of approximately 1.1 to 1.5 h (1, 4, 16, 20), similar to the mean of 1.1 h in the present study. Plasma concentrations of ZDV-G often exceed those of ZDV, and parallel decline of plasma concentration-time curves for ZDV and ZDV-G suggest formation-rate limited elimination of ZDV-G.

Although about 90% of a dose of ZDV is excreted in the urine as unchanged drug or ZDV-G, a second metabolite, AMT, has been identified in studies with human liver (8), human gastrointestinal bacteria (14), and rat liver microsomes and hepatocytes (7, 8) and in the plasma of rhesus monkeys (6). In addition, AMT has been detected in the plasma (but not bile) of patients receiving 2.5 mg of ZDV per kg by a 1-h intravenous infusion (31). There is evidence of the involvement of both cytochrome P450 reductase and cytochrome P450 in the reduction of ZDV to AMT (5, 8, 10, 26). A wide range of substrates and inhibitors of different cytochrome P450 isozymes were investigated in human liver microsomes. The most marked inhibition of the reduction of ZDV to AMT occurred with ketoconazole (10), a potent inhibitor of CYP3A (24). Induction in rats of CYPs 2B, 3A, and 4A resulted in increased hepatic microsomal formation of AMT and increases in the intrinsic clearance of approximately 1.6- to 3-fold (10). Although AMT is a relatively minor metabolite of ZDV, AMT is five- to sevenfold more toxic than ZDV in several in vitro tests (8), suggesting that alterations in the formation of AMT could affect the adverse event profile of ZDV.

The combination of ZDV and ritonavir caused a decrease in ZDV C_max and AUC compared to those for ZDV administered alone, suggesting that ritonavir induced the metabolism of ZDV. Although there was minimal change in ZDV-G pharmacokinetics, and ritonavir is a potent CYP3A inhibitor, the effect on the formation of AMT was unknown. Therefore, several samples for AMT analysis were selected in an attempt to provide a representative subset of all samples collected during the study (i.e., samples from each dose interval, samples with typical or high ZDV concentrations, and samples during the time of peak ZDV concentrations for both regimens).

AMT mean peak plasma concentrations of approximately 160 ng/ml (n = 6) were detected in the plasma of patients receiving 2.5 mg of ZDV per kg by intravenous infusion over 1 h (31) (a dose resulting in approximately 50% greater exposure of ZDV than the dose used in the present study). Assuming absolute bioavailability of about 60%, AMT peak concentrations of approximately 60 ng/ml could occur with the dosage regimen used in the present study. However, for both regimens, AMT concentrations were quite low. Most of the samples tested had AMT concentrations below the LLQ, with the majority of these below the lower limit of detection, regardless of whether ZDV was dosed alone or in combination with ritonavir. Thus, ritonavir had essentially no effect on the conversion of ZDV to AMT by CYP3A, and toxicity associated with high plasma concentrations of AMT did not occur in the presence of ritonavir.

The decrease in ZDV AUC by approximately 25% with no apparent change in ZDV-G AUC or AMT concentrations is not easily explained. A decrease in absorption could account for the change in ZDV, but a corresponding decrease in metabolite AUC would be expected. Rather than changes in absorption, enhanced ZDV metabolism by induction of glucuronyl transferase could have caused lower ZDV concentrations and relatively little change in metabolite AUC, as slightly more ZDV would be converted to ZDV-G rather than eliminated unchanged in the urine. However, enhanced ZDV metabolism would be expected to have a corresponding effect on ZDV , which was not observed. Thus the reduction in ZDV concentrations when administered concurrently with ritonavir are not completely explained by either reduced absorption (ZDV-G concentrations were unchanged) or by increased metabolism (ZDV  was unchanged). Also, if glucuronidation of ZDV was induced by ritonavir, the full extent of induction may not have been reached after 4 days of dosing and the effect on ZDV concentrations may have been underestimated in this study. Regardless of the actual mechanism(s) involved, plasma concentrations of ZDV were reduced by about 25% when administered concurrently with ritonavir, but ZDV-G, AMT, and ritonavir concentrations were similar between regimens.

In conclusion, ritonavir pharmacokinetics were not influenced by the addition of ZDV to the regimen of ritonavir alone, and dosage adjustment of ritonavir is unnecessary when it is administered concurrently with ZDV. In contrast, ZDV exposure apparently was reduced when it was coadministered with ritonavir; ZDV 0- to 24-h C_max and AUC_0-24 decreased 27 and 26%, respectively, with coadministration. However, there was little or no apparent difference in ZDV-G or AMT plasma concentrations between regimens. The mechanism of the interaction between ritonavir and ZDV is unclear. No differences in adverse events between the combination and single-drug regimens were apparent. The clinical relevance of a 26% reduction in ZDV exposure when the drug is administered with ritonavir is unknown; however, the long-term safety and efficacy of coadministration of ZDV and ritonavir continues to be evaluated clinically.