| |
Similar Safety and Efficacy of Once- and Twice-Daily Lopinavir/Ritonavir Tablets in Treatment-Experienced HIV-1-Infected Subjects at 48 Weeks
|
| |
| |
JAIDS Journal of Acquired Immune Deficiency Syndromes:
1 June 2010 - Volume 54 - Issue 2 - pp 143-151
"Through 48 weeks, 135 subjects (22.5%) discontinued prematurely, including 66 (22.0%) in the QD arm and 69 (23.1%) in the BID arm (P = 0.770).....Based on the ITT-TLOVR analysis, virologic efficacy was similar between treatment groups. At week 48, 166 of 300 QD subjects (55.3%) and 155 of 299 BID subjects (51.8%) were responders......Mean taking compliance at week 24 was 84.4% for QD and 78.1% for BID (P = 0.003), and mean correct dosing compliance was 79.6% and 68.1% for QD versus BID, respectively (P < 0.001). Lastly, mean timing compliance was also higher in the QD arm, at 65.8% compared with 58.2% in BID (P = 0.014)......Rates of drug-related adverse events of moderate or greater severity occurred with similar frequency in the 2 treatment groups, with the exception of nausea, which was reported in 2.7% of QD and 7.4% of BID subjects (P = 0.009; Table 3). There was no significant difference between treatment groups in the incidence of moderate-to-severe drug-related diarrhea (14.0% QD, 11.0% BID, P = 0.323) or the incidence of diarrhea requiring treatment with additional medication (16.0% QD; 16.4% BID; P = 0.912). Adverse events leading to discontinuation occurred in 13 QD subjects (4.3%) and 21 BID subjects (7.0%), with no statistically significant difference between groups (P = 0.163). Diarrhea contributed to premature discontinuation in 7 QD subjects (2.3%) and 6 BID subjects (2.0%; P > 0.999). .....Virologic response rates were lower among subjects with a greater number of protease inhibitor resistance-associated mutations at baseline, based on a dropouts-as-censored end point"....regarding resistance see Results below
Zajdenverg, Roberto MD; Podsadecki, Thomas J MD; Badal-Faesen, Sharlaa MD; Andrade-Villanueva, Jaime MD; Gathe, Joseph MD; Mingrone, Horacio MD; Fredrick, Linda M MS; Gaultier, Isabelle A MS; Woodward, W Chris DO; Bernstein, Barry M MD
From the *Projeto PraŤa Onze, UFRJ, Rio de Janeiro, Brazil; Abbott, Abbott Park, IL; University of Witwatersrand, Johannesburg, South Africa; Hospital Civil de Guadalajara, Guadalajara, Mexico; Therapeutic Concepts, Houston, TX; and HIV Outpatient Care Unit, Muniz Hospital, Buenos Aires City, Argentina.
Abstract
Objectives: To compare the safety and antiviral activity of once (QD) or twice (BID) daily lopinavir/ritonavir (LPV/r) in combination with investigator-selected nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs) in treatment-experienced subjects.
Methods: Subjects failing treatment with HIV-1 RNA > 1000 copies per milliliter received LPV/r tablets 800/200 mg QD (n = 300) or 400/100 mg BID (n = 299) with investigator-chosen nucleoside/nucleotide reverse transcriptase inhibitors. Efficacy was determined by the intent-to-treat time to loss of virologic response (ITT-TLOVR) algorithm. Safety, tolerability, adherence, impact of baseline protease mutations on virologic response, and emergence of resistance on therapy were assessed.
Results: Demographics were comparable across groups. By intent-to-treat time to loss of virologic response, 166 QD subjects (55.3%) and 155 BID subjects (51.8%) were responders at week 48 (P = 0.413), with similar mean increases in CD4+ T-cell count. QD subjects demonstrated better adherence than BID subjects. The occurrence of treatment-related moderate/severe adverse events was comparable for all events except nausea, which was reported more frequently among BID-treated subjects. Emergence of new protease resistance mutations on treatment was similarly infrequent in both groups.
Conclusion: LPV/r dosed QD resulted in increased treatment adherence and was as efficacious as BID LPV/r while providing similar safety, tolerability, and limited resistance evolution.
INTRODUCTION
Since being introduced in the 1980s, antiretroviral treatment (ART) has evolved from complicated and relatively toxic regimens into simplified and generally well-tolerated therapies. HIV infection is now considered a chronic disease requiring lifelong suppressive therapy. Although mortality rates have not been reduced to those of the general population, most young HIV-infected persons undergoing treatment have estimated survivals of at least 35 years after diagnosis.1 Despite these successes, treatment failures attributable to poor tolerability, suboptimal adherence, or transmitted resistance continue to occur, often necessitating changes in therapy.
Adherence is an important factor that affects the response to ART.2 Pill burden, dose frequency, and tolerability are the major factors that patients consider as influencing adherence.3 Simplifying treatment by decreasing the dosing frequency may be instrumental in increasing adherence for select patients. Several studies comparing once-daily (QD) and twice-daily (BID) dosing have confirmed that QD dosing increases adherence.4-8 For example, a comparison of LPV/r dosed QD and BID in treatment-naïve subjects revealed that adherence was better in the QD arm.7 However, few studies comparing QD with BID dosing of specific antiretroviral agents have been conducted in treatment-experienced populations.
Lopinavir/ritonavir (LPV/r) is a coformulation of the HIV-1 protease inhibitor lopinavir with low-dose ritonavir, which acts to increase plasma levels of lopinavir by inhibiting lopinavir metabolism. LPV/r dosed 400/100 mg BID is approved for treatment of naïve and experienced HIV-1-infected patients. LPV/r administered 800/200 mg QD has been compared with BID dosing in several studies of antiretroviral-naïve subjects.7-9 In the largest of these trials, study M05-730, 664 treatment-naïve subjects were randomized to receive LPV/r QD (n = 333) or BID (n = 331), with all subjects also receiving tenofovir DF and emtricitabine (FTC). Antiviral activity was comparable for LPV/r QD and BID through 96 weeks of treatment.8,10 The incidence and character of treatment-related adverse events were similar between dosing groups and consistent with previous LPV/r studies.8,11 Emergence of resistance was infrequent and occurred with similar frequency in the 2 treatment groups. LPV/r 800/200 mg QD has been approved for naïve patients in the United States and many other countries since 2005.
Study M06-802 is the first, large, randomized study of LPV/r tablets dosed QD versus BID in ART-experienced subjects failing their current regimen. The objectives of the study were to compare the safety, tolerability, and antiviral activity of LPV/r dosed QD and BID in combination with an optimized background regimen of at least 2 nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs) and to evaluate the emergence of resistance through 48 weeks of treatment. The impact of dosing frequency on adherence to LPV/r was also assessed.
DISCUSSION
Previous studies have revealed that therapy with LPV/r dosed QD provides similar clinical benefits as BID LPV/r in antiretroviral-naïve subjects.7-10 This study extends these findings to treatment-experienced subjects. In this large, randomized, 48-week trial, similar efficacy, safety, tolerability, and emergence of resistance were demonstrated with QD and BID LPV/r dosing in a population failing their previous antiretroviral regimen, although treatment compliance was higher in the QD group.
The study population was representative of what may be encountered in clinical practice. Reflective of treatment guidelines and clinical practice, subjects were selected for treatment with LPV/r based on genotypic analysis, treatment history, and the investigator's assessment that LPV/r plus ≥2 NRTIs was an appropriate treatment regimen. A modest prevalence of protease inhibitor resistance was noted at screening. A significant number of women, Hispanics, and non-white individuals were enrolled, suggesting these results may be broadly applicable to the HIV-infected population.
Overall, LPV/r dosed QD was shown to provide efficacy comparable to LPV/r dosed BID in all statistical analyses. Antiviral activity was also similar for QD- and BID-treated subjects regardless of baseline viral load or CD4+ T-cell count. Pharmacokinetic studies in naïve subjects have shown that lopinavir trough levels exceed the half maximal inhibitory concentration (IC50) for wild-type HIV regardless of whether LPV/r is dosed QD or BID, thus supporting either dosing strategy.9 It is likely that lopinavir trough levels in both the QD and the BID treatment groups also exceeded the IC50 of virus from the majority of subjects enrolled in this study, given the limited genotypic resistance in protease observed at screening. The high rate of treatment compliance observed in subjects taking LPV/r in this study may also have influenced viral suppression.
Although cross-study comparisons must be undertaken with caution, the discontinuation and response rates in this study are similar to those observed in recent trials utilizing LPV/r in treatment-experienced subjects. The TMC114-C214 (TITAN) study resulted in 29% of 297 BID LPV/r-dosed subjects discontinuing prematurely, including 7% and 11% for adverse events or virologic failure, respectively. Through 48 weeks, 60% of subjects in the LPV/r treatment group had HIV-1 RNA < 50 copies per milliliter17; however, a third class of antiretroviral drug was allowed. Premature discontinuation occurred in 17% of 103 BID LPV/r-treated subjects in the APV30003 (CONTEXT) study; adverse events contributed to discontinuation in 8% of subjects, whereas virologic failure accounted for discontinuation of < 1% of these subjects. Half (50%) of LPV/r-treated subjects demonstrated HIV-1 RNA < 50 copies per milliliter through 48 weeks.22 Lastly, the BMS-045 study treated 118 subjects with extensive antiretroviral history in the LPV/r BID group; 11% of LPV/r-treated subjects discontinued prematurely (2% due to adverse events), and 46% achieved HIV-1 RNA levels < 50 copies per milliliter through 48 weeks.23 The similar response rates observed in these studies support the broad applicability of findings from the M06-802 trial.
The efficacy of LPV/r QD and BID was maintained even in the presence of baseline protease inhibitor-associated mutations. Three mutation lists that have been shown to be predictive of response to LPV/r-based therapy in previous analyses were assessed in this study.12-16,20 Subjects showed similar response rates with as many as 4 mutations from list 1 and with 2 or fewer mutations from the list included in the US prescribing information (list 3), consistent with previous reports.15,16,23,24 Of particular note, LPV/r dosing frequency did not impact virologic response for any of the mutation lists evaluated. An insufficient number of subjects were enrolled with substantial lopinavir resistance to assess potential differences in activity of QD- and BID-dosed LPV/r in subjects with 3 or more mutations.
The evolution of resistance while on therapy was also unaffected by QD or BID dosing. A similar number of subjects from each arm met protocol-defined criteria for genotypic analysis. Detection of NRTI mutations that were not present at baseline may reflect the selection of novel NRTI resistance-associated mutations or the reemergence of archived mutations. The majority of changes in protease sequence while on treatment were polymorphisms or mutations at secondary sites,20 with few new primary protease inhibitor-associated mutations. Overall, these findings suggest a similar low emergence of resistance with QD dosing compared with BID dosing in subjects with limited baseline resistance.
A previous study of the tablet formulation in treatment-naive subjects indicated that QD and BID LPV/r dosing have similar safety and tolerability profiles.8 These findings were confirmed in ART-experienced subjects, where the incidence of treatment-related moderate-to-severe adverse events and grade 3+ laboratory abnormalities was similar among subjects receiving QD or BID dosing. As with previous studies of LPV/r, the most common adverse events were gastrointestinal in nature, with diarrhea the most frequently reported of these events. The rates of moderate-to-severe diarrhea related to treatment were not significantly different between QD and BID and were consistent with previous reports of LPV/r-based therapy.8,11,17,23 Diarrhea leading to study drug discontinuation was infrequent, occurring in only 2.3% of QD-treated and 2.0% of BID-treated subjects. Nausea was the only study drug-related adverse event of at least moderate severity that showed a difference between treatment groups; it was noted in a statistically significantly greater proportion of BID-treated subjects compared with QD-treated subjects.
As observed with previous LPV/r studies, the most frequently observed laboratory abnormalities were elevations in cholesterol and triglycerides. There were no differences between groups in the incidence of grade 3 or greater laboratory abnormalities, also suggesting a similar safety profile for LPV/r dosed QD and BID.
Several limitations of this study should be noted. First, the LPV/r dosing regimen was unblinded to subjects and investigators. As there has been extensive use of LPV/r, both QD and BID, in many of the research sites, previous experiences with each dosing regimen may have impacted adverse event reporting or management. Second, the number of subjects with moderate or high-level lopinavir resistance enrolled in the study was limited. Although this reflects appropriate selection of subjects for treatment with LPV/r, the impact of dosing interval on antiviral response in subjects with 3 or more mutations at baseline cannot be determined with certainty. Lastly, NRTIs were selected by the investigator. Although NRTI use was similar in the 2 study arms, small differences in the use of specific NRTIs may have impacted either safety or efficacy of the 2 treatment groups.
In conclusion, the safety, tolerability, and antiviral activity of LPV/r were similar whether dosed QD or BID in ART-experienced subjects. Notably, the emergence of new mutations in protease was infrequent and occurred at a similar rate in the 2 treatment arms, suggesting minimal emergence of resistance. Based on these findings, LPV/r QD should be considered as an option for treatment-experienced HIV-1-infected patients desiring a simplified dosing regimen.
ACKNOWLEDGMENTS
The authors acknowledge the efforts of all of the study participants who made this investigation possible. The authors acknowledge the investigators at each site for their contributions to this study. The authors wish to thank the M06-802 coordinators and study team for their support of the conduct and completion of this study. Elaine M. Smith, Ph.D. (medical writer, Abbott) drafted and managed the revision of this article.
RESULTS
Subjects
A total of 599 subjects were randomized and dosed: 300 in the QD arm and 299 in the BID arm. Subject disposition is presented in Figure 1. Through 48 weeks, 135 subjects (22.5%) discontinued prematurely, including 66 (22.0%) in the QD arm and 69 (23.1%) in the BID arm (P = 0.770). Noncompliance and loss to follow-up contributed to discontinuations in 5.0% and 6.7% of subjects, respectively; 2.7% of subjects discontinued due to withdrawal of consent. Study investigators provided all reasons for discontinuation without detailed narratives, so it is difficult to determine whether noncompliance reflected poor adherence to study medication or nonparticipation in study procedures. Virologic failure and adverse events/HIV-related events were provided as reasons for discontinuation of 3.7% and 6.0% subjects, respectively. There were no statistically significant differences between groups in the overall discontinuation rate or rates of discontinuation due to a particular reason. Five subjects (2 QD and 3 BID) died during the study; another BID subject died within 30 days after the last dose of study drug. The cause of death was generally attributed to comorbid disease expected in the population studied, with all deaths considered by the investigators as not related or probably not related to study drug.
There was no statistically significant difference in any demographic characteristic between LPV/r treatment groups (Table 1). Each group included a relatively large proportion of female subjects (34.4% overall). Furthermore, the study represented diverse racial (34.7% black) and ethnic (33.9% Hispanic) populations. Mean baseline plasma HIV-1 RNA levels were 4.26 log10 copies per milliliter in each arm. The mean baseline CD4+ T-cell count was statistically significantly lower in the QD group compared with that in the BID group (Table 1). However, the distribution of subjects with CD4+ T-cell counts <50 cells, ≥50 to <200, and ≥200 cells per cubic millimeter was not statistically significantly different between groups. The study population reflected a broad spectrum of ART experience; all but 3 subjects had prior NRTI exposure, and the majority (84.3%) had used at least one nonnucleoside reverse transcriptase inhibitor. Approximately half of the participants had been previously treated with at least one protease inhibitor. Although 17%-25% of subjects had ≥2 baseline protease inhibitor mutations, depending on which of the above lists were examined, only a total of 29 subjects (12 QD and 17 BID) demonstrated evidence of possible or high-level resistance to LPV/r at baseline on genotypic resistance testing using the ViroSeq interpretation algorithm.
Tenofovir DF was the most commonly used NRTI, included as part of 70.2% of regimens in this study. NRTI use was well balanced between arms, with no significant differences in individual NRTIs, total or active NRTIs used, or whether the NRTIs were dosed QD or BID (Table 2).
Virologic and Immunologic Efficacy
Based on the ITT-TLOVR analysis, virologic efficacy was similar between treatment groups. At week 48, 166 of 300 QD subjects (55.3%) and 155 of 299 BID subjects (51.8%) were responders (P = 0.413; Fig. 2A). The difference between QD and BID was 3.5% (95% CI -4.5, 11.5); as the lower margin of the 95% CI was greater than -12%, the criteria for noninferiority of the LPV/r QD regimen compared with the BID regimen was met. The nonresponders in the ITT analysis included 27 QD subjects (9.0%) and 32 BID subjects (10.7%) who never achieved 2 consecutive HIV-1 RNA levels <50 copies per milliliter but were on study at week 48. Similar proportions of subjects in each treatment group discontinued before confirmed virologic suppression: 47 QD-dosed LPV/r subjects (15.7%) and 41 BID-dosed LPV/r subjects (13.7%). Lastly, 6 (2.0%) and 15 (5.0%) subjects from the LPV/r QD and BID dosing groups, respectively, discontinued while suppressed.
Observed data analysis showed a similar response between groups. At week 48, 171 of 225 QD subjects (76.0%) and 161 of 223 BID subjects (72.2%; P = 0.389; difference 3.8%, 95% CI -4.3, 11.9) had plasma HIV-1 RNA levels <50 copies per milliliter (Fig. 2B).
Additional ITT analyses (NC = F, M = F, and last observation carried forward) also revealed similar efficacy between treatment groups. There were no statistically significant differences in the proportion of subjects achieving HIV-1 RNA <50 copies per milliliter between treatment groups within subgroups defined by baseline HIV-1 RNA or CD4+ T-cell counts.
Mean changes in CD4+ T-cell counts were comparable between QD and BID arms (Fig. 2C). From baseline through week 48, the mean CD4+ T-cell count increased by 135 and 122 cells per cubic millimeter, for QD and BID subjects, respectively (P = 0.281).
Adherence
Compliance data were obtained from 273 QD subjects and 274 BID subjects who had medication event monitoring system cap results available. Adherence during the first 24 weeks of treatment was statistically significantly higher in the QD regimen for the 3 predefined measures of adherence. Mean taking compliance at week 24 was 84.4% for QD and 78.1% for BID (P = 0.003), and mean correct dosing compliance was 79.6% and 68.1% for QD versus BID, respectively (P < 0.001). Lastly, mean timing compliance was also higher in the QD arm, at 65.8% compared with 58.2% in BID (P = 0.014).
Adverse Events
Rates of drug-related adverse events of moderate or greater severity occurred with similar frequency in the 2 treatment groups, with the exception of nausea, which was reported in 2.7% of QD and 7.4% of BID subjects (P = 0.009; Table 3). There was no significant difference between treatment groups in the incidence of moderate-to-severe drug-related diarrhea (14.0% QD, 11.0% BID, P = 0.323) or the incidence of diarrhea requiring treatment with additional medication (16.0% QD; 16.4% BID; P = 0.912). Adverse events leading to discontinuation occurred in 13 QD subjects (4.3%) and 21 BID subjects (7.0%), with no statistically significant difference between groups (P = 0.163). Diarrhea contributed to premature discontinuation in 7 QD subjects (2.3%) and 6 BID subjects (2.0%; P > 0.999). The prevalence of gastrointestinal adverse events generally decreased over time (data not shown). Other adverse events were less commonly noted, and no clinically significant differences in character or incidence between treatment groups were observed.
Table 3 includes the incidence of grade 3+ laboratory abnormalities occurring in ≥ 2.0% of subjects in either treatment group. There were no statistically significant differences between dosing groups in the proportion of subjects experiencing grade 3 or higher laboratory abnormalities. Grade 3 or 4 lipid elevations were the most commonly observed laboratory abnormalities and were noted with similar frequency in the QD and BID groups. These lipid elevations did not result in study drug discontinuation in any subject, and none were associated with an adverse event of pancreatitis. The mean changes from baseline to week 48 for these clinical laboratory parameters were not statistically significantly different between the QD and BID dosing groups. Mean total cholesterol increased by 0.460 mmol/L (17.7 mg/dL) from baseline through week 48 for QD subjects and 0.401 mmol/L (15.4 mg/dL) for BID subjects (P = 0.514), whereas the mean increase from baseline through week 48 for triglycerides was 0.708 mmol/L (62.6 mg/dL) and 0.623 mmol/L (55.1 mg/dL) for the QD and BID treatment groups, respectively (P = 0.593). Mean changes from baseline to week 48 in high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and the low-density lipoprotein to high-density lipoprotein ratio were also comparable for QD- and BID-treated subjects.
Resistance
Impact of Baseline Protease Resistance on Virologic Response
Virologic response rates were lower among subjects with a greater number of protease inhibitor resistance-associated mutations at baseline, based on a dropouts-as-censored end point. For list 1, overall response rates were similar among subjects with 0-4 baseline protease mutations and were lower for the small number of subjects with > 4 mutations. However, this trend was not statistically significant. There was no statistically significant impact of the number of baseline protease mutations from list 2 on virologic response. For list 3, the number of baseline mutations significantly affected antiviral efficacy; response rates were similar among subjects with 0-2 baseline protease mutations and were lower among subjects with 3 or more mutations (Table 4). No clinically significant differences were observed between the QD and BID groups with respect to virologic response in the presence of 2 or fewer baseline protease mutations. The number of subjects with more than 2 mutations in each group was relatively small, limiting assessment in these subjects. Similar results were observed based on the FDA ITT-TLOVR end point. Virologic response, however, was less sensitive to the number of prior protease inhibitor-based treatment regimens in the QD group than that in the BID group (Table 5).
Based on a dropouts-as-censored end point, only 2 individual protease mutations were identified with a statistically significant relationship to virologic response at week 48. The presence of a V82A/F/T/S protease mutation at baseline was associated with lower virologic response [4 of 14 subjects (28.6%) versus 329 of 518 subjects (63.5%) without the mutation; P = 0.011]. However, this mutation always occurred in the presence of at least one other lopinavir resistance-associated mutation. A positive influence of any change at codon 30 was revealed; 16 of 18 subjects with this mutation (88.9%) were responders at week 48, whereas 317 of 514 subjects with no alterations at codon 30 (61.7%) were responders (P = 0.023). The absence of statistically significant differences for other mutations may have been affected by a relatively low prevalence of baseline protease mutations. In these small samples, no obvious differences in the impact of particular mutations on virologic response were observed between LPV/r dosing groups.
Emergence of Postbaseline Resistance Mutations
Postbaseline genotyping results were available for 152 subjects (75 QD and 77 BID) with inadequate virologic suppression during the study. At the time of virologic rebound, NRTI-associated mutations19 not detected at baseline were observed in 16 of 75 subjects (21.3%) in the QD group and 19 of 77 subjects (24.7%) in the BID group. The most common new mutation was M184V/I, appearing in isolates from 7 QD and 10 BID subjects.
Specific protease mutations that emerged during treatment were similar in the QD and BID groups. Although 65 subjects (42.8%) showed postbaseline changes in protease, the majority of these were polymorphisms or mutations at secondary sites.20,21 Only 6 of 75 QD subjects (8.0%) and 12 of 77 subjects from the BID group (15.6%) showed treatment-emergent mutations at the primary protease resistance codons (30, 32, 48, 50, 82, 84, and 90)21; the most common of which were at codon 82 in 4 of 67 QD (6.0%) and 7 of 67 BID (10.4%) subjects who did not have a mutation at codon 82 at baseline. The frequency of emergence of new postbaseline protease mutations was not statistically significantly different between the QD and BID groups.
METHODS
Study Design
Study M06-802 was a phase 3, randomized, open-label, multicenter global trial. Study protocols were approved by institutional ethics review boards at participating sites, and all subjects provided written informed consent. This trial was registered with clinicaltrials.gov (NCT00358917).
Study drug administration occurred from September 2006 through November 2008 at 120 sites in 17 countries in North America, South America, Europe, Africa, and Australia. Eligible subjects were randomized in a 1:1 ratio to receive LPV/r 400/100 mg BID or 800/200 mg QD for 48 weeks using the LPV/r 200/50 mg tablet formulation. Participants also received at least 2 investigator-selected NRTIs, based on treatment history and results of genotypic resistance testing at the screening visit, but were not allowed any other class of antiretroviral drug (ie, no nonnucleoside reverse transcriptase inhibitors, fusion inhibitors, or integrase inhibitors). The primary end point was virologic efficacy [by the Food and Drug Administration (FDA) intent-to-treat time to loss of virologic response algorithm, ITT-TLOVR, see details below] at week 48; immunological response, safety, tolerability, adherence, and emergence of viral resistance were also evaluated.
Subjects
ART-experienced HIV-1-infected individuals ≥18 years who were on a stable antiretroviral regimen for at least 12 weeks but failing therapy were eligible for this study. In addition, based on HIV-1 drug resistance test results from the screening visit and prior treatment history, the investigator must have considered LPV/r plus at least 2 NRTIs to be an appropriate treatment regimen for the subject. These inclusion criteria were chosen to approximate the spectrum of treatment-experienced patients in a clinical setting who would be considered candidates for switching to LPV/r-based therapy. Treatment failure was defined as 2 previous consecutive plasma HIV-1 RNA levels >400 copies per milliliter and the most recent plasma HIV-1 RNA level >1000 copies per milliliter. There was no limit on CD4+ T-cell count. Although inclusion criteria allowed for any level of prior ART exposure, subjects were required to be naïve to LPV/r. Subjects were excluded if any of the following laboratory results were observed at screening: hemoglobin ≤8.0 g/dL, neutrophil count ≤750 cells per microliter, platelet count ≤50,000 cells per cubic millimeter, or alanine aminotransferase (SGPT) or aspartate aminotransferase (SGOT) ≥5.0 x upper limit of normal.
Outcome Measures
Virology and Immunology
Plasma levels of HIV-1 RNA were measured using the Roche Ultrasensitive Amplicor HIV-1 assay, version 1.5 (limit of quantitation = 50 copies/mL). CD4+ T-cell cell counts were determined by standard flow cytometry.
Adherence
Adherence to LPV/r was monitored through the first 24 weeks of study by the use of a medication event monitoring system featuring a pill bottle cap that was manufactured by Advanced Analytical Research on Drug Exposure, Ltd (Zurich, Switzerland).
Adverse Events
All treatment-emergent adverse events were recorded by investigators and coded using terminology from the Medical Dictionary for Regulatory Activities (MedDRA). Laboratory hematology and chemistry values outside of the normal range were assigned a severity grade of 1 through 4 based on criteria modified from the NCI Common Terminology Criteria for Adverse Events v3.0.
Resistance
Genotypic sequencing was performed using the ViroSeq assay and interpretation algorithm (Celera Diagnostics, Alameda, CA) at the screening visit. Postbaseline HIV-1 genotypic resistance testing was performed according to protocol-specified criteria and provided to the investigator in real time.
1. If the subject's plasma HIV-1 RNA was ≥50 copies per milliliter and at the previous visit, the plasma HIV-1 RNA was <50 copies per milliliter, a confirmatory plasma HIV-1 RNA and a sample for HIV-1 drug resistance genotyping were obtained within 4 weeks. If the confirmatory plasma HIV-1 RNA was >400 copies per milliliter, HIV-1 drug resistance genotyping was conducted on the sample collected at the time of the confirmatory plasma HIV-1 RNA level.
2. For subjects with a plasma HIV-1 RNA increase >0.5 log10 (3-fold increase) above study nadir and >400 copies per milliliter on 2 consecutive measurements obtained at least 14 days apart while the subject was receiving antiretroviral therapy, HIV-1 drug resistance genotyping was performed on the plasma sample collected at the time of the second consecutive plasma HIV-1 RNA rebound.
3. For subjects who never demonstrated plasma HIV-1 RNA <400 copies per milliliter by week 24, real-time HIV-1 drug resistance genotyping was performed on the plasma sample collected at week 24.
All treatment-emergent mutations were reported. Virologic response (ie, subjects with plasma HIV-1 RNA levels < 50 copies/mL at week 48) was assessed against baseline protease inhibitor mutations using 3 lists of protease mutations associated with reduced response to protease inhibitors.
List 1: L10F/I/R/V, K20M/R, L24I, V32I, L33F, M36I, M46I/ L, I47V/A, G48V, I50V, F53L, I54 (any change), A71V/T, G73S, V82A/F/T/S, I84V, and L90M, consistent with in vitro studies and findings from previous clinical trials of LPV/r.12-14
List 2: Any mutation at amino acid positions D30, V32, M36, M46, I47, G48, I50, F53, I54, G73, V82, I84, N88, and L90, per FDA request.
List 3: L10F/I/R/V, K20M/N/R, L24I, L33F, M36I, I47V, G48V, I54L/T/V, V82A/C/F/S/T, and I84V, consistent with US prescribing information for LPV/r.15,16
Statistical Analyses
All subjects who were randomized and received at least one dose of study medication were included in the analyses.
Baseline Demographics and Subject Disposition
Baseline characteristics, baseline NRTI use, and study discontinuation were compared between groups by 1-way analysis of variance (ANOVA) for continuous variables or Fisher exact test or Pearson χ2 test for categorical variables.
Antiviral Efficacy
The primary end point was the proportion of subjects responding with HIV-1 RNA <50 copies per milliliter at week 48 by the FDA ITT-TLOVR algorithm. The time of failure was defined as the earliest of any of the following events: death, permanent discontinuation of the study drug or lost to follow-up, or confirmed plasma HIV-1 RNA levels ≥50 copies per milliliter from 2 consecutive visits or 1 value ≥50 copies per milliliter followed by permanent discontinuation of the study drug or lost to follow-up. If the time of virologic failure defined above was immediately preceded by a single missing scheduled visit or multiple consecutive missing scheduled visits, then the time of virologic failure was replaced by the first time of such missing visit. Subjects who never achieved confirmed plasma HIV-1 RNA levels <50 copies per milliliter on 2 consecutive visits before the following events were considered to fail at study day 1: death, permanent discontinuation of the study drug or lost to follow-up, or the last available visit. Subjects who remained on study through week 48 without demonstrating a loss of virologic response were considered responders through week 48. The difference between groups was evaluated by Fisher exact test. To test for noninferiority of QD dosing compared with BID dosing, the 95% confidence interval (CI) for the difference in response rates (QD-BID) was determined. The QD regimen was to be considered noninferior to the BID regimen if the lower limit of the 95% CI was at or above -12%. This margin has been used in similar clinical trials to evaluate noninferiority of ART regimens.8,17 A sample size of approximately 600 subjects was chosen to provide >80% power to detect noninferiority at a lower 95% CI margin of -12% with a type I error rate (α) of 0.05.
The proportions of subjects with HIV-1 RNA levels <50 copies per milliliter at week 48 were also calculated using an observed data analysis, in which subjects with missing values were excluded from the analysis, and other intent-to-treat (ITT) analyses where noncompleters were considered failures (NC = F), where subjects with missing values were considered failures (M = F), and where missing values were replaced by the most recent nonmissing value (last observation carried forward).
The efficacy of QD and BID dosing was ascertained for subgroups of subjects based on baseline disease characteristics including HIV-1 RNA levels (<100,000 or ≥100,000 copies/mL) and CD4+ T-cell counts (<50, ≥50 to <200, and ≥200 cells/mm3). The mean change in CD4+ T-cell counts from baseline through 48 weeks was compared between arms using 1-way ANOVA with treatment group as the factor.
Adherence
Compliance to prescribed LPV/r dosing was assessed using 3 measures of adherence. "Taking compliance" is the percentage of prescribed doses taken, "correct dosing compliance" is the percentage of days in which the correct number of doses was taken, and "timing compliance" measures the percentage of doses taken within ±3 hours of the prescribed dosing interval: for QD, the percentage of interdose intervals between 21 and 27 hours and for BID, the percentage of interdose intervals between 9 and 15 hours.18 Differences in the distributions of the proportions were compared between LPV/r QD- and BID-dosed groups using the Wilcoxon rank sum test.
Adverse Events
The proportion of subjects reporting study drug-related adverse events of at least moderate severity was compared between groups by Fisher exact test. Mean changes from baseline in laboratory variables were compared between groups by 1-way ANOVA. Proportions of subjects with grade 3+ laboratory changes were compared between groups using Fisher exact test.
Resistance
The relationship between virologic response and individual baseline mutations was evaluated by Fisher exact test, using a dropouts-as-censored approach to define virologic response. The influence of the number of baseline protease mutations from each list defined above on virologic response was assessed using logistic regression.
The incidence of emergent resistance-associated mutations was determined in subjects experiencing virologic rebound and was compared between treatment groups by Fisher exact test.
| |
| |
| |
|
|
|
