Pharmacokinetics of Indinavir and Ritonavir Administered at 667 and 100 Milligrams, Respectively, Every 12 Hours Compared with Indinavir Administered at 800 Milligrams Every 8 Hours in Human Immunodeficiency Virus-Infected Patients

Abstract

Human immunodeficiency virus (HIV) patients on nucleoside or nucleotide reverse transcriptase inhibitors with HIV RNA at <1,000 copies/ml were randomized in an open-label study to administration of combined indinavir/ritonavir (IDV/RTV) at 667/100 mg every 12 h (q12h) or IDV alone at 800 mg q8h to determine the regimens' pharmacokinetics. On day 14, plasma IDV and RTV levels were determined over 24 h. Noncompartmental pharmacokinetics (minimum concentration of drug in serum [C_min], area under the concentration-time curve from 0 to 24 h [AUC_0-24], and maximum concentration of drug in serum [C_max]) were expressed as geometric mean values with 90% confidence intervals (CI). The primary hypothesis was that the lower bound of the protocol-specified 90% CI for the geometric mean C_min ratio of the combination compared to IDV alone regimen would be ≥2. Twenty-seven patients were enrolled, and 24 (15 male; average age, 42 years) completed the study. The C_min, AUC_0-24, and C_max for IDV/RTV compared to IDV alone were 1,511 versus 250 nM, 119,557 versus 77,034 nM · h, and 10,428 versus 10,407 nM, respectively. Corresponding relationships for IDV/RTV compared to IDV alone were a 6.0-fold increase in C_min (90% CI, 4.0, 9.3), an increase in AUC_0-24 (1.5-fold, 90% CI, 1.2, 2.0), and no increase in C_max. Adverse events were similar and generally mild, with no cases of nephrolithiasis. The geometric mean ratio of IDV C_min for IDV/RTV compared to IDV was at least 2 by a lower bound of the 90% CI, satisfying the primary hypothesis. The C_max was not increased, suggesting an IDV/RTV 667/100-mg toxicity profile may be similar to that of unboosted IDV.

Single-protease inhibitor regimens significantly reduced the morbidity and mortality associated with human immunodeficiency virus (HIV) following their introduction (19, 20). More recently, boosted protease inhibitor regimens combined ritonavir (RTV) with a second protease inhibitor to achieve higher sustained levels of the second protease inhibitor than seen when it was given as part of a single-protease inhibitor regimen. The boosted protease inhibitor regimens may be more effective in suppressing HIV strains with resistance to antiretroviral therapy, including resistance to the second protease inhibitor (2, 9, 10, 18, 22, 32). In addition, boosted protease inhibitor regimens may be administered less frequently and without regard to meals, potentially improving patient adherence (2, 9, 10, 18, 22, 32). RTV's inhibition of the cytochrome P-450 CYP3A4 enzyme, the primary enzyme in the metabolism of most protease inhibitors, changes the pharmacokinetics of the second protease inhibitor, with elevations of minimum drug concentration (C_min) and area under the concentration-time curve (AUC) contributing to improved clinical efficacy (2, 10). In addition, RTV's inhibition of P-glycoprotein transport proteins may also contribute to increased concentration of a concomitantly administered protease inhibitor in plasma (33).

Indinavir (IDV) is approved for administration every 8 h (q8h) at 800 mg without food (28). Combinations of IDV at 400 to 800 mg and RTV at 100 to 400 mg twice a day (b.i.d.) demonstrated at least equivalent to (and in many cases much higher than) the C_min and AUC at 24 h (AUC₂₄) of IDV at 800 mg q8h in normal volunteers and can be administered with food (25). The trough levels maintained with the b.i.d. combinations were above the concentration necessary to inhibit 95% of viral growth of wild-type HIV-1 seen in the absence of drug (8). The most-studied boosted IDV regimen is IDV at 800 mg plus RTV at 100 mg q12h (4, 7, 21, 31). A combination regimen of IDV/RTV at 400/400 mg provided suppression of HIV RNA to undetectable levels (<80 copies/ml) in 59.5% of HIV-infected antiretroviral-naïve individuals, but patients presented with side effects characteristic of RTV, as well as IDV (17). Pharmacokinetic and clinical data with IDV/RTV at 400/100 mg appear comparable to those seen with the IDV 800-mg q8h regimen, but no direct comparison has been performed (5, 6, 12, 15). There are no pharmacokinetic studies that directly compare a regimen of IDV/RTV q12h, which includes an IDV dose of <800 mg and low-dose RTV (100 mg), directly with the standard IDV 800-mg q8h regimen.

In the present study, pharmacokinetic parameters (C_min, C_max, and AUC from 0 to 24 h [AUC_0-24]) of a lower-dose regimen consisting of IDV/RTV at 667/100 mg q12h were compared to those of the standard IDV 800-mg q8h regimen. The 667-mg IDV q12h regimen offers greater convenience and potentially better adherence than the IDV 800-mg q8h regimen due to less-frequent dosing. The primary objective was to compare the C_min of the IDV/RTV 667/100-mg regimen to that of the IDV 800-mg q8h regimen, as well as fully characterize the pharmacokinetics of the lower-dose regimen. The tolerability of both regimens was also evaluated as a secondary study endpoint.

MATERIALS AND METHODS

Study design.

This was a randomized, open-label, parallel group, 14-day study comparing regimens containing IDV (Merck & Co., Inc.) in combination with RTV (Abbott Pharmaceuticals) or IDV alone. HIV-1-seropositive men and nonpregnant women at least 18 years of age were enrolled at two study sites. The study was conducted with HIV-positive patients to provide the closest correlation to the treatment setting (3, 11). The patients needed to have HIV suppressed to <1,000 copies/ml on nucleoside or nucleotide therapy to allow for introduction of protease inhibitor therapy without disadvantaging the subjects. IDV pharmacokinetics appear to differ between patients with detectable and undetectable plasma HIV RNA. The ability to make a one-drug perturbation can be done safely only in the patient with fully suppressed HIV. One site was an outpatient clinic, which required patients to return to the site for blood sample collection on predefined days. These patients returned for their last evening dose on day 13, remained through 24-h blood collection on day 14, and could leave after their C_{24 h} level was collected on the morning of day 15. The second clinic housed patients at the site for days 1 to 14. Eligible patients were on a dual- or triple-nucleoside and/or nucleotide reverse transcriptase regimen, which did not contain protease inhibitors or nonnucleoside reverse transcriptase inhibitors. Their viral load was well controlled, and their CD4 cell count was ≥100 cells/mm³ within 45 days of study entry. Patients were excluded who had known hypersensitivity or allergy to IDV or RTV. Patients were also excluded who required concomitant medication or had a concurrent condition that was contraindicated with use of the study medications or required concomitant medication or had a concurrent condition that could interfere with the evaluation of the study medications' pharmacokinetics.

Patients were randomized equally between two treatment groups: patients in group 1 received IDV/RTV at 667/100 mg q12h; patients in group 2 received IDV at 800 mg q8h. Patients in group 1 received two 333-mg capsules of IDV and one 100-mg capsule of RTV q12h. Patients in group 2 received two 400-mg capsules of IDV q8h. Patients and study site staff were blinded to treatment allocation assignment until after enrollment. Allocation was performed with a computer-generated allocation schedule. At one study site, randomization took place in the pharmacy out of view of the study staff, while at the second study site, the study staff used sealed envelopes to maintain blinding until eligibility was confirmed. After assignment was determined, study medications were administered in an open-label fashion. The medications were added to the patients' prior nucleoside/nucleotide reverse transcriptase inhibitor regimens. Study medication was administered on days 1 to 14. Patients were seen in the clinic for screening and randomization and had follow-up visits with the investigator on days 7, 12, 13, 14, and 15. Blood was drawn for trough drug levels in the clinic on days 1, 12, 13, and 14. The evening dose and meal on day 13 were to be observed by the study site staff. On day 14, patients remained in the clinical research facility for 24 h, all dosing was observed, and multiple blood draws were performed for drug levels after each dose of drug. Blood draws extended into day 15. Patients had a final clinical assessment 14 days after completion of the study.

Patients in group 1 took both doses of study medication on days 1 to 10 and their a.m. dose on days 11 to 13 with or without food. The p.m. doses on days 11 to 13 and both doses on day 14 were administered with water and within 5 min after the consumption of a muffin and 6 oz of skim milk. This was done to confirm that the patients had the same meal and that there was not a food effect with the IDV/RTV regimen.

Patients in group 2 took all their doses of study medication on days 1 to 10 and their a.m. doses on days 11 to 13 1 h before or 2 h after a large meal or with a predefined low-fat meal. The p.m. doses on days 11 to 13 and all doses on day 14 were administered with a Quaker low-fat chewy granola bar. Patients in group 2 did not consume any other food (except water) within 2 h prior to or 1 h after receiving their p.m. dose of study drug on days 11 to 13 and all doses on day 14. This was done to be consistent with the food alternatives listed in the IDV prescribing information.

All patients were instructed to maintain vigorous hydration while receiving study medication (six 8-oz glasses of water throughout the day). Patients were allowed to continue taking their prior concomitant medications. Any additional medications were discouraged, and the following were specifically excluded during the study: sidenafil, St. John's wort, HMG-coenzyme A reducatase inhibitors, and any drug which utilizes the CYP3A4 pathway. In addition, females were not allowed to use oral contraceptives. Patients who smoked were allowed to do so during the study.

Pharmacokinetic sampling and analysis.

On day 13, patients in groups 1 and 2 were seen at the clinic site to start collection of blood samples for pharmacokinetic analyses of IDV and RTV. The site staff collected 8-ml samples of blood in the p.m. of day 13 as a trough and on day 14 at prespecified time points of 0 (a.m. only), 0.5, 1, 1.5, 2, 3, 4, 6, 8, and 12 (group 1 patients only) h postdose. Blood samples were packaged and transported to Covance Bioanalytical Services, LLC (Indianapolis, Ind.), for analysis. Plasma samples were assayed for IDV and RTV concentration. IDV, RTV, and an internal standard, reserpine, were extracted from human plasma by protein precipitation. After dry down under nitrogen, the residue was reconstituted and analyzed by liquid chromatography with tandem mass spectrometric detection. The lower limit of quantitation was 10.0 nM for both IDV and RTV. For IDV, intraday and interday precision ranged from 2.0 to 8.6% and 6.5 to 13.8%, respectively. For RTV, intraday and interday precision ranged from 3.3 to 9.3% and 8.5 to 13.9%, respectively. IDV concentrations were converted to a molar basis with a molecular weight of 613.81. Noncompartmental pharmacokinetic parameters were calculated and are expressed as geometric mean values with protocol-specified 90% confidence intervals (CI).

Pharmacokinetic parameters and safety measures.

The pharmacokinetic endpoints in this study were C_min, the steady-state trough plasma IDV concentration of the dosing interval; C_max, the maximum plasma IDV concentration during the dosing interval; and AUC_0-24, the IDV AUC from 0 to 24 h. The protocol-specified endpoint was an average over the 24-h dosing interval because we did not have sufficient grounds to suspect diurnal variation. An average of the natural log C_min was used: that is, the natural log C_min for IDV at 800 mg q8h was defined as the mean value of natural log (C_{8 h}, C_{16 h}, and C_{24 h}) on day 14 and natural log C_min for IDV at 667 mg plus RTV at 100 mg q12h was defined as the mean value of C_{12 h} and C_{24 h} on day 14. The C_min, C_max, and AUC_0-24 were also calculated for RTV on day 14.

Clinical and laboratory adverse events were recorded throughout the study for analyses of the general safety and tolerability of the IDV/RTV combination as a second primary endpoint. Laboratory values (serum chemistries, hematology, and urinalysis) were also recorded at baseline and day 14, including a test for pregnancy of all female participants.

Statistical analysis.

The primary study hypothesis was after 2 weeks of administration of IDV/RTV at 667/100 mg q12h or IDV at 800 mg q8h, the C_min of IDV with the combination regimen would be at least twofold greater than that achieved with the IDV-alone regimen. Power was calculated based on this hypothesis. Determination of differences between treatment groups for pharmacokinetic endpoints was based upon comparisons of geometric mean ratios of 90% CI. Assuming the true geometric mean ratio of the actual pharmacokinetic values was 6, a sample of 12 patients in each regimen would yield a 90% probability that the lower bound of the two-sided 90% CI for the geometric mean ratio would be >2 (16). This calculation is based on the standard deviation (SD) of 0.89 for C_min (on the log scale) (16). Because there was only one primary hypothesis in the study, no multiplicity adjustment was needed.

An analysis of variance (ANOVA) model was used to analyze the C_min data on the natural log scale to estimate and test the difference in natural log C_min between the groups. The standard error from the ANOVA was used to construct a 90% CI for the difference of natural log (C_min). A CI for the geometric mean ratio of C_min (IDV/RTV at 667/100 mg q12h versus IDV at 800 mg q8h) was obtained by exponentiating the CI for the difference of natural log (C_min). If the lower bound of the two-sided 90% CI for the geometric mean ratio was at least 2, then the hypothesis was satisfied.

A similar approach was used to analyze C_max and AUC_0-24. In addition to calculations of geometric mean ratios, geometric mean and 90% CI summary statistics were calculated for C_min, C_max, and AUC_0-24. Mean concentration-over-time curves for both study medications were constructed.

All relevant pharmacokinetic parameters for RTV were summarized (geometric means and 90% CI) within dose regimens (C_min, C_max, and AUC_0-24).

Adverse experiences, vital signs, and laboratory safety data were tabulated, and total cholesterol, serum triglycerides, and total serum bilirubin levels were analyzed.

RESULTS

Study patients.

Twenty-seven patients were enrolled in the study. A total of three patients discontinued participation prior to completing the study. Two patients discontinued in the IDV/RTV group (one experienced a clinical adverse event [fever] on day 11, and one withdrew consent on day 4), and one patient discontinued in the IDV-alone group: one laboratory adverse event (elevation of blood urea nitrogen [BUN] and creatinine on day 9) prior to day 14. The fever leading to the withdrawal of the IDV/RTV patient was assessed as not drug related by the investigator. The elevation of BUN leading to the withdrawal of the IDV-alone patient was assessed as possibly drug related by the investigator.

Twenty-four patients (12 per treatment arm) had evaluable data on day 14. These patients had a mean age of 42 years, 63% were male, 50% were black, 25% were white, 21% were Hispanic American, and 4% were of other ethnicities. The weights of the patients in each treatment group were similar, with means of 186 (SD, 29.3) and 181 (SD, 62) lb in groups 1 and 2, respectively.

Pharmacokinetic profile.

Mean IDV concentration-over-time curves are provided for 24 h on day 14 for both IDV/RTV and IDV alone (Fig. 1). As illustrated by Fig. 1, the mean IDV level for the IDV/RTV regimen was maintained above 1,000 nM throughout the 24-h dosing cycle, well above minimum levels maintained with the IDV-alone regimen.

FIG. 1.

IDV mean concentration log plot over time.

The morning IDV trough levels on days 12, 13, and 14 for both regimens are provided in Fig. 2. The day 12, 13, and 14 values were relatively similar within each treatment group. Thus, these values were important in confirming the achievement of steady state by the time the day 14 analyses were performed.

FIG. 2.

Individual trough plasma IDV concentrations and geometric means (90% CI) of morning doses on days 12, 13, and 14 (log plot).

Primary hypothesis.

The IDV C_min ratio (IDV/RTV at 667/100 mg q12h compared to IDV at 800 mg q8h) was 6.05 and was statistically significant (P < 0.001) (Table 1). The lower bound of the 90% CI of the IDV C_min ratio of 3.96 was >2-fold, satisfying the requirements of the primary hypothesis. The geometric means of IDV C_min based on the average across the dosing intervals were 1,511 nM (90% CI, 1,119, 2,039) for the IDV/RTV 667/100-mg q12h regimen and 250 nM (90% CI, 185, 337) for the IDV 800-mg q8h regimen (Table 1; Fig. 3A). Geometric mean IDV C_min values were calculated from predose concentrations on day 14 (from 0 to 24 h), and those for IDV/RTV (prior to 12- and 24-h doses) and IDV alone (prior to 8-, 16-, and 24-h doses) were calculated and are presented in Fig. 3A. Geometric mean IDV C_min levels for both the 12- and 24-h time points with IDV/RTV were several fold higher than corresponding mean levels for any of the IDV-alone time points. The geometric mean IDV C_min level appeared higher for the 24-h dose, obtained at 8 a.m., than the 12-h dose, obtained at 8 p.m.

TABLE 1.

Comparison of IDV pharmacokinetic parameters between treatment groups on day 14

Parameter (n = 12)a	Result for:						P value
	IDV/RTV at 667/100 q12h		IDV at 800 q8h		Between-treatment comparison of IDV/RTV vs IDV
	Geometric mean	90% CIb	Geometric mean	90% CIb	Geometric mean ratio	90% CIb
Cmin (nM)	1,511	1,119, 2,039	250	185, 337	6.05	3.96, 9.25	<0.001
Cmax (nM)	10,428	9,209, 11,808	10,407	9,190, 11,784	1.00	0.84, 1.19	0.985
AUC0-24 (nM · h)	119,557	98,476, 145,150	77,034	63,451, 93,525	1.55	1.18, 2.04	0.012

^aAverage natural log (parameter) of dosing intervals was used in an ANOVA model.

^b90% CI for geometric mean are based on the ANOVA model.

FIG. 3.

Mean change in IDV pharmacokinetic parameters. (A) Individual C_min plasma IDV concentrations (average and by dosing interval) and geometric means (90% CI) on day 14 (log plot). (B) Individual C_max plasma IDV concentrations (average and by dosing interval) and geometric means (90% CI) on day 14 (log plot). For panels A and B, the average was obtained by first computing the average of the natural log (C_{12 h} and C_{24 h}) for IDV/RTV q12h and the average of the log (C_{8 h}, C_{16 h}, and C_{24 h}) for IDV q8h and then reexpressing the results in the original units via exponentiation. (C) Individual AUC plasma IDV concentrations (from 0 to 24 h and by dosing interval) and geometric means (90% CI) on day 14 (log plot).

Secondary endpoints.

The geometric means of IDV C_max based on the average across the dosing intervals were 10,428 nM (90% CI, 9,209, 11,808) for IDV/RTV at 667/100 mg q12h and 10,407 nM (90% CI, 9,190, 11,784) for IDV at 800 mg q8h (Table 1; Fig. 3B). These values were not significantly different, with a geometric mean C_max ratio for IDV/RTV of 667/100 mg q12h compared to IDV at 800 mg q8h of 1.00 (90% CI, 0.84, 1.19; P = 0.985) (Table 1). Geometric mean IDV C_max values by dosing interval for IDV/RTV (0 to 12 and 12 to 24 h) and IDV alone (0 to 8, 8 to 16, and 16 to 24 h) were calculated and are presented in Fig. 3B. Variability in geometric mean IDV C_max values was seen between the IDV/RTV dosing intervals of 0 to 12 and 12 to 24 h, as well as the IDV-alone dosing intervals of 0 to 8, 8 to 16, and 16 to 24 h. With both regimens, the first dose of the day was associated with the highest geometric mean IDV C_max. The geometric mean IDV C_max with the first dose of the day appeared to be lower for IDV/RTV than IDV alone, and the second of the two IDV/RTV doses appeared to have a lower geometric mean IDV C_max than either of the subsequent IDV-alone doses.

The geometric means of IDV AUC_0-24 were 119,557 nM (90% CI, 98,476, 145,150) for IDV/RTV at 667/100 mg q12h and 77,034 nM (90% CI, 63,451, 93,525) for IDV at 800 mg q8h (Table 1; Fig. 3C). The geometric mean IDV AUC_0-24 ratio for IDV/RTV at 667/100 mg q12h over IDV at 800 mg q8h was statistically significant: 1.55 (90% CI, 1.18, 2.04; P = 0.012) (Table 1). Geometric mean IDV AUC values by dosing interval for IDV/RTV (0 to 12 and 12 to 24 h) and IDV alone (0 to 8, 8 to 16, and 16 to 24 h) were calculated and are presented in Fig. 3C. Variability in geometric mean IDV AUC values was seen between the IDV/RTV dosing intervals of 0 to 12 and 12 to 24 h, as well as the IDV-alone dosing intervals of 0 to 8, 8 to 16, and 16 to 24 h. With both regimens, the first dose of the day appeared to be associated with the highest geometric mean IDV AUC.

The RTV geometric means for C_min, C_max, and AUC_0-24 for the IDV/RTV combination regimen are presented in Table 2.

TABLE 2.

RTV pharmacokinetic parameters on day 14

Parameter (n = 12)	Result for IDV/RTV at 667/100 q12h
	Geometric mean	90% CIb
Cmin (nM)	880a	524, 1,010
Cmax (nM)	2,387a	1,767, 2,753
AUC0-24 (nM · h)	34,060	25,045, 38,926

^aThe average was obtained by first computing the average of the natural log (C_12h and C_24h) for IDV/RTV q12h and then reexpressing the results in the original units via exponentiation.

^b90% CI for the geometric mean are based on a standard error from a t test.

Safety.

The incidence of clinical adverse events was similar in both treatment groups, and the events were generally mild. Among the 27 patients treated, 7 of 14 (50%) IDV/RTV patients and 7 of 13 (54%) patients receiving IDV alone reported an adverse event. There were no cases of nephrolithiasis. Nausea was the only adverse event reported in more than one patient in either treatment group (two patients receiving IDV alone reported nausea of mild severity). The most common classes of adverse events were gastrointestinal (abdominal pain, diarrhea, dry mouth/lips, dyspepsia, nausea, vomiting) and nervous system adverse events (dizziness, headache, hypoesthesia, sensory loss, and syncope).

There were four patients who experienced one or more drug-related adverse events in the IDV/RTV treatment group: (i) headache; (ii) chills; (iii) diarrhea and nausea; (iv) and dry lips, dry mouth, feelings of puffiness, loss of sensation in the left lateral thigh, and numbness of bilateral insteps and palms. In addition, there were four patients who experienced one or more drug-related adverse events in the IDV-alone treatment group: (i) diarrhea, gastroenteritis, and vomiting; (ii) elevated BUN and creatinine, heartburn, and nausea; (iii) nausea; and (iv) headache. There were no serious adverse events or deaths. One patient receiving the combination regimen discontinued treatment due to fever considered by the investigator to be not related to treatment. One patient in the IDV-alone treatment group experienced a possibly drug-related laboratory adverse event, elevation of BUN and creatinine, which led to study discontinuation. There were no other laboratory adverse events.

Total cholesterol increased with IDV/RTV and IDV alone, respectively, from day 1 (182.4 and 200.6 mg/dl) to day 14 (212.8 and 212.3 mg/dl). Likewise, serum triglycerides increased with IDV/RTV and IDV alone, respectively, from day 1 (126.0 and 163.4 mg/dl) to day 14 (183.3 and 189.8 mg/dl). Increases in both total cholesterol and serum triglycerides appeared to be larger with the IDV/RTV regimen than IDV alone, while the final day 14 levels achieved with both regimens appeared similar. Increases in total serum bilirubin with IDV/RTV and IDV alone, respectively, from day 1 (0.5 and 0.7 mg/dl) to day 14 (1.2 and 0.9 mg/dl) also appeared to be larger with the IDV/RTV regimen than with IDV alone.

DISCUSSION

Improvement in protease inhibitor therapy, specifically the use of less frequent and lower doses of protease inhibitors, is possible with the concomitant administration of low-dose RTV. The IDV/RTV 800/100-mg q12h regimen is the best studied of the IDV/RTV regimens. This regimen demonstrated clinical efficacy comparable to that of IDV at 800 mg q8h (HIV RNA levels of <50 copies/ml in 64 and 59% of patients, respectively) in the BEST study and offers greater potential to overcome resistance with higher maintained levels of drug in blood (4, 8, 9, 25, 28). The present study was the first to evaluate the 24-h steady-state pharmacokinetics of the lower dose of IDV at 667 mg in combination with RTV at 100 mg. This combination regimen produced a 6-fold increase in IDV C_min compared to IDV at 800 mg q8h, as well as an increase of 1.6-fold in IDV AUC_0-24. The IDV C_max levels of the two regimens were comparable. These results suggest that regimens including IDV/RTV at 667/100 mg q12h should provide comparable efficacy to IDV 800-mg q8h regimens secondary to increased C_min in antiretroviral naïve patients harboring wild-type virus.

The design of this study had many similarities to previous IDV/RTV pharmacokinetic studies to allow comparison of results (25, 30). Pharmacokinetic evaluations were conducted on day 14 to ensure achievement of steady-state IDV levels, since indinavir has a half-life of typically 1 to 2 h (25, 30). Concentration-time curves and day 12, 13, and 14 C_min values demonstrated that this was accomplished. As in previous IDV pharmacokinetic studies, the concomitant antiretroviral agents, the nucleosides and the nucleotide (tenofovir), were known not to substantially affect the pharmacokinetics of IDV (29). Unique characteristics of the study included dosing of the IDV/RTV group without regards to meals, inclusion of HIV-positive patients receiving nucleoside and nucleotide therapy, and evaluation of full 24-h pharmacokinetics on day 14, rather than a single dosing interval.

Patients receiving IDV alone adhered to the IDV prescribing information regarding timing around meals, while patients on IDV/RTV administered their medication regardless of meals. The results are consistent with three dose-ranging pharmacokinetic studies, which observed higher maintained IDV levels with IDV/RTV than with IDV alone, regardless of administration of the combination regimen with meals (13, 25, 30).

Unlike earlier IDV/RTV pharmacokinetic studies with healthy subjects, this study included HIV-positive patients who were receiving concomitant nucleosides or nucleotides (13, 25). A study of 901 newly treated clinic patients found a higher likelihood of viral suppression among patients who were antiretroviral naïve than those previously treated (27). Comparison between the study by Saah et al. and Gerber et al. suggested that the total exposure, C_min, and C_max of IDV and RTV are lower in HIV-infected patients than in seronegative subjects (11, 25). Slain et al. observed greater variability in hepatic activity of CYP3A4 in HIV-positive patients compared to HIV-negative healthy controls, which may have contributed to these findings (26). In addition, Acosta et al. found that protease inhibitor-naïve patients with undetectable HIV RNA who were started on a protease inhibitor regimen had statistically higher C_8h and AUC_0-8 values than similarly treated patients with detectable HIV RNA at baseline (3). Thus, pharmacokinetic differences with protease inhibitors may exist between healthy subjects compared to HIV-positive patients (possibly in relation to viral load). The inclusion of HIV-positive patients with HIV RNA levels up to 1,000 copies/ml and prior, although limited, exposure to antiretroviral therapy in this study allow for more accurate extrapolation of the study results to the clinical setting.

Concomitant nucleoside or nucleotide inhibitors were included in the patients' treatment regimens, as they would be prescribed. The prescribing information for these agents indicates that they do not affect IDV pharmacokinetics, and they were not expected to affect the results of this study.

Most previous pharmacokinetic studies of protease inhibitors collected samples for the duration of a single dosing interval and estimated AUC_0-24 by multiplying the results from the single dosing interval by the number of dosing intervals (25, 30). These studies did not evaluate possible diurnal variation in drug metabolism. As was observed in the present study, some variability may exist between dosing intervals, which can be most accurately accounted for with a full 24-h evaluation. Hsu et al. also reported slight variability in pharmacokinetic endpoints over 24 h between dosing intervals (13). In the present study, IDV/RTV group C_min levels appeared to increase with the second dose, while C_max and AUC appeared to decrease with the second dose, possibly in association with a diurnal effect. A potential diurnal effect was also observed in a study by Gerber et al. of IDV/RTV as salvage therapy (11). Patients in the IDV-alone group experienced a decrease in C_max and AUC after the first dose, possibly attributable to a food effect. However, C_min with IDV alone was greatest after the third dose. Regardless of differences in pharmacokinetic levels between doses, mean C_min levels by dose with IDV/RTV were greater than those with IDV alone. Further study is needed to adequately characterize diurnal variation in IDV pharmacokinetics, although the clinical significance of this phenomenon is unclear.

The three pharmacokinetic parameters evaluated in this study correlate well to patient outcomes: C_min and AUC_0-24 to the regimen's potential to maintain suppression of HIV RNA and C_max to the regimen's potential to cause exposure-related side effects (1, 14, 23). The fourfold increase in the lower bound of the 90% CI of C_min with IDV/RTV compared to IDV alone confirmed the primary hypothesis and, along with increases in AUC, suggests that IDV/RTV should provide efficacy at least comparable to that with IDV alone. The additional benefit of the IDV/RTV 667/100-mg q12h regimen is the potential for greater convenience and patient compliance. Patients can take their dose of medication in the morning and evening when meals and work are less of a concern.

C_min is known to correlate with viral suppression. Condra et al. determined that the 95% inhibitory concentration of wild-type HIV-1 of IDV is 68 nM (8). The C_min of IDV/RTV (1,511 nM) and IDV alone (250 nM) in the present study were well above this level, with IDV alone consistent with its historical reference (251 nM) (8). The C_min of IDV/RTV at 667/100 mg was lower than C_min levels with IDV/RTV at 400/400 mg (1,891 nM), 800/100 mg (2,233 nM), and 800/200 mg (5,344 nM) given with a high-fat meal reported by Condra et al. (8). The higher C_min levels achieved with combination IDV/RTV regimens appear necessary to overcome resistant strains of HIV, while C_min levels achieved with IDV 800-mg q8h regimens may be inadequate to overcome some resistant strains of HIV.

Data from Saah et al. suggest that higher IDV doses (800 mg) and/or RTV doses (>100 mg) might provide better efficacy, but might also contribute to greater toxicity (9, 25). Thus it appears that it is most appropriate to individualize dosing with an IDV/RTV regimen, optimally choosing a dose of IDV of ≤800 mg and RTV of 100 mg to account for prior treatment failures, potential for resistance, and ability to tolerate side effects.

In addition to the RTV/IDV 667/100-mg regimen, the pharmacokinetics and clinical efficacy of the even-lower-dose IDV/RTV 400/100-mg regimen have been studied (5, 12, 15). Twenty HIV-infected patients were switched from an IDV 800-mg q8h regimen (baseline) to an IDV/RTV 400/100-mg q12h regimen and monitored for 48 weeks in a study by Ghosn et al. (12). From baseline to week 4, mean IDV C_min more than doubled, while C_max values were halved. All patients' plasma HIV RNA levels were <200 copies/ml at week 48. Katlama et al. reported viral suppression of <400 copies/ml among 65% of patients by intention-to-treat analysis and 96% of patients on treatment in a study of 40 HIV-positive, antiretroviral-naïve patients treated with an IDV/RTV 400/100-mg q12h regimen (15). The median IDV C_min concentration at week 4 was 429 ng/ml. Direct comparator studies with the IDV/RTV 400/ 100-mg regimen and the IDV/RTV 800/100-mg regimen, which has demonstrated efficacy to serve as a control, are needed.

Pharmacokinetic studies are generally too short and small to adequately evaluate the tolerability and safety of a medication. A previous study by Lamotte et al. comparing C_min, C_max, and adverse event profiles of four IDV/RTV b.i.d. regimens did find that lowering the IDV dose improved tolerability (16). Similarly, RTV doses above 100 mg in the combination regimens have been associated with twice as many side effects (9). In the present study, the IDV/RTV and IDV-alone regimens had similar adverse event profiles and relatively few adverse events were reported. In the only clinical study directly comparing IDV at 800 mg q8h to IDV/RTV at 800/100 mg q12h, the IDV-alone regimen in the BEST study appeared to be better tolerated (4). This perception may have been perpetuated by the switch design, with fewer side effects among patients remaining on a treatment with which they were comfortable, as opposed to being switched to an alternative regimen (24). While additional data are needed to evaluate the safety of IDV/RTV at 667/100 mg, the present pharmacokinetic data, particularly C_max values, suggest that IDV/RTV at 667/100 mg q12h would be as well tolerated as the IDV 800-mg q8h regimen.

The results of this study suggest that IDV/RTV at 667/100 mg q12h would be an appropriate alternative to IDV at 800 mg q8h, with comparable efficacy and tolerability and improved convenience in protease inhibitor-naïve individuals. This regimen may offer some tolerability advantages compared to other b.i.d. IDV/RTV regimens, which include higher doses of either medication. Clinical experience is needed with the IDV/RTV 667/100-mg regimen to confirm the expectations generated by this pharmacokinetic study. However, the data presented in this study support its use in clinical practice. This study provides us with a pharmacokinetic characterization of an additional IDV/RTV treatment regimen for patients requesting new, effective, well-tolerated, and convenient HIV therapies. This characterization includes the observation of differences in pharmacokinetic endpoints between dosing intervals. However, regardless of the dosing regimen, IDV/RTV at 667/100 mg q12h maintained higher IDV C_min and AUC_0-24 values, with no differences in C_max compared to IDV at 800 mg q8h.

In conclusion, the geometric mean ratio of IDV C_min for the regimen including IDV/RTV at 667/100 q12h relative to IDV 800 at q8h was at least 4. The IDV AUC_0-24 with the IDV/RTV regimen was also greater than that of the IDV-alone regimen, while the two regimens' C_max values did not differ. The adverse event profiles of the two regimens were also similar. This suggests that IDV/RTV at 667/100 mg q12h would be an effective and well-tolerated alternative to the IDV 800-mg q8h regimen, which would offer better convenience.