Heather J. Ribaudo

Prevention of HIV-1 Infection with Early Antiretroviral Therapy

BACKGROUND Antiretroviral therapy that reduces viral replication could limit the transmission of human immunodeficiency virus type 1 (HIV-1) in serodiscordant couples. METHODS In nine countries, we enrolled 1763 couples in which one partner was HIV-1-positive and the other was HIV-1-negative; 54% of the subjects were from Africa, and 50% of infected partners were men. HIV-1-infected subjects with CD4 counts between 350 and 550 cells per cubic millimeter were randomly assigned in a 1:1 ratio to receive antiretroviral therapy either immediately (early therapy) or after a decline in the CD4 count or the onset of HIV-1-related symptoms (delayed therapy). The primary prevention end point was linked HIV-1 transmission in HIV-1-negative partners. The primary clinical end point was the earliest occurrence of pulmonary tuberculosis, severe bacterial infection, a World Health Organization stage 4 event, or death. RESULTS As of February 21, 2011, a total of 39 HIV-1 transmissions were observed (incidence rate, 1.2 per 100 person-years; 95% confidence interval [CI], 0.9 to 1.7); of these, 28 were virologically linked to the infected partner (incidence rate, 0.9 per 100 person-years, 95% CI, 0.6 to 1.3). Of the 28 linked transmissions, only 1 occurred in the early-therapy group (hazard ratio, 0.04; 95% CI, 0.01 to 0.27; P<0.001). Subjects receiving early therapy had fewer treatment end points (hazard ratio, 0.59; 95% CI, 0.40 to 0.88; P=0.01). CONCLUSIONS The early initiation of antiretroviral therapy reduced rates of sexual transmission of HIV-1 and clinical events, indicating both personal and public health benefits from such therapy. (Funded by the National Institute of Allergy and Infectious Diseases and others; HPTN 052 ClinicalTrials.gov number, NCT00074581.).

Low-Frequency HIV-1 Drug Resistance Mutations and Risk of NNRTI-Based Antiretroviral Treatment Failure A Systematic Review and Pooled Analysis

CONTEXT Presence of low-frequency, or minority, human immunodeficiency virus type 1 (HIV-1) drug resistance mutations may adversely affect response to antiretroviral treatment (ART), but evidence regarding the effects of such mutations on the effectiveness of first-line ART is conflicting. OBJECTIVE To evaluate the association of preexisting drug-resistant HIV-1 minority variants with risk of first-line nonnucleoside reverse transcriptase inhibitor (NNRTI)-based antiretroviral virologic failure. DATA SOURCES Systematic review of published and unpublished studies in PubMed (1966 through December 2010), EMBASE (1974 through December 2010), conference abstracts, and article references. Authors of all studies were contacted for detailed laboratory, ART, and adherence data. STUDY SELECTION AND DATA ABSTRACTION Studies involving ART-naive participants initiating NNRTI-based regimens were included. Participants were included if all drugs in their ART regimen were fully active by standard HIV drug resistance testing. Cox proportional hazard models using pooled patient-level data were used to estimate the risk of virologic failure based on a Prentice weighted case-cohort analysis stratified by study. DATA SYNTHESIS Individual data from 10 studies and 985 participants were available for the primary analysis. Low-frequency drug resistance mutations were detected in 187 participants, including 117 of 808 patients in the cohort studies. Low-frequency HIV-1 drug resistance mutations were associated with an increased risk of virologic failure (hazard ratio (HR], 2.3 [95% confidence interval {CI}, 1.7-3.3]; P < .001) after controlling for medication adherence, race/ethnicity, baseline CD4 cell count, and plasma HIV-1 RNA levels. Increased risk of virologic failure was most strongly associated with minority variants resistant to NNRTIs (HR, 2.6 [95% CI, 1.9-3.5]; P < .001). Among participants from the cohort studies, 35% of those with detectable minority variants experienced virologic failure compared with 15% of those without minority variants. The presence of minority variants was associated with 2.5 to 3 times the risk of virologic failure at either 95% or greater or less than 95% overall medication adherence. A dose-dependent increased risk of virologic failure was found in participants with a higher proportion or quantity of drug-resistant variants. CONCLUSION In a pooled analysis, low-frequency HIV-1 drug resistance mutations, particularly involving NNRTI resistance, were significantly associated with a dose-dependent increased risk of virologic failure with first-line ART.

Impact of Efavirenz on Neuropsychological Performance and Symptoms in HIV-Infected Individuals

Context Neurologic toxicity is the most commonly reported adverse effect of the antiretroviral drug efavirenz. Contribution In this substudy of a randomized, controlled trial, 12 of 200 (6%) HIV-infected individuals discontinued treatment with efavirenz because of central nervous system symptoms or mood disorders versus 0 of 103 individuals (0%) who were not receiving the drug. Although patients taking efavirenz had more neuropsychological symptoms, such as bad dreams, in the first week of therapy, no statistically significant neuropsychological differences were found at weeks 4, 12, and 24. Implications Some adverse neuropsychological effects associated with efavirenz are probably transient. The Editors Efavirenz is a non-nucleoside reverse transcriptase inhibitor approved for treatment of HIV infection. The drug is potent, is generally well tolerated, and can be administered once daily, making it a preferred treatment option for HIV infection (1, 2). The most commonly reported adverse effect with efavirenz is neurologic toxicity, with more than 50% of patients reporting symptoms in open-label studies (1, 3). Our randomized, controlled study prospectively characterized aspects of the neurologic toxicity of 3 protease inhibitorsparing antiretroviral regimens for the initial treatment of HIV infection. Methods This investigator-initiated trial was a substudy of the AIDS Clinical Trials Group study A5095, a randomized, double-blind trial of 3 antiretroviral regimens: zidovudine and lamivudine in combination with efavirenz; abacavir; or abacavir and efavirenz in combination (4). For simplicity, we will refer to 2 groups: patients who received efavirenz (with or without abacavir) and those who did not. Randomization was performed centrally without reference to center. The study was supported by the National Institutes of Health (NIH) and was approved by the institutional review boards at each of the participating institutions, with each patient providing informed consent to participate in the substudy. All patients at sites taking part in the substudy were invited to participate before randomization for the parent study (Figure 1). Unblinding and within-class substitutions were allowed in cases of treatment-limiting toxicity (we substituted stavudine for zidovudine, didanosine for abacavir, and nevirapine for efavirenz). Participants had not previously received antiretroviral therapy, and their baseline plasma HIV-1 RNA levels were greater than 400 copies/mL. Parent study A5095 enrolled 1147 participants, of whom 303 at 36 clinical trials units volunteered to participate in the additional evaluations for A5097s. Participants were recruited between March 2001 and January 2002. Figure 1. CONSORT diagram of substudy 5097s. The primary measures of neuropsychological performance were the Trail Making Tests (Parts A and B) and the Digit Symbol Substitution Test (part of the Wechsler Adult Intelligence Scale III [5]). A summary neuropsychological Z score (NPZ3) was derived from the sum of the scores from these 3 tests and standardized for age. Positive scores indicated above-normal function, whereas negative scores indicated below-normal function. The entire score was coded as missing if any component of the NPZ3 was not available. The Neurologic AIDS Research Consortium provided administrator training at each site. These tests assessed functioning in the areas of motor persistence, sustained attention, response speed, visuomotor coordination, and conceptual shifting and tracking. Neuropsychometric measures were collected at baseline and at weeks 1, 4, 12, and 24. Testing was performed at each time point to assess symptoms that might be associated with efavirenz use, sleep disorders, anxiety, depression, and history of drug abuse. The instruments are summarized in Table 1. The symptom questionnaire developed for this study is shown in the Appendix Figure. Appendix Figure. Sample participant questionnaire. Table 1. Testing Instruments Whole blood was collected from all participants to determine efavirenz trough concentrations in plasma (13). These data were used to explore relationships between drug exposure and other variables that were evaluated in the study. Statistical Analysis Our substudy was designed to compare neurologic changes from baseline in patients who received efavirenz with changes in those who did not. The study had 90% power to detect a standard deviation of 0.4 for change in the summary neuropsychological performance score from baseline to week 1. We presented descriptive statistics for the study sample and used nonparametric tests to determine treatment differences. Using the nonparametric methods of Hodges and Lehmann (14) and Proc-StatXact software, version 4.0.1 (Cytel Software Corp., Cambridge, Massachusetts), we estimated treatment differences for continuous outcomes with corresponding exact confidence intervals. Generalized estimating equation modeling (a regression method) and the Wei-Johnson test (a nonparametric method for analyzing incomplete 2-sample data) (15) were used to compare treatment groups longitudinally; both methods assumed that data were missing completely at random. We used the Spearman correlation coefficient, a rank-based method that is robust to extreme observations, to evaluate correlations. All significance testing was performed at an level of 0.05 with no adjustment for multiple testing. All reported P values were 2-sided. To assess the potential effect of any missing data, we performed multiple imputation, analyzed 2 worst-case scenarios (Appendix Table 1 and Appendix Table 2), and conducted an as-treated analysis that excluded patients who discontinued efavirenz therapy. We used SAS software (SAS Institute, Inc., Cary, North Carolina) to perform statistical analyses. Test sources included Elsevier Science (Oxford, United Kingdom) for the Pittsburgh Sleep Quality Index, Mind Garden (Redwood City, California) for the State-Trait Anxiety Inventory for Adults, and the National Institute of Mental Health (Bethesda, Maryland) for the Center for Epidemiologic Studies Depression Scale. Role of the Funding Sources This investigator-initiated protocol was supported by the NIH. Drugs used in the study were donated by pharmaceutical companies whose representatives participated in team discussions. The study was monitored by NIH-contracted monitors and was supervised by a data safety monitoring committee that was appointed by the National Institute of Allergy and Infectious Diseases. The NIH-supported biostatistical team working with the AIDS Clinical Trials Group and the Neurologic AIDS Research Consortium performed the statistical analyses. The protocol team, led by the first author, had final responsibility for the study protocol, case report forms, statistical analysis plan, progress of the study, analysis, and reporting of the data, regardless of outcome. The final version was the sole responsibility of the authors. The team had full access to the data files of the study. Results Baseline Evaluations Recruitment characteristics are displayed in Figure 1; demographic characteristics of the study participants are presented in Table 2. The treatment groups were balanced at baseline with respect to demographic characteristics, neuropsychological measures, and responses to the symptom questionnaire. The sleep disturbance component of the global sleep index demonstrated a baseline difference; the patients who eventually received efavirenz had marginally more sleep disturbances (P= 0.048) (data not shown). Other components of the sleep index, including quality, latency, duration, efficiency, use of sleeping medication, and daytime dysfunction, were similar between groups. Alcohol abuse, drug use, and affective disturbances were infrequent and similar for both groups. Table 2. Baseline Characteristics and Evaluations by Treatment Group Disposition of Study Participants The study allowed for drug substitution from the same class of antiretroviral agents in cases of treatment-limiting toxicity. Table 3 summarizes the modifications that occurred and the respective reasons. Appendix Table 3 gives further details of timing of modifications and the ethnicity of the individuals. Primary end point data (the change in NPZ3 from baseline to week 1) were observed in 283 of the 303 (93.4%) participants. Table 3. Reasons for Modification of Treatment Appendix Table 1. Neuropsychological Performance Appendix Table 2. Results of Sensitivity Analysis by Generalized Estimating Equation Method Appendix Table 3. Status and Ethnicity of Patients Requiring Modification of Treatment Prospective Evaluations Median NPZ3 scores improved in both groups during the study, with the greatest change occurring in the first week of treatment (Figure 2). No statistically significant differences in changes in neuropsychological performance were observed between the groups at any time point. We conducted conventional longitudinal analyses to further investigate differences in neuropsychological scores between the treatment groups. On the basis of these analyses, we had insufficient evidence to conclude that there were treatment differences (generalized estimating equation modeling in which treatment was the only independent variable and an exchangeable correlation structure was assumed, P= 0.176; Wei-Johnson test, P= 0.196). Multiple sensitivity analyses were performed, including single and multiple imputation methods, as-treated analyses, and 2 forms of worst-case scenarios. Details of these analyses are shown in Appendix Table 1. Multiple imputation and as-treated analyses generally provided similar results to observed data at specific weeks; however, the worst-case analyses at weeks 4, 12, and 24 displayed significant differences between groups. These results suggest that differences between groups might exist if the worst-case scenario were true, that is, if patients without data who were receiving efavirenz had wors

Effects of early versus delayed initiation of antiretroviral treatment on clinical outcomes of HIV-1 infection: results from the phase 3 HPTN 052 randomised controlled trial

BACKGROUND Use of antiretroviral treatment for HIV-1 infection has decreased AIDS-related morbidity and mortality and prevents sexual transmission of HIV-1. However, the best time to initiate antiretroviral treatment to reduce progression of HIV-1 infection or non-AIDS clinical events is unknown. We reported previously that early antiretroviral treatment reduced HIV-1 transmission by 96%. We aimed to compare the effects of early and delayed initiation of antiretroviral treatment on clinical outcomes. METHODS The HPTN 052 trial is a randomised controlled trial done at 13 sites in nine countries. We enrolled HIV-1-serodiscordant couples to the study and randomly allocated them to either early or delayed antiretroviral treatment by use of permuted block randomisation, stratified by site. Random assignment was unblinded. The HIV-1-infected member of every couple initiated antiretroviral treatment either on entry into the study (early treatment group) or after a decline in CD4 count or with onset of an AIDS-related illness (delayed treatment group). Primary events were AIDS clinical events (WHO stage 4 HIV-1 disease, tuberculosis, and severe bacterial infections) and the following serious medical conditions unrelated to AIDS: serious cardiovascular or vascular disease, serious liver disease, end-stage renal disease, new-onset diabetes mellitus, and non-AIDS malignant disease. Analysis was by intention-to-treat. This trial is registered with ClinicalTrials.gov, number NCT00074581. FINDINGS 1763 people with HIV-1 infection and a serodiscordant partner were enrolled in the study; 886 were assigned early antiretroviral treatment and 877 to the delayed treatment group (two individuals were excluded from this group after randomisation). Median CD4 counts at randomisation were 442 (IQR 373-522) cells per μL in patients assigned to the early treatment group and 428 (357-522) cells per μL in those allocated delayed antiretroviral treatment. In the delayed group, antiretroviral treatment was initiated at a median CD4 count of 230 (IQR 197-249) cells per μL. Primary clinical events were reported in 57 individuals assigned to early treatment initiation versus 77 people allocated to delayed antiretroviral treatment (hazard ratio 0·73, 95% CI 0·52-1·03; p=0·074). New-onset AIDS events were recorded in 40 participants assigned to early antiretroviral treatment versus 61 allocated delayed initiation (0·64, 0·43-0·96; p=0·031), tuberculosis developed in 17 versus 34 patients, respectively (0·49, 0·28-0·89, p=0·018), and primary non-AIDS events were rare (12 in the early group vs nine with delayed treatment). In total, 498 primary and secondary outcomes occurred in the early treatment group (incidence 24·9 per 100 person-years, 95% CI 22·5-27·5) versus 585 in the delayed treatment group (29·2 per 100 person-years, 26·5-32·1; p=0·025). 26 people died, 11 who were allocated to early antiretroviral treatment and 15 who were assigned to the delayed treatment group. INTERPRETATION Early initiation of antiretroviral treatment delayed the time to AIDS events and decreased the incidence of primary and secondary outcomes. The clinical benefits recorded, combined with the striking reduction in HIV-1 transmission risk previously reported, provides strong support for earlier initiation of antiretroviral treatment. FUNDING US National Institute of Allergy and Infectious Diseases.

Pharmacogenetics of Plasma Efavirenz Exposure after Treatment Discontinuation: An Adult AIDS Clinical Trials Group Study

BACKGROUND Efavirenz has a long plasma half-life and a low genetic barrier to resistance. Simultaneously stopping treatment with all agents in efavirenz-containing regimens may result in functional efavirenz monotherapy that selects for drug-resistant human immunodeficiency virus type 1. Lower plasma efavirenz clearance is associated with a cytochrome P450 2B6 gene (CYP2B6) polymorphism (516G-->T) that is more frequent among African American individuals than among European American individuals. METHODS We characterized relationships between this polymorphism and predicted plasma efavirenz concentration-time profiles after discontinuation of therapy with use of data obtained from subjects receiving therapy. Pharmacokinetic parameters were estimated using population-based methods. Concentrations after discontinuation of therapy were predicted from subject-specific estimates. RESULTS. Median estimated efavirenz half-lives were 23, 27, and 48 h for patients with CYP2B6 position 516 GG (78 patients), GT (60), and TT (14) genotypes, respectively (P<.001). After therapy was stopped, plasma efavirenz concentrations in patients with GG, GT, and TT genotypes were predicted to exceed 46.7 ng/mL (the estimated protein-adjusted 95% inhibitory concentration for wild-type virus) for a median of 5.8 days (interquartile range [IQR], 4.4-8.3 days), 7.0 days (IQR, 5.0-8.0 days), and 14 days (IQR, 11.1-21.2 days), respectively (P<.001). Plasma efavirenz levels were predicted to exceed 46.7 ng/mL for >21 days in 5% of subjects with GG genotype, 5% of subjects with GT genotype, and 29% of subjects with TT genotype. CONCLUSIONS The CYP2B6 position 516 TT genotype or a prolonged measured elimination half-life may predict increased risk of developing drug resistance among patients who discontinue efavirenz-containing regimens. This has implications for strategies to safely discontinue antiretroviral regimens while avoiding the emergence of drug resistance.

Highly Active Antiretroviral Therapy and Adverse Birth Outcomes Among HIV-Infected Women in Botswana

BACKGROUND It is unknown whether adverse birth outcomes are associated with maternal highly active antiretroviral therapy (HAART) in pregnancy, particularly in resource-limited settings. METHODS We abstracted obstetrical records at 6 sites in Botswana for 24 months. Outcomes included stillbirths (SBs), preterm delivery (PTD), small for gestational age (SGA), and neonatal death (NND). Among human immunodeficiency virus (HIV)-infected women, comparisons were limited to HAART exposure status at conception, and those with similar opportunities for outcomes. Comparisons were adjusted for CD4(+) lymphocyte cell count. RESULTS Of 33,148 women, 32,113 (97%) were tested for HIV, of whom 9504 (30%) were HIV infected. Maternal HIV was significantly associated with SB, PTD, SGA, and NND. Compared with all other HIV-infected women, those continuing HAART from before pregnancy had higher odds of PTD (adjusted odds ratio [AOR], 1.2; 95% confidence interval [CI], 1.1, 1.4), SGA (AOR, 1.8; 95% CI, 1.6, 2.1) and SB (AOR, 1.5; 95% CI, 1.2, 1.8). Among women initiating antiretroviral therapy in pregnancy, HAART use (vs zidovudine) was associated with higher odds of PTD (AOR, 1.4; 95% CI, 1.2, 1.8), SGA (AOR, 1.5; 95% CI, 1.2, 1.9), and SB (AOR, 2.5; 95% CI, 1.6, 3.9). Low CD4(+) was independently associated with SB and SGA, and maternal hypertension during pregnancy with PTD, SGA, and SB. CONCLUSIONS HAART receipt during pregnancy was associated with increased PTD, SGA, and SB.

Pre-existing Minority Drug-Resistant HIV-1 Variants, Adherence, and Risk of Antiretroviral Treatment Failure

BACKGROUND The clinical relevance of detecting minority drug-resistant human immunodeficiency virus type 1 (HIV-1) variants is uncertain. METHODS To determine the effect of pre-existing minority nonnucleoside reverse-transcriptase inhibitor (NNRTI)-resistant variants on the risk of virologic failure, we reanalyzed a case-cohort substudy of efavirenz recipients in AIDS Clinical Trials Group protocol A5095. Minority K103N or Y181C populations were determined by allele-specific polymerase chain reaction in subjects without NNRTI resistance by population sequencing. Weighted Cox proportional hazards models adjusted for recent treatment adherence estimated the relative risk of virologic failure in the presence of NNRTI-resistant minority variants. RESULTS The evaluable case-cohort sample included 195 subjects from the randomly selected subcohort (51 with virologic failure, 144 without virologic failure), plus 127 of the remaining subjects who experienced virologic failure. Presence of minority K103N or Y181C mutations, or both, was detected in 8 (4.4%), 54 (29.5%), and 11 (6%), respectively, of 183 evaluable subjects in the random subcohort. Detection of minority Y181C mutants was associated with an increased risk of virologic failure in the setting of recent treatment adherence (hazard ratio, 3.45 [95% confidence interval, 1.90-6.26]) but not in nonadherent subjects (hazard ratio, 1.39 [95% confidence interval, 0.58-3.29]). Of note, 70% of subjects with minority Y181C variants achieved long-term viral suppression. CONCLUSIONS In adherent patients, pre-existing minority Y181C mutants more than tripled the risk of virologic failure of first-line efavirenz-based antiretroviral therapy. CLINICAL TRIALS REGISTRATION NCT00013520.

Preexisting Resistance to Nonnucleoside Reverse-Transcriptase Inhibitors Predicts Virologic Failure of an Efavirenz-Based Regimen in Treatment-Naive HIV-1–Infected Subjects

A case-cohort study was used to determine the effect of baseline nonnucleoside reverse-transcriptase inhibitor (NNRTI) resistance, as assessed by viral genotyping, on the response to efavirenz-containing regimens in AIDS Clinical Trials Group A5095. The sample included a random cohort of efavirenz-treated subjects plus unselected subjects who experienced virologic failure. Of 220 subjects in the random cohort, 57 (26%) had virologic failure. The prevalence of baseline NNRTI resistance was 5%. The risk of virologic failure for subjects with baseline NNRTI resistance was higher than that for subjects without such resistance (hazard ratio 2.27 [95% confidence interval], 1.15-4.49; P = .018). These results support resistance testing before starting antiretroviral therapy.

Efficacy and Tolerability of 3 Nonnucleoside Reverse Transcriptase Inhibitor–Sparing Antiretroviral Regimens for Treatment-Naive Volunteers Infected With HIV-1: A Randomized, Controlled Equivalence Trial

Background Non-nucleoside reverse transcriptase (NNRTI) inhibitor-based antiretroviral therapy is not suitable for all treatment-naive HIV-infected persons.BACKGROUND Nonnucleoside reverse transcriptase inhibitor-based antiretroviral therapy is not suitable for all treatment-naive HIV-infected persons. OBJECTIVE To evaluate 3 nonnucleoside reverse transcriptase inhibitor-sparing initial antiretroviral regimens to show equivalence for virologic efficacy and tolerability. DESIGN A phase 3, open-label study randomized in a 1:1:1 ratio with follow-up for at least 96 weeks. (ClinicalTrials.gov: NCT00811954). SETTING 57 sites in the United States and Puerto Rico. PATIENTS Treatment-naive persons aged 18 years or older with HIV-1 RNA levels greater than 1000 copies/mL without resistance to nucleoside reverse transcriptase inhibitors or protease inhibitors. INTERVENTION Atazanavir, 300 mg/d, with ritonavir, 100 mg/d; raltegravir, 400 mg twice daily; or darunavir, 800 mg/d, with ritonavir, 100 mg/d, plus combination emtricitabine, 200 mg/d, and tenofovir disoproxil fumarate, 300 mg/d. MEASUREMENTS Virologic failure, defined as a confirmed HIV-1 RNA level greater than 1000 copies/mL at or after 16 weeks and before 24 weeks or greater than 200 copies/mL at or after 24 weeks, and tolerability failure, defined as discontinuation of atazanavir, raltegravir, or darunavir for toxicity. A secondary end point was a combination of virologic efficacy and tolerability. RESULTS Among 1809 participants, all pairwise comparisons of incidence of virologic failure over 96 weeks showed equivalence within a margin of equivalence defined as -10% to 10%. Raltegravir and ritonavir-boosted darunavir were equivalent for tolerability, whereas ritonavir-boosted atazanavir resulted in a 12.7% and 9.2% higher incidence of tolerability discontinuation than raltegravir and ritonavir-boosted darunavir, respectively, primarily because of hyperbilirubinemia. For combined virologic efficacy and tolerability, ritonavir-boosted darunavir was superior to ritonavir-boosted atazanavir, and raltegravir was superior to both protease inhibitors. Antiretroviral resistance at the time of virologic failure was rare but more frequent with raltegravir. LIMITATION The trial was open-label, and ritonavir was not provided. CONCLUSION Over 2 years, all 3 regimens attained high and equivalent rates of virologic control. Tolerability of regimens containing raltegravir or ritonavir-boosted darunavir was superior to that of the ritonavir-boosted atazanavir regimen. PRIMARY FUNDING SOURCE National Institute of Allergy and Infectious Diseases.

Effect of CYP2B6, ABCB1, and CYP3A5 polymorphisms on efavirenz pharmacokinetics and treatment response: an AIDS Clinical Trials Group Study.

In AIDS Clinical Trials Group protocols 384, A5095, and A5097s, we characterized relationships between 22 polymorphisms in CYP2B6, ABCB1, and CYP3A5; plasma efavirenz exposure; and/or treatment responses. A stepwise logistic regression procedure selected polymorphisms associated with reduced drug clearance adjusted for body mass index and the composite CYP2B6 516/983 genotype. Relationships between selected polymorphisms and treatment responses were characterized by competing risk methodology. Association analyses involved 821 individuals (317 for pharmacokinetics and 643 for treatment response). Models that included CYP2B6 516/983 genotype best predicted pharmacokinetics. Slow-metabolizer genotypes were associated with increased central nervous system events among white participants and decreased virologic failure among black participants.