A Double-Blind, Placebo-Controlled Trial of Maraviroc in Treatment-Experienced Patients Infected with Non-R5 HIV-1

Transcription

1 MAJOR ARTICLE A Double-Blind, Placebo-Controlled Trial of Maraviroc in Treatment-Experienced Patients Infected with Non-R5 HIV-1 Michael Saag, 1 James Goodrich, 2 Gerd Fätkenheuer, 3 Bonaventura Clotet, 4 Nathan Clumeck, 5 John Sullivan, 6 Mike Westby, 6 Elna van der Ryst, 6 and Howard Mayer, 2 for the A Study Group 1 School of Medicine, University of Alabama, Birmingham; 2 Pfizer Global Research and Development, New London, Connecticut; 3 Universitätsklinik Köln, Köln, Germany; 4 Hospital Universitario Germans Trias i Pujol, Badalona, Barcelona, Spain; 5 Le Centre Hospitalier Universitaire St-Pierre, Brussels, Belgium; and 6 Pfizer Global Research and Development, Sandwich, United Kingdom Background. Maraviroc, a CCR5 antagonist, is active against R5 but not X4 or dual- or mixed-tropic strains of human immunodeficiency virus type 1 (HIV-1). A phase 2b study was conducted to determine the safety and efficacy of maraviroc in combination with optimized background therapy in treatment-experienced patients infected with dual- or mixed-tropic HIV-1. Methods. Treatment-experienced patients with an HIV-1 RNA level 5000 copies/ml who had received 3 classes of drugs and/or were infected with virus resistant to 2 drug classes and were infected with non-r5 HIV-1 were randomized to receive optimized background therapy plus maraviroc (once or twice daily) or placebo. The primary end point was change in HIV-1 RNA level from baseline to 24 weeks. Results. Among 167 patients infected with dual- or mixed-tropic HIV-1, baseline mean HIV-1 RNA levels were 5 log 10 copies/ml and median CD4 cell counts were 50 cells/ L. From baseline to 24 weeks, patients who received placebo demonstrated a mean decrease in HIV-1 RNA levels of 0.97 log 10 copies/ml, compared with mean decreases of 0.91 and 1.20 log 10 copies/ml for those who received maraviroc once (P.83) or twice (P.38) daily, respectively. Mean increases in CD4 cell counts from baseline were 36 cells/ L for patients who received placebo, 60 cells/ L among patients who received maraviroc once daily, and 62 cells/ L among patients who received maraviroc twice daily. The incidences of serious adverse events were similar among groups. Conclusions. In this exploratory study involving extensively treatment-experienced patients with advanced, non-r5 HIV-1 infection, neither superiority nor noninferiority was statistically demonstrated for either maraviroc dosage compared with placebo at 24 weeks of treatment. Trial registration. Clinicaltrials.gov identifier NCT Human immunodeficiency virus type 1 (HIV-1) strains are categorized as R5 (CCR5 tropic), X4 (CXCR4 tropic), or R5X4 (dual-tropic strains that use both CCR5 Received 16 July 1008; accepted 22 December 2008; electronically published 7 May Potential conflicts of interest: M.S. has received consultancy fees from Avexa, Boehringer-Ingelheim, Bristol Myers Squibb, Gilead, GlaxoSmithKline, Merck Sharp and Dohme, Monogram Biosciences, Panacos, Pfizer, Progenics, Roche, Tibotec, and Virco and has received grant support from Achillion Pharmaceuticals, Boehringer- Ingelheim, Gilead Sciences, GlaxoSmithKline, Merck Sharp and Dohme, Panacos, Pfizer, Progenics, Roche, Serono, and Tibotec. J.G. is an employee of Pfizer. G.F. has received consultancy fees from Gilead, GlaxoSmithKline, Monogram, Pfizer, Roche, Schering-Plough, Tibotec, and Bristol Myers Squibb; has received lecture fees from Abbott, Boehringer-Ingelheim, Bristol Myers Squibb, Gilead, GlaxoSmithKline, Merck Sharp and Dohme, Pfizer, Schering-Plough, and Tibotec; and has received grant support from Pfizer. B.C. has received consultancy fees from Janssen, Merck Sharp and Dohme, Gilead, and Bristol Myers Squibb and has received lecture fees from Janssen, Merck Sharp, and Dohme and Gilead. N.C. has received consultancy fees from Gilead, Pfizer, Abbott, and Merck Sharp and Dohme. J.S., M.W., E.v.d.R., and H.M. are employees of Pfizer and hold stock options in Pfizer. The Journal of Infectious Diseases 2009; 199: by the Infectious Diseases Society of America. All rights reserved /2009/ $15.00 DOI: / and CXCR4) [1]. Patients infected with dual- or mixedtropic HIV-1 harbor a mixture of R5, X4, and/or dualtropic viruses [1]. Most extensively antiretroviralexperienced patients continue to be infected with only R5 virus [2 4]. CCR5 antagonists only inhibit R5 strains; X4 and R5X4 strains (i.e., CXCR4-using strains) can infect cells in the presence or absence of a CCR5 antagonist. Maraviroc, an oral CCR5 antagonist, has demonstrated potent activity in vitro and in vivo against R5 HIV-1, including variants that are resistant to multiple drug classes [5, 6]. Analyses from 2 phase 3 studies of maraviroc involving treatment-experienced patients infected Presented in part: XVI International AIDS Conference, Toronto, Canada, August 2006 (presentation THLB0215). Financial support: Pfizer Global Research and Development. Editorial support was provided by Health Interactions and was funded by Pfizer. Reprints or correspondence: Michael S. Saag, MD, Director of AIDS Outpatient Clinic, University of Alabama at Birmingham School of Medicine, th St. S., Birmingham, AL (msaag@uab.edu) JID 2009:199 (1 June) Saag et al.

2 with R5 virus have clearly demonstrated efficacy and safety during 48 weeks of treatment [7, 8]. Evaluation of the safety of a CCR5 antagonist in patients infected with a mixture of viruses that may include CXCR4-using variants is important, because such patients could inadvertently receive a CCR5 antagonist. We present results from a study evaluating the safety of maraviroc treatment in patients infected with non-r5 virus and the efficacy of maraviroc in patients infected with dual- or mixed-tropic virus. METHODS Study design and conduct. A is a randomized, placebo-controlled, double-blind, phase 2b trial with a primary end point at 24 weeks conducted at 76 centers in Australia, Europe, Canada, and the United States. Patients were excluded from the study if they had only R5 HIV-1 detected during screening (which was performed using the original Trofile assay [Monogram Biosciences]). Eligibility criteria included HIV-1 RNA level 5000 copies/ml, resistance to 2 antiretroviral drug classes approved at the time of study initiation (i.e., nucleoside analogues, nonnucleoside reverse-transcriptase inhibitors, protease inhibitors, and fusion inhibitors) or 3 months experience receiving 1 agent from 3 of these drug classes, and detection of X4 or dual- or mixed-tropic HIV-1 or indeterminate tropism at screening (indeterminate tropism was defined as assay results that failed to determine tropism in 3 consecutive tests of the same specimen). Patients were excluded if, on the basis of resistance testing results, enfuvirtide or 1 nonnucleoside reverse-transcriptase inhibitors or protease inhibitors were not available as active treatment options. Other criteria were identical to those of the MOTIVATE studies [7], as were the study design, conduct, monitoring, and statistical analyses, except as indicated. Written informed consent was obtained from all patients, and the protocol was approved by institutional review boards or independent ethics committees at all study centers. Patients were randomized 1:1:1 to receive maraviroc at 300 mg once daily, maraviroc at 300 mg twice daily, or placebo; all patients received a selected optimized background therapy that included 3 6 antiretroviral drugs (with or without low-dose ritonavir). Patients who were currently receiving potent CYP3A4 inhibitors, such as protease inhibitors (excluding tipranavir) or delavirdine, received 150-mg doses of maraviroc. After initiation of study treatment (baseline), study visits were scheduled at weeks 4, 8, 16, 24, 32, 40, and 48. In addition to measurement of CD4 and CD8 cell counts and HIV-1 RNA level, study evaluations included physical examination; assessment of vital signs, symptoms, and adverse events; serum chemistry and hematology measurements; and review of concomitant medications. A plasma sample was collected for determination of HIV-1 coreceptor tropism from patients who experienced protocol-defined treatment failure, from patients who had HIV-1 RNA levels 500 copies/ml at weeks 4, 8, 16, 32, and 40, and from patients who discontinued treatment. HIV-1 RNA was isolated from plasma for drug resistance testing only for patients who experienced protocol-defined treatment failure. A fasting metabolic assessment (i.e., measurement of total cholesterol, high-density liprotein, low-density lipoprotein, triglycerides, glucose, and glycosylated hemoglobin levels), 12-lead electrocardiogram, and orthostatic blood-pressure monitoring were performed at baseline and at weeks 24 and 48, or at discontinuation. Treatment failure was prospectively defined as reaching 1of the following 4 HIV-1 RNA measurement end points (confirmed by consecutive measurements meeting the end point within 14 days): (1) an increase in HIV-RNA level to 3 times baseline level, at or after week 2; (2) a decrease in HIV-1 RNA level from baseline of 0.5 log 10 copies/ml, at or after week 8; (3) a decrease in HIV-1 RNA from baseline of 1.0 log 10 copies/ml at or after week 8 in a patient who had previously achieved a decrease from baseline of 2.0 log 10 copies/ml; or (4) an increase in HIV-1 RNA level to 5000 copies/ml in a patient who had previously had levels 400 copies/ml on 2 consecutive visits. Subjects who discontinued treatment because of treatment failure or for other reasons were followed up according to the study protocol until week 48. The primary end point was the change in HIV-1 RNA level (measured as log 10 -transformed copies/ml) in plasma from baseline to 24 weeks. Secondary efficacy end points were identical to those used in the MOTIVATE studies [7]. For the primary end point, the differences in least-square mean HIV-1 RNA levels and a 2-sided 97.5% confidence interval (CI) between groups are presented. For secondary efficacy end points, 2-sided 95% CIs between groups are presented. This study was part of a regulatory submission; therefore, the primary end point was adjusted for multiple comparisons to maintain an overall significance level of.05. Safety data were analyzed through 24 weeks for all patients who received 1 dose of study medication, regardless of the result of HIV-1 tropism screening, with use of methodologies identical to those used in the MOTIVATE studies [7]. Data were collected and analyzed by the study sponsor according to a predefined statistical analysis plan and were reviewed periodically by a data and safety monitoring board. Statistical analysis. The primary efficacy population included all randomized patients infected with dual- or mixedtropic HIV-1 at screening who received 1 dose of study medication. Based on the assumption that 79% of patients with non-r5 virus would have dual- or mixed-tropic HIV-1 (the remainder having an X4 tropism result or an indeterminate tropism result), 192 patients would have to be randomized to include 150 patients infected with dual- or mixed-tropic HIV-1 in the study. Assuming a standard deviation of 0.8 with a 2-sided significance level of.025 (using Bonferroni adjustment for multiple comparisons) and assuming 50 patients infected with dual- Maraviroc in Patients with Non-R5 HIV Infection JID 2009:199 (1 June) 1639

3 or mixed-tropic HIV-1 per treatment group, there was 80% power to detect a difference in HIV-1 RNA level (the primary efficacy end point) of 0.5 log 10 copies/ml between each maraviroc arm and the placebo arm. Using the same assumptions about the study population with a 1-sided significance level of.0125 and assuming that maraviroc treatment groups would demonstrate decreases in HIV RNA levels that were 0.25 log 10 copies/ml greater than that demonstrated by the placebo group, there was 80% power to demonstrate noninferiority with a margin of 0.25 log 10 copies/ml. An analysis of covariance model was used, with HIV-1 RNA level at screening (randomization stratification factor), enfuvirtide use as part of optimized background therapy (randomization stratification factor), and treatment arm as the main effects. Efficacy data were analyzed using an intention-to-treat approach. Change from baseline was imputed as 0 for patients who discontinued treatment for any reason and for patients for whom a baseline value was missing. Further sensitivity analyses were performed using the primary end point, including analysis with a last-observation-carriedforward approach to discontinuations and a nonparametric analysis in which patients who discontinued treatment were ranked lower than those who completed treatment. The percentage of patients with an undetectable viral load ( 50 copies/ml) was summarized for the primary efficacy population; values for patients who had missing data or who discontinued treatment were imputed as 0. Subgroup analyses of the proportion of patients with undetectable viral load, according to randomization stratification criteria and by overall susceptibility score at screening, were also conducted, imputing values of 0 for patients who had missing data or discontinued treatment. Least-squares mean change in CD4 cell count from baseline was analyzed using an analysis of covariance model, with baseline CD4 cell count, enfuvirtide use, and treatment arm as the main effects, and using a last-observation-carried-forward approach. Further sensitivity analyses were performed on CD4 cell count data, including an analysis that imputed values of 0 for patients for whom data was missing or who discontinued treatment. Time to protocol-defined treatment failure was summarized using Kaplan-Meier curves. The difference between each of the maraviroc arms and the placebo arm was analyzed using a stratified log-rank test. RESULTS Study population. Of 190 patients who were randomized, 186 received 1 dose of study medication (figure 1). Demographic and baseline clinical characteristics were similar across treatment arms (table 1). Mean time since receipt of HIV diagnosis was 13.5 years (range, years); 57 (31%) of 186 patients had previously received enfuvirtide, and the median baseline CD4 cell count was 50 cells/ L, which is indicative of a population with advanced HIV infection. At screening, 167 (90%) of 186 patients who had received 1 dose of study drug were infected with dual- or mixed-tropic HIV-1 (the primary efficacy population). In addition, 1 patient, who was randomized in error, had only R5 HIV-1 detected at screening, and the remainder had either an X4 tropism result or an indeterminate tropism result (table 1). Of the 167 patients in the primary efficacy population, 98 (59%) had a overall susceptibility score of 2 at screening, and 94 (56%) received enfuvirtide as part of optimized background therapy. Premature discontinuation and treatment duration. Among the 186 patients who were randomized and treated, 38 (61%) of 62 patients in the placebo arm, 38 (60%) of 63 patients in the maraviroc once-daily arm, and 32 (52%) of 61 patients in the maraviroc twice-daily arm discontinued study treatment (maraviroc or placebo) by week 24, which was the primary efficacy analysis timepoint. The most common reason for discontinuation in all groups was lack of efficacy (determined by the investigator, guided by the 4 prespecified criteria for treatment failure). At week 24, the median duration of treatment was 127 days for patients who received placebo, compared with 119 days for patients who received maraviroc once daily and 176 days for patients who received maraviroc twice daily. Time to protocol-defined treatment failure was not significantly different among groups (maraviroc once daily vs. placebo, P.75; maraviroc twice daily vs. placebo, P.25) (figure 2A). Efficacy. The least-squares mean HIV-1 RNA level in plasma from baseline to week 24 decreased 0.97 log 10 copies/ml in the placebo group, 0.91 log 10 copies/ml in the maraviroc once-daily group (difference vs. placebo, 0.06 log 10 copies/ml; 97.5% CI, 0.53 to 0.64 log 10 copies/ml; P.83), and 1.20 log 10 in the maraviroc twice-daily group (difference vs. placebo, 0.23 log 10 copies/ml; 97.5% CI, 0.83 to 0.36 log 10 copies/ml; P.38) (table 2). The statistical analysis did not indicate superiority or noninferiority for either maraviroc dosage, compared with placebo. Results at week 48 were similar; mean HIV-1 RNA levels decreased 0.84 log 10 copies/ml in the placebo group, 0.62 log 10 copies/ml in the maraviroc once-daily group (difference vs. placebo, 0.23; 97.5% CI, 0.35 to 0.81 log 10 copies/ml), and 1.11 log 10 copies/ml in the maraviroc twice-daily group (difference vs. placebo, 0.26 log 10 copies/ml; 97.5% CI, 0.86 to 0.33 log 10 copies/ml). The sensitivity analyses yielded findings similar to those of the primary analysis. Using the last-observation-carried-forward approach, the least-squares mean HIV-1 RNA level in plasma from baseline to week 24 decreased 1.20 log 10 copies/ml in the placebo group, 1.18 log 10 copies/ml in the maraviroc once-daily group (difference vs. placebo, log 10 copies/ml; 97.5% CI, to log 10 copies/ml), and 1.50 log 10 copies/ml in the maraviroc twice-daily group (difference vs. placebo, log 10 copies/ml; 97.5% CI, to log 10 copies/ml) JID 2009:199 (1 June) Saag et al.

4 Figure 1. Disposition of study participants at 24 weeks (primary efficacy analysis timepoint). *Number of patients who had a dual- or mixed-tropic HIV-1 tropism result at screening (primary efficacy population). Lack of efficacy as determined by the investigator, but guided by the 4 prespecified criteria for treatment failure (see Methods for details). A total of 7 deaths were recorded during the study, but only 5 of these deaths were listed by investigators as the reason for discontinuation of study treatment. The results of a sensitivity analysis that used a stratified Wilcoxon test, with patients who discontinued treatment ranked below those who completed treatment, were consistent with the results of the primary analysis. A slightly higher proportion of patients in the maraviroc twice-daily arm (14 [27%] of 52) achieved undetectable viremia (HIV-1 RNA level 50 copies/ml) at 24 weeks, compared with 12 [21%] of 57 patients in the maraviroc once-daily and 9 (16%) of 58 patients in the placebo group; however, these differences were not statistically significant (table 2; figure 2B). Similar results were observed at 48 weeks, with 27% of patients in the maraviroc twice-daily arm, 18% of patients in the maraviroc once-daily arm, and 22% of patients in the placebo arm achieving undetectable HIV-1 RNA levels. The proportion of patients whose optimized background therapy included enfuvirtide and who achieved undetectable viremia at 24 weeks was somewhat greater in the once- and twicedaily maraviroc arms (7 [20%] of 35 and 7 [24%] of 29 patients, respectively) than in the placebo arm (3 [9%] of 34 patients), although the numbers were small and statistically significant differences were not observed. Similarly, when analyzed by HIV-1 RNA level at screening, a somewhat greater proportion of patients with a lower viral load at screening ( 100,000 copies/ml) who were randomized to receive maraviroc achieved a viral load 50 copies/ml at 24 weeks (7 [27%] of 26 and 8 [35%] of 23 patients in the once- and twice-daily arms, respectively), compared with those who received placebo (4 [15%] 26 patients). The percentage of patients with HIV-1 RNA levels 100,000 copies/ml at screening who achieved HIV-1 RNA levels 50 copies/ml ranged from 16% 17% in the maraviroc once-daily and placebo groups to 21% in the maraviroc twice-daily group. Just 1 patient with a overall susceptibility score 2 at screening (in the maraviroc once-daily group) achieved undetectable viremia at week 24. The proportions of patients with an overall susceptibility score of 2 4 who achieved undetectable viremia were slightly greater in the maraviroc once- and twice-daily groups (11 [31%] of 35 and 14 [36%] of 39 patients, respectively), compared with the placebo group (9 [23%] of 40 patients). Maraviroc in Patients with Non-R5 HIV Infection JID 2009:199 (1 June) 1641

5 Table 1. Demographic and baseline clinical characteristics of all patients who received 1 dose of study medication. Treatment group Characteristic Placebo Maraviroc once daily Maraviroc twice daily All treated patients No. of patients Age, mean years (range) 45 (23 65) 43 (16 59) 43 (16 62) Male sex 53 (85.5) 53 (84.1) 55 (90.2) Race a White 40 (64.5) 46 (73.0) 44 (72.1) Black 18 (29.0) 17 (27.0) 13 (21.3) Asian 3 (4.8) 0 (0.0) 1 (1.6) Other 1 (1.6) 0 (0.0) 3 (4.9) Duration since diagnosis of HIV-1 infection, mean years (range) 12.9 ( ) 13.8 ( ) 13.8 ( ) Prior enfuvirtide experience 22 (35.5) 18 (28.6) 17 (27.9) HIV-1 tropism at screening Not phenotyped or not reported 2 (3.2) 3 (4.8) 5 (8.2) R5 0 (0.0) 1 (1.6) 0 (0.0) X4 2 (3.2) 2 (3.2) 4 (6.6) Dual or mixed tropism 58 (93.5) 57 (90.5) 52 (85.2) Treated patients with dual- or mixed-tropic HIV-1 at screening No. of patients HIV-1 RNA load, b mean log 10 copies/ml CD4 cell count, b median cells/ L HIV-1 RNA level at screening 100,000 copies/ml 25 (46.3) c 26 (45.6) 23 (44.2) 100,000 copies/ml 29 (53.7) c 31 (54.4) 29 (55.8) Genotypic susceptibility score at screening 0 4 (6.9) 5 (8.8) 6 (11.5) 1 19 (32.8) 30 (52.6) 18 (34.6) 2 7 (12.1) 12 (21.1) 10 (19.2) 3 27 (46.6) 9 (15.8) 17 (32.7) Missing 1 (1.7) 1 (1.8) 1 (1.9) Phenotypic susceptibility score at screening 0 2 (3.4) 0 (0.0) 2 (3.8) 1 13 (22.4) 12 (21.1) 6 (11.5) 2 11 (19.0) 23 (40.4) 13 (25.0) 3 31 (53.4) 21 (36.8) 30 (57.7) Missing 1 (1.7) 1 (1.8) 1 (1.9) Overall susceptibility score at screening d 0 2 (3.4) 1 (1.8) 2 (3.8) 1 15 (25.9) 19 (33.3) 11 (21.2) 2 13 (22.4) 21 (36.8) 14 (26.9) 3 27 (46.6) 15 (26.3) 24 (46.2) Missing 1 (1.7) 1 (1.8) 1 (1.9) PI and/or delavirdine in OBT e 53 (98.1) f 52 (91.2) 48 (92.3) Enfuvirtide in OBT 30 (55.6) f 35 (61.4) 29 (55.8) NOTE. Data are no. (%) of patients, unless otherwise indicated. All patients received optimized background therapy (OBT) in addition to study medication. HIV, human immunodeficiency virus, PI, protease inhibitor. a Patients of Hispanic or Latino ethnicity (15.6% of patients) could be included in any of the 3 racial designation groups included in this table. b Baseline values for each patient were calculated on the basis of up to 3 measurements made before administration of study medication (i.e., at screening, randomization, and baseline visit). c HIV-1 RNA level stratification data were missing for 4 patients. d Susceptibility score was based on an algorithm that assigns a net assessment of phenotypic and genotypic information for each OBT medication with use of a binary system. e The maraviroc dose was 150 mg for patients who were concurrently receiving a PI (excluding tipranavir) and/or delavirdine as part of their OBT; otherwise, the maraviroc dose was 300 mg. Only data for patients receiving a PI other than tipranavir are included. f Data were missing for 4 patients. 1642

6 Figure 2. Treatment response in the primary efficacy population (i.e., all patients with dual- or mixed-tropic virus who received 1 dose of study medication) over 24 weeks (primary efficacy analysis timepoint). A, Kaplan-Meier plot of time to protocol-defined treatment failure. HIV-1 RNA value from baseline was imputed if data were missing or if patient discontinued before 24 weeks. B, Proportion of patients with HIV-1 RNA levels 50 copies/ml. C, Mean change in CD4 cell count from baseline (using last-observation-carried-forward approach). OBT, optimized background therapy. 1643

7 Table 2. Efficacy at 24 weeks (primary efficacy analysis timepoint) in patients infected with dual- or mixed-tropic human immunodeficiency virus type 1 (HIV-1) at screening (primary efficacy population). Variable Placebo (n 58) Treatment group Maraviroc once daily (n 57) Maraviroc twice daily (n 52) Maraviroc once daily vs. placebo Difference (95% CI) P Maraviroc twice daily vs. placebo Change from baseline in plasma HIV-1 RNA level, least-squares mean log 10 copies/ml a ( 0.53 to 0.64) b ( 0.83 to 0.36) b.38 HIV-1 RNA level 50 copies/ml, % of patients c ( 7 to 20) 11 ( 3 to26) Change in CD4 cell count, least-squares mean cells/ L d All patients ( 1 to 49) (1 52).04 Patients receiving blinded therapy at week 24 e Patients who discontinued because of treatment failure f NOTE. The maraviroc dose was 150 mg for patients who were concurrently receiving a protease inhibitor (excluding tipranavir) and/or delavirdine as part of their optimized background therapy; otherwise, the maraviroc dose was 300 mg. CI, confidence interval. a Patients who discontinued the study at or before week 24 for any reason had their missing values imputed as baseline values. b 97.5% CI. c Data were missing for 4 patients in the placebo group. d The last-observation-carried-forward analysis was used to impute missing values. e There were 24 patients in the placebo group, 23 patients in the maraviroc once-daily group, and 26 patients in the maraviroc twice-daily group. f There were 23 patients in the placebo group, 33 patients in the maraviroc once-daily group, and 21 patients in the maraviroc twice-daily group. P The change in CD4 cell count from baseline was a secondary end point. With use of the last-observation-carried-forward approach, the increase in CD4 cell count from baseline to 24 weeks was greater for both maraviroc treatment groups, compared with the placebo group (table 2; figure 2C). Unadjusted for multiple comparisons, the difference in the increase in CD4 cell count from baseline for the maraviroc twice-daily arm (62 cells/ L) versus the placebo arm (36 cells/ L) was statistically significant (difference, 27 cells/ L; 95% CI, 1 52 cells/ L; P.04). The magnitude of the difference in increase in CD4 cell count from baseline between the maraviroc twice-daily (47 cells/ L) and placebo arms (32 cells/ L) was lower when CD4 cell count values were analyzed using values of 0 for all patients who discontinued treatment or had missing data (difference, 14 cells/ L; 95% CI, 11 to 39 cells/ L). Mean CD4 cell counts increased slightly in all groups between weeks 24 and 48. By week 48, the increase from baseline was 51 cells/ L in the placebo group versus 65 and 78 cells/ Lin the maraviroc once- and twice-daily groups, respectively (using last-observation-carried-forward approach). At 48 weeks, there were no statistically significant differences in CD4 cell count response between treatment groups. A change in tropism result between screening and baseline was observed in 18 (11%) of 167 patients with dual- or mixedtropic HIV-1 at screening. Sixteen patients had results change from dual- or mixed tropic to R5, and 2 had results change to X4. Most patients (140 [84%] of 167) did not have a result for tropism tests at week 24, primarily because of discontinuation or having a viral load 500 copies/ml. Of the 75 patients with dual- or mixed-tropic virus at screening who experienced treatment failure and for whom a tropism test result was available, 31 (41%) had a different HIV-1 tropism result at the time of treatment failure. Patients who received maraviroc and experienced treatment failure were more likely to have X4 HIV-1 at treatment failure (12 [36%] of 33 and 12 [57%] of 21 patients who received maraviroc once and twice daily, respectively) than were those who received placebo and experienced treatment failure (2 [9%] of 23 patients). Increases in CD4 cell count from baseline to treatment failure in this subgroup with X4 HIV-1 (48 and 33 cells/ L for patients who received maraviroc once and twice daily, respectively) were similar to the increase in CD4 cell count observed among maraviroc-treated patients who experienced treatment failure with dual- or mixed-tropic virus (34 and 14 cells/ L for patients who received maraviroc once and twice daily, respectively). These increases were similar to the mean increase in CD4 cell count at 24 weeks of 36 cells/ L observed among patients in the placebo group. Safety. The most commonly reported adverse events that occurred during treatment across all groups were diarrhea, fatigue, headache, nausea, injection site reactions, pyrexia, and bronchitis; there was no clear pattern to their incidence (table 3). Study drug related discontinuation attributable to adverse events was uncommon, and the rates were similar across groups (3.2%, 0%, and 1.6% for the placebo, maraviroc once-daily, and maraviroc twice-daily groups, respectively) JID 2009:199 (1 June) Saag et al.

8 Table 3. Treatment-emergent adverse events among all patients who received 1 dose of study medication. Treatment group, no. (%) of patients Variable Placebo (n 62) Maraviroc once daily (n 63) Maraviroc twice daily (n 61) Experienced 1 adverse event (all grades) All causality 55 (88.7) 54 (85.7) 56 (91.8) Treatment related 38 (61.3) 28 (44.4) 30 (49.2) Grade 2 4 adverse events occurring in 5% of patients in any treatment group Bronchitis 0 (0.0) 1 (1.6) 4 (6.6) Diarrhea 5 (8.1) 8 (12.7) 5 (8.2) Fatigue 3 (4.8) 4 (6.3) 3 (4.9) Headache 3 (4.8) 4 (6.3) 3 (4.9) Injection site reaction 2 (3.2) 1 (1.6) 4 (6.6) Nausea 5 (8.1) 1 (1.6) 1 (1.6) Pyrexia 2 (3.2) 3 (4.8) 4 (6.6) Category C (AIDS-defining) adverse events Esophageal candidiasis 1 (1.6) 0 (0.0) 2 (3.3) Cytomegalovirus chorioretinitis 0 (0.0) 1 (1.6) 0 (0.0) Herpes simplex virus infection 0 (0.0) 1 (1.6) 0 (0.0) Histoplasmosis 0 (0.0) 1 (1.6) 0 (0.0) Mycobacterium avium infection 0 (0.0) 1 (1.6) 0 (0.0) Pneumocystis jiroveci pneumonia 0 (0.0) 3 (4.8) 1 (1.6) Pneumonia 1 (1.6) 0 (0.0) 0 (0.0) Encephalitis 1 (1.6) 0 (0.0) 0 (0.0) Total number of events Patients experiencing event 2 (3.2) 6 (9.5) 3 (4.9) Maximum liver function test values a Aspartate aminotransferase Grade 3, 5.0 to 10.0 ULN 2 (3.4) 1 (1.6) 1 (1.6) Grade 4, 10.0 ULN 0 (0.0) 1 (1.6) 0 (0.0) Alanine aminotransferase Grade 3, 5.0 to 10.0 ULN 2 (3.4) 1 (1.6) 0 (0.0) Grade 4, 10.0 ULN 1 (1.7) 0 (0.0) 0 (0.0) NOTE. The maraviroc dose was 150 mg for patients who were concurrently receiving a protease inhibitor (excluding tipranavir) and/or delavirdine as part of their optimized background therapy; otherwise, the maraviroc dose was 300 mg. ULN, upper limit of normal. a There were 58 patients in the placebo group, 63 patients in the maraviroc once-daily group, and 61 patients in the maraviroc twice-daily group who had 1 observation for the given liver function test while receiving study treatment or within 7 days of receiving the last dose of study medication. None of the 9 serious adverse events reported in each of the maraviroc arms and only 1 of the 10 reported in the placebo arm was considered to be treatment related. Seven deaths were recorded. Three deaths occurred among patients in the placebo group, which were attributed to (1) aggravated renal failure, ascites, hemothorax, and disease aggravation; (2) progressive multifocal leukoencephalopathy; and (3) multiple cerebral lesions. Two deaths occurred among patients in the maraviroc oncedaily group, which were attributed to (1) pneumonia and possible chest mass and (2) HIV disease progression. Two deaths also occurred among patients in the maraviroc twice-daily group, which were attributed to (1) bacterial pneumonia and (2) pneumocystis pneumonia. None of the deaths were considered to be treatment related. There were 3, 7, and 3 category C (AIDSdefining) events that occurred among patients in the placebo, maraviroc once-daily, and maraviroc twice-daily groups, respectively (table 3). During the treatment period, 1 patient in each treatment group received a diagnosis of malignancy: basal cell carcinoma (placebo), anal carcinoma (maraviroc once daily), and bowenoid papulosis (maraviroc twice daily). No cases of lymphoma were reported. The overall incidence of grade 3 or 4 laboratory abnormalities was similar across treatment arms: 17% in the placebo group, 19% in the maraviroc once-daily group, and 18% in the maraviroc twice-daily group. The incidence of liver function test abnormalities was low and similar between groups (table 3). No new or unique safety findings emerged between the 24- and 48- week timepoints. Maraviroc in Patients with Non-R5 HIV Infection JID 2009:199 (1 June) 1645

9 DISCUSSION The findings of this exploratory study demonstrate that the addition of maraviroc to optimized background therapy in an extensively treatment-experienced patient population infected with dual- or mixed-tropic HIV-1 does not confer a significant additional reduction in plasma viremia, compared with placebo, over 24 weeks (primary efficacy analysis timepoint) of treatment. Neither superiority nor noninferiority of either maraviroc dosage versus placebo was statistically demonstrated with respect to the primary virologic end point (HIV-1 RNA levels). This may reflect the fact that the variability in virological response was higher than had been assumed in the sample size calculation. The observation that the population with a low median baseline CD4 cell count who received maraviroc twice daily experienced a statistically significantly greater increase in mean CD4 cell count at 24 weeks than those who received placebo was unexpected, although this finding should be interpreted with caution, because the difference was not corrected for multiple comparisons and was not apparent at 48 weeks. The reason for the unexpected increase in CD4 cell counts among maraviroc recipients remains undetermined, but a possible explanation may be that the suppression of R5 virus is an important factor in determining immunologic recovery. Alternatively, a decrease in immune activation as a result of a CCR5 blockade, above and beyond that caused by antiviral activity, may be responsible for the release of CD4 cells from lymphoid reservoirs into plasma. The increase in CD4 cell population resulted mostly from an increase in total lymphocyte count rather than a specific effect on CD4 lymphocytes per se (data not shown), which suggests that maraviroc has an effect on inhibiting adhesion molecules elaborated within lymphatic tissue, resulting in the release of lymphocytes into the circulation. This premise is supported by recent work of Silvestri et al. [9], which demonstrated reduced CCR5 expression in gut-associated lymphoid tissue and reduced immune activation in several species of nonhuman primates with chronic, nonprogressing simian immunodeficiency virus infection. This hypothesis needs to be further validated. A third possibility is that CCR5 blockade may lead to prolongation of the lifespan of CD4 cells attributable to a reduction in apoptosis or other mechanisms. More patients in the maraviroc arms had X4 virus at time of treatment failure, compared with placebo recipients, which is consistent with the selective suppression of R5 virus in these patients. However, X4 virus did not appear to be associated with an adverse immunologic outcome in this patient population, because CD4 cell count increases for patients in the maraviroc arms did not differ with regard to tropism results at the time of treatment failure. Contrary to previous hypotheses, we saw no evidence of synergy or additive effects of maraviroc and enfuvirtide in this study population. Rather, maraviroc-enfuvirtide combination regimens perform as all other antiretroviral regimens, with activity linked to the number of active drugs in the regimen [10]. The adverse event profile of maraviroc was comparable to that of placebo in this study. Serious adverse events (including deaths), grade 3/4 adverse events, and discontinuations attributable to adverse events occurred with similar frequency across the 3 treatment arms. In spite of the advanced level of disease in the study population, only 13 new category C (AIDS-defining) events were reported. Concerns have been raised previously about the risk of developing lymphoma for patients receiving CCR5 antagonists [11]. It is therefore reassuring that, in this 24-week study and in the 48-week follow up, no cases of lymphoma were reported among these severely immunocompromised patients who were at high risk of lymphoma. Analyses from the MOTIVATE studies of maraviroc in treatment-experienced patients, which were based on a similar design to A but enrolled only patients infected with R5 virus at screening, demonstrated that the addition of maraviroc to optimized background therapy provided significant virologic and immunologic benefit and a similar safety profile to optimized background therapy alone in the patient population [7, 8]. Notably, when evaluating the tropism results at screening and baseline in these studies, 8% of the 1042 patients infected with R5 virus at screening had a different tropism result at baseline [8]. The original Trofile assay used in these studies had a reported 100% sensitivity for detecting CXCR4-using variants present at the 10% level and 85% sensitivity for variants present at the 5% level [12]. The ability to detect CXCR4-using viral subpopulations will therefore depend on the proportion of these minority strains in the total virus population, which can vary over time and may at times be below the limit of detection of the assay [13]. These observations reinforce the importance of the study reported here, which indicates that treatment of patients infected dual- or mixed-tropic HIV-1 with a maraviroccontaining regimen does not result in a negative virologic or immunologic outcome. This study has important limitations. Postdiscontinuation data were not consistently available, and it was not possible to control for the treatment regimen or the quality of patient care after the study. Therefore, any adverse events that became apparent after discontinuation may not have been fully evaluated. Second, because more than one-half of the patients in the maraviroc arms discontinued treatment prior to week 24, the mean duration of maraviroc exposure was 4 and 6 months for the once- and twice-daily arms, respectively. Finally, this study evaluated the efficacy of maraviroc in patients with dual- or mixedtropic virus. The low incidence of obligate CXCR4 use by HIV-1, even in advanced disease, precludes the evaluation of efficacy in patients with X4 HIV-1. The few patients who had an X4 tropism result at baseline in this study had viral load and CD4 cell count 1646 JID 2009:199 (1 June) Saag et al.

10 responses that were within the ranges observed for the primary efficacy population. In conclusion, this study demonstrates that maraviroc combined with an optimized background regimen in treatmentexperienced patients with dual- or mixed-tropic HIV-1 confers little to no virologic benefit but has no apparent adverse effect on clinical outcome. Administration of maraviroc to patients infected with dual- or mixed-tropic virus did not result in a decrease in CD4 cell count; rather, overall CD4 cell counts increased in both the maraviroc arms. Furthermore, this study indicates that maraviroc combined with optimized background therapy is generally well tolerated in patients with non-r5 virus, with no evidence of safety concerns when compared with placebo. Results at week 48 were consistent with findings at week 24. Acknowledgments We thank the study participants, study investigators, and study site staff. References 1. Berger EA, Doms RW, Fenyo EM, et al. A new classification for HIV-1. Nature 1998; 391: Whitcomb JM, Huang W, Fransen S, et al. Analysis of baseline enfuvirtide (T20) susceptibility and co-receptor tropism in two phase III study populations [abstract 557]. In: Program and abstracts of the 10th Conference on Retroviruses and Opportunistic Infections (Boston) Coakley E, Benhamida J, Chappey C, et al. An evaluation of tropism profiles and other characteristics among 3988 individuals screened for the A , A (MOTIVATE 1) and A (MOTIVATE 2) studies for maraviroc [abstract 8]. In: Program and abstracts of the 2nd International Workshop on Targeting HIV Entry (Boston) Wilkin TJ, Su Z, Kuritzkes DR, et al. HIV type 1 chemokine coreceptor use among antiretroviral-experienced patients screened for a clinical trial of a CCR5 inhibitor: AIDS Clinical Trial Group A5211. Clin Infect Dis 2007; 44: Dorr P, Westby M, Dobbs S, et al. Maraviroc (UK-427,857): a potent, orally bioavailable and selective small-molecule inhibitor of the chemokine receptor CCR5 with broad-spectrum anti-human immunodeficiency virus type 1 activity. Antimicrob Agents Chemother 2005; 49: Fätkenheuer G, Pozniak AL, Johnson MA, et al. Efficacy of short-term monotherapy with maraviroc, a new CCR5 antagonist, in patients infected with HIV-1. Nat Med 2005; 11: Gulick R, Lalezari J, Goodrich J, et al.; MOTIVATE Study Teams. Maraviroc for previously treated patients with R5 HIV-1 infection. N Engl J Med 2008; 359: Fatkenheuer G, Nelson M, Lazzarin A, et al.; for the MOTIVATE 1 and MOTIVATE 2 Study Teams. Subgroup analyses of maraviroc in previously treated R5 HIV-1 infection. N Engl J Med 2008; 359: Silvestri G, Paiardini M, Pandrea I, Lederman MM, Sodora DL. Understanding the benign nature of SIV infection in natural hosts. J Clin Invest 2007; 117: Hammer SM, Eron JJ Jr., Reiss P, et al. Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society-USA panel. JAMA 2008; 300: Gulick RM, Su Z, Flexner C, et al. Phase 2 study of the safety and efficacy of vicriviroc, a CCR5 inhibitor, in HIV-1 infected, treatment-experienced patients: AIDS clinical trials group J Infect Dis 2007; 196: Whitcomb JM, Huang W, Fransen S, et al. Development and characterization of a novel single-cycle recombinant-virus assay to determine human immunodeficiency virus type 1 coreceptor tropism. Antimicrob Agents Chemother 2007; 51: Westby M, Lewis M, Whitcomb JM, et al. Emergence of CXCR4-using human immunodeficiency virus type 1 (HIV-1) variants in a minority of HIV-1 infected patients following treatment with the CCR5 antagonist maraviroc is from a pretreatment CXCR4-using virus reservoir. J Virol 2006; 80: Maraviroc in Patients with Non-R5 HIV Infection JID 2009:199 (1 June) 1647