| |
TMC278, Rilpivirine: 2 Phase 3 Studies ECHO/THRIVE Published & Commentary pdfs attached
|
| |
| |
Download the PDF here
The Lancet 16 July 2011
"Overall, QT-intervals corrected according to Fridericia's formula (QTcF) increased from baseline to week 48 in both groups, with no notable differences between groups; mean increases were 12·0 ms (95% CI 10·1-13·8) for rilpivirine and 14·1 ms (12·3-16·0) for efavirenz. There were few adverse events potentially related to conduction abnormalities or to rate and rhythm disturbances (two patients in the rilpivirine group and six in the efavirenz group). One patient in the rilpivirine group discontinued because of a grade 3 QT prolongation (QTcF increased >60 ms [77 ms] at week 48), which was reported by the investigator as an asymptomatic adverse event. No concomitant medications were regarded as having caused the increase in QTcF."
Rilpivirine: a step forward in tailored HIV treatment
Rik Schrijvers a, Belete Ayele Desimmie a, Zeger Debyser a
Molecular Medicine, KU Leuven, 3000 Leuven, Belgium
The results of two highly anticipated clinical trials, ECHO1 and THRIVE,2 are reported in The Lancet. These studies assessed the efficacy and safety of the latest antiretroviral drug against HIV-1 infection, rilpivirine.
Despite the encouraging case of cure in a patient with HIV-1 infection after treatment for acute myeloid leukaemia with a stem-cell transplantation from a CCR5 delta32 homozygous donor,3 no feasible HIV-eradication strategy exists. Standard care is aimed at suppression of plasma viral loads below the limit of detection. This strategy has unequivocally changed the outcome for individuals infected with HIV. However, lifelong treatment is needed because the virus rebounds from latently infected cells once antiretroviral treatment is withdrawn. Furthermore, life expectancy of individuals infected with HIV and treated with combination antiretroviral therapy is lower than that of the general population.4 This excess mortality has been partly attributed to toxic effects of antiretrovirals. Therefore treatment not only needs to be directed at improved efficacy but also at improved tolerability for the patient. Here, the short-term and long-term side-effects in combination with the convenience of a regimen (ie, dosing frequency and the number of pills or the availability of a coformulation) guide therapy choices from the expanding range of antiretrovirals.
In this context, the ECHO and THRIVE investigators report two large international randomised, double-blind, double-dummy, phase 3 clinical trials comparing efavirenz with rilpivirine in combination with two nucleoside or nucleotide reverse transcriptase inhibitors (N[t]RTIs) in treatment-naive adults with HIV-1 infection.
Rilpivirine (TMC278) is the newest member of the class of non-nucleoside reverse transcriptase inhibitors (NNRTIs) that inhibit the HIV-1 reverse transcriptase in an allosteric manner.5 Nevirapine was approved as the first NNRTI by the US Food and Drug Administration (FDA) in 1996, followed by delavirdine in 1997 and efavirenz in 1998. The first next-generation NNRTI, etravirine (TMC125), was approved in 2007 for the treatment of patients with evidenced virological failure and resistance to NNRTI in a two-pill twice-daily regimen.6, 7 The newest NNRTI, rilpivirine, is the result of rational drug design.
The efficacy of efavirenz is established,8 and it is recommended as the first-choice treatment option for treatment-naive adults with HIV-1 infection in combination with two N(t)RTIs, emtricitabine and tenofovir.9 Efavirenz is generally well tolerated and the most common side-effects are rash and CNS toxic effects, necessitating a regimen switch in some patients. Efavirenz is teratogenic and has a low genetic barrier for drug resistance, usually leading to substantial cross-resistance with nevirapine and delavirdine.10
THRIVE and ECHO were designed to assess non-inferiority of rilpivirine to efavirenz. The rationale was sound because of the promising efficacy and safety results of a phase 2b clinical trial.11 These two parallel and independently run trials were started as a first and second confirmatory trial.12 Both trials were identical in design; in THRIVE the background N(t)RTIs were investigator-selected (emtricitabine/tenofovir, abacavir/lamivudine, or zidovudine/lamivudine) and in ECHO were fixed (emtricitabine/tenofovir). FDA approval-which was the reason to report two independent trials12-was granted before publication.13 From both studies, 1368 patients with a median baseline viral load of 5 log10 copies per mL and a median CD4 cell count of 256 cells per μL received at least one dose of rilpivirine or efavirenz.14 Analysis (time to loss of virological response) of the proportion of patients with a plasma viral load lower than the limit of detection (<50 copies per mL) at 48 weeks showed that rilpivirine was not inferior to efavirenz. The proportion of patients with a confirmed response at week 48 was 83% in ECHO and 86% in THRIVE for rilpivirine compared with 83% and 82%, respectively, for efavirenz. Analyses of patients who received at least one dose of study drug and the more conservative per-protocol analysis showed that the non-inferiority criteria were met. Response rates for patients stratified with a baseline viral load of 100 000 copies or fewer per mL were 90% for rilpivirine versus 84% for efavirenz, and 77% for rilpivirine versus 81% for efavirenz in patients with more than 100 000 copies per mL.
Rilpivirine had a better tolerability with fewer grade 2-4 adverse events (16% in the rilpivirine vs 31% in the efavirenz group)14 and fewer adverse events leading to discontinuation of the regimen (3% in the rilpivirine group vs 8% in the efavirenz group). Adverse events commonly attributed to efavirenz, such as rash or CNS side-effects, were less often reported in patients in the rilpivirine group than in the efavirenz group.14 Lipid perturbations were significantly lower in the rilpivirine group than in the efavirenz group although no significant difference in the ratio of total cholesterol to HDL was reported. Both rilpivirine and efavirenz can be coformulated with emtricitabine or tenofovir in a one-pill once-daily regimen, and an open-label comparative study of the triple combination is ongoing (NCT01309243).
Of note, nearly 11% of patients in the rilpivirine group had virological failure compared with nearly 6% in the efavirenz group.14 The authors suggest that this difference might be because of an increased effect on virological failure of suboptimum adherence to rilpivirine than occurred with efavirenz. In an analysis of the pooled data, the proportion of patients with virological failure with baseline viral load of 100 000 copies or fewer per mL was 5·2% (19 of 368 patients) for rilpivirine and 4·8% (16 of 330 patients) for efavirenz.15, 16 The proportion of patients with virological failure and baseline viral load of more than 100 000 copies per mL was 17% (53 of 318) for rilpivirine and 7% (23 of 352) for efavirenz.15, 16 These data suggest that a baseline viral load of more than 5 log10 copies per mL was a predictor of virological failure in the rilpivirine group. This consideration has been adopted in the FDA recommendations.13
What do the trials teach us about cross-resistance? In those patients who failed therapy, 63% in the rilpivirine group developed at least one NNRTI-resistance-associated mutation (mainly E138K, but also K101E, H221Y, V189I, Y181C, or V90I) compared with 54% in the efavirenz group (mainly K103N, but also V106M, Y188C, or K101E). Of the patients with virological failure on rilpivirine who were phenotypically resistant to rilpivirine, 45%, 87%, and 90% were cross-resistant to nevirapine, efavirenz, and etravirine, respectively,1, 15 practically excluding further use of NNRTIs in these patients. This outcome is different from efavirenz resistance, which generally leaves etravirine6, 7 and probably rilpivirine15 as alternative treatment options. Additional studies are needed to clarify these issues. Also unresolved is the question of why proportionally more NRTI-resistance-associated mutations were reported in patients who had virological failure on rilpivirine (68%, mainly M184I) than efavirenz (32%, mainly M184V).15 Because of the importance of the remaining questions about virological failure and cross-resistance, we believe that the absence of a fully published pooled analysis is a missed opportunity.
However, rilpivirine should be embraced as an example of the continuous effort to generate patient-tailored drugs that are highly convenient, have minimal side-effects, and are sufficiently efficacious. The presented data show that rilpivirine is a valid and safe alternative to efavirenz in the treatment of antiretroviral-naive patients infected with HIV-1 and is associated with fewer side-effects. However, although apparently non-inferior, rilpivirine should be started cautiously as patients with a baseline plasma viral load of more than 100 000 copies per mL were more prone to virological failure (and development of resistance). Additional studies explaining the increased rate of virological failure with rilpivirine and studies about resistance or cross-resistance are warranted. The results of ongoing trials assessing the potential and safety of switching from a protease-inhibitor-based (NCT01252940) or efavirenz-based (NCT01286740) regimen to rilpivirine in virologically suppressed patients are eagerly anticipated to expand the rational use of rilpivirine in daily practice.
RS and BAD declare that they have no conflicts of interest. ZD coordinated an FP7 project with Tibotec (ie, money provided by EU not by Tibotec [THINC 2008-2010]).
------------------------------------
The Lancet 16 July 2011
Rilpivirine versus efavirenz with two background nucleoside or nucleotide reverse transcriptase inhibitors in treatment-naive adults infected with HIV-1 (THRIVE): a phase 3, randomised, non-inferiority trial
Dr Calvin J Cohen MD a , Prof Jaime Andrade-Villanueva MD b, Bonaventura Clotet MD c, Jan Fourie MD d, Prof Margaret A Johnson MD e, Prof Kiat Ruxrungtham MD f, Hao Wu MD g, Carmen Zorrilla MD h, Herta Crauwels PhD i, Laurence T Rimsky PhD i, Simon Vanveggel MSc i, Katia Boven MD j, on behalf of the THRIVE study group
Summary
Background
The non-nucleoside reverse transcriptase inhibitor (NNRTI), rilpivirine (TMC278; Tibotec Pharmaceuticals, County Cork, Ireland), had equivalent sustained efficacy to efavirenz in a phase 2b trial in treatment-naive patients infected with HIV-1, but fewer adverse events. We aimed to assess non-inferiority of rilpivirine to efavirenz in a phase 3 trial with common background nucleoside or nucleotide reverse transcriptase inhibitors (N[t]RTIs).
Methods
We undertook a 96-week, phase 3, randomised, double-blind, double-dummy, non-inferiority trial in 98 hospitals or medical centres in 21 countries. We enrolled adults (≥18 years) not previously given antiretroviral therapy and with a screening plasma viral load of 5000 copies per mL or more and viral sensitivity to background N(t)RTIs. We randomly allocated patients (1:1) using a computer-generated interactive web-response system to receive oral rilpivirine 25 mg once daily or efavirenz 600 mg once daily; all patients received an investigator-selected regimen of background N(t)RTIs (tenofovir-disoproxil-fumarate plus emtricitabine, zidovudine plus lamivudine, or abacavir plus lamivudine). The primary outcome was non-inferiority (12% margin on logistic regression analysis) at 48 weeks in terms of confirmed response (viral load <50 copies per mL, defined by the intent-to-treat time to loss of virologic response [TLOVR] algorithm) in all patients who received at least one dose of study drug.
Findings
From May 22, 2008, we screened 947 patients and enrolled 340 to each group. 86% of patients (291 of 340) who received at least one dose of rilpivirine responded, compared with 82% of patients (276 of 338) who received at least one dose of efavirenz (difference 3·5% [95% CI -1·7 to 8·8]; pnon-inferiority<0·0001). Increases in CD4 cell counts were much the same between groups. 7% of patients (24 of 340) receiving rilpivirine had a virological failure compared with 5% of patients (18 of 338) receiving efavirenz. 4% of patients (15) in the rilpivirine group and 7% (25) in the efavirenz group discontinued treatment due to adverse events. Grade 2-4 treatment-related adverse events were less common with rilpivirine (16% [54 patients]) than they were with efavirenz (31% [104]; p<0·0001), as were rash and dizziness (p<0·0001 for both) and increases in lipid levels were significantly lower with rilpivirine than they were with efavirenz (p<0·0001).
Interpretation
Despite a slightly increased incidence of virological failures, a favourable safety profile and non-inferior efficacy compared with efavirenz means that rilpivirine could be a new treatment option for treatment-naive patients infected with HIV-1.
Funding
Tibotec.
Introduction
Recent changes in treatment guidelines1, 2 for HIV-1 recommend early initiation of highly active antiretroviral therapy. For first-line treatment in particular, short-term and long-term tolerability are very important for initiation and staying on treatment. Efavirenz-based regimens are recommended for individuals infected with HIV-1 who are treatment naive.1, 2 However, efavirenz is associated with neurological and psychiatric adverse events, rash, teratogenicity, and increases in concentrations of LDL cholesterol and triglycerides.3, 4
Rilpivirine (TMC278 [Edurant]; Tibotec Pharmaceuticals, County Cork, Ireland) is a US FDA-approved non-nucleoside reverse transcriptase inhibitor (NNRTI),5 which can be given once per day.6 In a phase 2b dose-ranging trial7 of treatment-naive patients with HIV-1, once-daily rilpivirine showed much the same efficacy to once-daily efavirenz for 96 weeks, both given with a background regimen of two nucleoside or nucleotide reverse transcriptase inhibitors (N[t]RTIs). Equivalent efficacy of rilpivirine and efavirenz was sustained for 192 weeks.8 Rilpivirine had a better neurological, rash, and lipid profile than did efavirenz,7, 8 and did not show teratogenic potential in preclinical studies.9
TMC278 against HIV, in a once-daily regimen versus efavirenz (THRIVE) was a 96-week trial that aimed to assess the efficacy, safety, and tolerability of rilpivirine versus efavirenz, with two investigator-chosen background N(t)RTIs in treatment-naive patients with HIV-1 infection. We aimed to show non-inferiority of rilpivirine compared with efavirenz in terms of the percentage of patients with confirmed response (viral load <50 copies per mL defined by the time-to-loss of virological response [TLOVR] algorithm). Because choice of N(t)RTI is often made in clinical practice on the basis of characteristics of patients and local availability, THRIVE was designed to assess rilpivirine with one of three combination N(t)RTI regimens. In this analysis, we report data from the primary analysis at 48 weeks. The results of a companion phase 3 trial (ECHO),10 which compared rilpivirine with efavirenz with a background regimen of tenofovir-disoproxil-fumarate and emtricitabine, are reported separately.
Methods
Trial design and patients
We undertook our phase 3, randomised, double-blind, double-dummy, active-controlled trial in 98 academic medical centres, independent non-profit centres, or hospitals in 21 countries (USA and Puerto Rico, Canada, Australia, Europe [seven countries], South Africa, Asia [four countries], and Latin America [6 countries]). The trial had a 6-week screening period, a 96-week treatment period, and a 4-week follow-up period.
Eligible patients were adults (≥18 years) who were naive to antiretroviral therapy, with a screening plasma viral load of 5000 copies per mL or more and viral sensitivity to the background N(t)RTIs, as assessed with the vircoTYPE HIV-1 assay. Main exclusion criteria were HIV-2 infection, documented presence of at least one of 39 NNRTI resistance-associated mutations (RAMs)11 active clinically significant disease (eg, pancreatitis, cardiac dysfunction, active and significant psychiatric disorder, adrenal insufficiency, or hepatic impairment), renal impairment, and for women, pregnancy or breastfeeding.
The protocol was reviewed and approved by independent ethics committees and institutional review boards at participating sites or at a central institutional review board for some sites (eg, in the USA), and the trial was undertaken in accordance with the principles of good clinical practice and the Declaration of Helsinki. All patients provided written consent.
Randomisation and masking
We randomly allocated patients with a computer-generated interactive web-response system in a one-to-one ratio to receive oral rilpivirine 25 mg once daily or efavirenz 600 mg once daily after investigator selection of the background N(t)RTI regimen, which included tenofovir-disoproxil-fumarate plus emtricitabine, zidovudine plus lamivudine, or abacavir plus lamivudine. Randomisation was stratified by background regimen and screening viral load (≤100 000 copies per mL, 100 001-500 000 copies per mL, and >500 000 copies per mL). Investigators, the sponsor, and patients were masked to NNRTI treatment assignment.
Procedures
We used a double-dummy regimen in which rilpivirine (or matching placebo) was taken with a meal, whereas efavirenz (or matching placebo) was taken on an empty stomach in the evening. N(t)RTIs were taken according to the locally applicable procedures and package inserts, but preferably at the same time as rilpivirine or efavirenz for abacavir plus lamivudine and tenofovir-disoproxil-fumarate plus emtricitabine. For zidovudine plus lamivudine (taken twice daily), the first dose was preferably taken in the morning with rilpivirine (or placebo), and the second dose was preferably taken in the evening with efavirenz (or placebo).
Disallowed drugs were all investigational drugs, drugs that could reduce rilpivirine exposure (eg, those with a potent cytochrome 3A4-inducing effect or proton-pump inhibitors), drugs disallowed for efavirenz or the background regimen (as per the package inserts) and any anti-HIV therapy other than those used in the trial. Antacids (≥2 h before or ≥4 h after) and histamine H2-receptor antagonists (≥12 h before or ≥4 h after) after rilpivirine were allowed. Switches between N(t)RTIs were allowed only if intolerance occurred and were guided by resistance testing.
Our primary outcome was non-inferiority of rilpivirine to efavirenz in terms of percentage of all patients who received at least one dose of rilpivirine or efavirenz who had a confirmed virological response (defined by the intent-to-treat TLOVR algorithm) at 48 weeks. We used a non-inferiority margin of 12% (lower limit of two-sided 95% CI) to establish non-inferiority of rilpivirine from efavirenz, which is in accordance with the margin of 10-12% suggested by the FDA for HIV drug development.12 Secondary outcomes were non-inferiority with a 10% margin and superiority (if non-inferiority was shown), antiviral activity in time, changes from baseline in CD4 cell count, safety, tolerability, HIV genotypic and phenotypic characteristics (in virological failures), adherence (assessed by the Modified Medication Adherence Self-Report Inventory), pharmacokinetics, and pharmacokinetic and pharmacodynamic relations.
We followed-up patients at weeks 2, 4, 8, 12, and 16, and every 8 weeks thereafter. We collected urine and blood samples for haematology and biochemistry, urinalysis, immunology, plasma viral load, and viral genotype and phenotype assessments. We assessed plasma viral load (concentration of HIV-1 RNA) with the Amplicor HIV-1 monitor test version 1.5 (Roche, Basel, Switzerland).
In the primary analysis, patients were regarded as non-responders if they discontinued treatment prematurely for any reason or if they had virological failure. Patients with virological failure were classified as never suppressed (never achieving viral load <50 copies per mL before week 48) or as rebounder (after achievement of two consecutive viral load values of <50 copies per mL but then having viral load ≥50 copies per mL at two consecutive assessments).
Virological failure in the resistance analysis was assessed in all patients who had received at least one dose of study drug and included all failures of treatment in the database, irrespective of time of failure (at, before, or after week 48), treatment status or reason for discontinuation, provided the following criteria were met: never achieved two consecutive viral load values of fewer than 50 copies per mL and had an increase in viral load 0-5 log10 copies per mL or more above the nadir (never suppressed) or first achieved two consecutive viral load values <50 copies per mL followed by two consecutive (or single, when last available) viral load values of 50 copies per mL or more (rebounder).
Virco (Mechelen, Belgium) did the viral phenotypic assessments with the Antivirogram assay and genotypic assessments with the VircoTYPE HIV-1 assay. We coded adverse events with MedDRA (version 11.0), and established severity of adverse events according to the division of AIDS grading scale.13 An independent international data and safety monitoring board monitored safety and efficacy throughout the trial.
We estimated glomerular filtration rate at baseline, weeks 2, and 24 on the basis of serum creatinine (eGFRcreat) with the modification of diet in renal disease trial formula14 and serum cystatin C concentrations (eGFRcyst) with the Hoek formula.15 We did an electrocardiograph at screening and weeks 2, 12, 24, and 48.
Statistical analysis
We assessed non-inferiority of rilpivirine to efavirenz in patients who received at least one dose of rilpivirine or efavirenz, irrespective of protocol adherence, and in a per-protocol population of all randomly allocated patients who received study drugs, excluding those with major protocol violations.
Assuming a response rate of 75% at week 48 for both treatment groups,16-22 we needed to enrol 340 patients in each group to establish non-inferiority of rilpivirine to efavirenz, with a maximum allowable difference of 12% at 95% power. We assessed the primary efficacy endpoint with a predicted-response analysis by use of a logistic regression analysis adjusted for the stratification factors (baseline log10 plasma viral load and background N[t]RTIs). We also did a sensitivity analysis for the subpopulation, excluding patients who discontinued for reasons other than virological failure according to the resistance analysis criteria.
In the analysis of mean change in absolute CD4 cell count from baseline, for premature discontinuations, we imputed data with baseline values (non-completer was classified as failure). For other missing values, we used the last observation carried forward method.
We undertook preplanned statistical analyses with Fisher's exact test (5% significance level) for prespecified adverse events that reported a significant difference in the phase 2b trial.7 For adverse events, we made no adjustment for multiple comparisons between groups. We used a non-parametric Wilcoxon rank-sum test to compare changes in lipid concentrations between treatment groups.
Role of the funding source
The study sponsor was involved in the design and conduct of the trial, and in data collection and analysis. All authors had full access to the 48-week clinical trial report. The corresponding author had final responsibility for the decision to submit for publication.
Results
Figure 1 shows the trial profile and table 1 shows the baseline characteristics of patients. The study started on May 22, 2008. The cut-off date for the 48-week analysis was Jan 28, 2010, and for the 96-week analysis was Jan 7, 2011. 326 patients (48%) were from the USA, Canada, Europe, and Australia. As reported by the investigators, discontinuations were mainly due to adverse events (15 of 340 [4%] in the rilpivirine group compared with 25 of 338 [7%] in the efavirenz group), virological failure (investigator-reported virological failure; 13 [4%] vs 8 [2%]), loss to follow-up (10 [3%] vs 6 [2%]), and withdrawn consent (2 [1%] vs 11 [3%]). Six patients (2%) in each group had major protocol violations, which included use of a disallowed drug (5 [1%] vs 4 [1%]), use of disallowed background therapy (1 [<1%] in both groups), or selection criteria not met (0 vs 1 [<1%]).
One of 204 (<1%) patients in the rilpivirine group switched from the initial background regimen of tenofovir-disoproxil-fumarate plus emtricitabine, five of 101 (5%) switched from zidovudine plus lamivudine, and one of 35 (3%) switched from abacavir plus lamivudine; the corresponding values in the efavirenz group were one of 202 (<1%) patients, five of 103 (5%) patients, and one of 33 (3%) patients.
In the primary analysis, 86% of patients (291 of 340) assigned to receive rilpivirine had a confirmed viral load of fewer than 50 copies per mL at 48 weeks, compared with 82% (276 of 338) for efavirenz (table 2). The lower 95% CI of estimated difference in confirmed response at 48 weeks in the logistic regression model was greater than -12% and -10%, confirming non-inferiority at the 12% (primary endpoint) and 10% margins (p<0·0001). However, we did not note superiority. 7% (24 of 340) of patients in the rilpivirine group had virological failure compared with 5% (18 of 338) in the efavirenz group. Results from the predicted-response analysis that adjusted for stratification factors were equivalent to those for the main analysis (table 2). Rilpivirine remained non-inferior to efavirenz in the per-protocol analysis (table 2). Figure 2 shows percentages of responders in both groups from baseline to 48 weeks (primary analysis).
In a sensitivity analysis excluding patients who discontinued for reasons other than virological failure (defined as in the resistance analysis), 91% (291 of 319) patients in the rilpivirine group responded, compared with 93% (276 of 296) in the efavirenz group (difference -2·0%, 95% CI -6·3 to 2·2).
Mean CD4 cell counts continuously increased from baseline to 48 weeks for rilpivirine and efavirenz (figure 2). At week 48, the mean change from baseline in CD4 cell count was 189 cells per μL (95% CI 174-203) with rilpivirine and 171 cells per μL (155-187) with efavirenz (p=0·09).
The Modified Medication Adherence Self-Report Inventory data were not available for all patients. 89% (243 of 272) of patients who self-reported better than 95% adherence responded to treatment in the rilpivirine group as did 90% (206 of 230) in the efavirenz group. 64% (23 of 36) of patients who were less than or equal to 95% adherent responded in the rilpivirine group (median adherence 92·2%), compared with 62% (24 of 39) in the efavirenz group (median adherence 91·5%). In the rilpivirine group, 91% (170 of 187) of patients with a baseline viral load of 100 000 copies per mL or fewer responded, compared with 80% (94 of 118) for 100 001-500 000 copies per mL, and 77% (27 of 35) for more than 500 000 copies per mL; the corresponding numbers for the efavirenz group were 84% (140 of 167), 82% (112 of 136), and 69% (24 of 35). The background N(t)RTI regimen had no significant effect on response. However, given the small numbers of patients with low (≤95%) adherence as assessed by the Modified Medication Adherence Self-Report Inventory and high baseline viral loads, findings from such patients should be interpreted with caution.
8% of patients (27 of 340) in the rilpivirine group had virological failure according to the resistance analysis (including those without emerging mutation at failure) compared with 6% (20 of 338) in the efavirenz group (table 3).
We did safety analyses with all available data, including those for patients treated beyond 48 weeks. Adverse events were generally mild-to-moderate (grade 1 or 2). Prevelance of any grade 2-4 adverse events at least possibly related to treatment was lower in the rilpivirine group than it was in the efavirenz group (table 4). Rash was the main adverse event leading to discontinuation in the efavirenz group (five patients), but no discontinuations related to rash occurred in the rilpivirine group. All other adverse events that caused discontinuation from various system organ classes occurred in less than 1% of patients in either group.
Neurological events of interest (cluster headache, cranial neuropathy, disturbance in attention, dizziness, facial palsy, headache, lethargy, memory impairment, mononeuropathy, paraesthesia circumoral, photophobia, restlessness, sensation of pressure in ear, somnolence, uveitis, vertigo, or blurred vision) possibly related to treatment (any grade) occurred in 18% of patients (62 of 340) in the rilpivirine group compared with 39% (132 of 338) in the efavirenz group (p<0·0001). Individual neurological adverse events occurring in 2% or more of patients included dizziness (10% of patients [33 of 340] in the rilpivirine group vs 28% [94 of 338] in the efavirenz group; p<0·0001), headache (6% [20] vs 8% [27]), somnolence (4% [13] vs 8% [28]), and disturbance in attention (1% [three] vs 2% [seven]). Psychiatric events of interest (abnormal dreams, affective disorder, aggression, agitation, anxiety, confusional state, depressed mood, depression, euphoric mood, homicidal ideation, insomnia, irritability, libido decreased, major depression, mood swings, nervousness, nightmare, panic attack, phobia, post-traumatic stress disorder, sleep disorder, social phobia, sopor, stress symptoms, or suicide attempt) at least possibly related to treatment (any grade) occurred in 15% of patients (52 of 340) in the rilpivirine group compared with 20% (69 of 338) in the efavirenz group (p=0·09). Psychiatric adverse events occurring in 2% of patients or more were abnormal dreams or nightmares (7% [24 of 340] patients in the rilpivirine group vs 11% [38 of 338] in the efavirenz group; p=0·06), insomnia (6% [20] vs 5% [16]), and sleep disorder (2% [seven] vs 3% [nine]). Most neurological (98%) and psychiatric (96%) adverse events of interest were grade 1 or 2.
3% of patients (nine of 340) in the rilpivirine group had rash at least possibly related to treatment (any grade), compared with 13% (43 of 338) in the efavirenz group (p<0·0001). Of all treatment-related rashes (grouped term), 100% were grade 1-2 in the riplivirine group, with 99% grade 1-2 and 1% grade 3 in the efavirenz group. Rash resolved with continued dosing in both treatment groups, apart from for five patients who discontinued because of rash in the efavirenz group.
Table 4 shows rates of serious adverse events and treatment-emergent grade 3-4 laboratory abnormalities. There was one death in the rilpivirine group (bronchopneumonia) and three in the efavirenz group (one cerebral toxoplasmosis and dysentery, one cerebrovascular accident, and one respiratory failure). All four deaths were unrelated to treatment. With the exception of increased LDL-cholesterol (6% with efavirenz), individual grade 3-4 laboratory abnormalities occurred in 3% or less of patients.
Mean increases in total cholesterol, LDL-cholesterol, and triglyceride concentrations from baseline to week 48 were significantly lower with rilpivirine than they were with efavirenz (p<0·0001; table 5). The mean increase in HDL-cholesterol was significantly lower with rilpivirine than it was with efavirenz (p<0·0001). The ratio of total cholesterol to HDL-C did not differ between groups.
We noted a small increase from baseline in mean serum creatinine at the first on-treatment assessment, which remained stable over 48 weeks with rilpivirine (range 4·11-7·16 μmol/L), but no change with efavirenz. Treatment-associated changes in glomerular filtration rate differed according to the estimation used. There was a maximum mean decrease in eGFRcreat of 5-9 mL/min per 1·73 m2 from baseline during treatment with rilpivirine, corresponding to the change in creatinine concentration, with glomerular filtration rate remaining in the healthy range for all patients. eGFRcyst increased in both groups at week 2 (3 mL/min for rilpivirine vs 5 mL/min for efavirenz) and at week 24 (22 mL/min vs 31 mL/min). There were no grade 3-4 creatinine abnormalities, abnormalities reported as adverse events, or renal-related trial discontinuations.
Overall, QT-intervals corrected according to Fridericia's formula (QTcF) increased from baseline to week 48 in both groups, with no notable differences between groups; mean increases were 12·0 ms (95% CI 10·1-13·8) for rilpivirine and 14·1 ms (12·3-16·0) for efavirenz. There were few adverse events potentially related to conduction abnormalities or to rate and rhythm disturbances (two patients in the rilpivirine group and six in the efavirenz group). One patient in the rilpivirine group discontinued because of a grade 3 QT prolongation (QTcF increased >60 ms [77 ms] at week 48), which was reported by the investigator as an asymptomatic adverse event. No concomitant medications were regarded as having caused the increase in QTcF.
Discussion
We showed that oral rilpivirine once daily is non-inferior in terms of efficacy to efavirenz at 48 weeks when given in combination with background N(t)RTIs. Both rilpivirine and efavirenz had high response rates. In our study, response rates to efavirenz were among the highest reported when compared with earlier studies in treatment-naive patients (panel).25-29 For both groups, the proportion of patients with virological failure was low. Within these virological failures, the rate of rebound after suppression was also low and much the same between groups with less than 5% of patients never suppressed in either group. The proportion of patients who discontinued due to adverse events and other reasons was lower for rilpivirine than it was for efavirenz, resulting in similar response rates.
Panel
Research in context
Systematic review
Before approval of rilpivirine, two main non-nucleoside reverse transcriptase inhibitors (NNRTIs) were available for the first-line treatment of patients with HIV-1 infections in combination with reverse transcriptase inhibitors: efavirenz and nevirapine. Both drugs are equally effective in suppression of HIV infection24 but cause different side-effects that can restrict their use. Treatment with efavirenz can lead to rash, impaired mental function, vertigo, abnormal dreams, and fetal malformations, whereas nevirapine treatment is associated with severe rash and liver damage due to hypersensitivity reactions. These adverse effects emphasise the unmet need for additional first-line NNRTI treatment options with a good safety profile.
We searched the PubMed database to May, 2011, for randomised controlled trials and clinical trials published in English with the search term "rilpivirine". Rilpivirine has only been assessed in one phase 2b study.7, 8 This randomised, controlled, dose-finding study of 368 patients showed that rilpivirine provided long-term (96 weeks) efficacy and tolerability in treatment-naive adults with HIV-1 infections, with comparable response rates with efavirenz. In this study, all daily rilpivirine doses (25 mg, 75 mg, and 150 mg) resulted in much the same response rates. Grade 2-4 adverse events at least possibly related to study medication, including nausea, dizziness, abnormal dreams or nightmares, rash, somnolence, and vertigo were less frequent with TMC278 than they were with efavirenz in the context of an open-label trial (only the dose of rilpivirine was masked).7
Interpretation
The phase 3 ECHO10 and THRIVE trials were independently undertaken and powered to investigate the non-inferior efficacy of rilpivirine 25 mg once daily in combination with different reverse transcriptase inhibitors backbones for first-line therapy in adults with HIV-1 infection, compared with efavirenz, the preferred NNRTI for treatment-naive patients. Both studies met the primary objective of non-inferiority and also showed rilpivirine to have a more favourable side-effect profile versus efavirenz, with a reduced incidence of rash and central nervous system adverse reactions. However, the virological failure rate was slightly higher with rilpivirine than it was with efavirenz, and exploratory analyses are ongoing to examine the reasons for this difference in more detail.
Rilpivirine was better tolerated than was efavirenz in terms of lower incidences of discontinuations due to adverse events, especially due to central nervous system side-effects such as insomnia, depression, dizziness, or due to rash, when compared with patients taking efavirenz. Consequently, rilpivirine might be suitable for use in some treatment-naive patients, which efavirenz is not, such as women of child bearing age or potential and patients with certain pre-existing psychiatric conditions.
The ECHO trial10 examined rilpivirine with emtricitabine and tenofovir-disoproxil-fumarate, a combination that has been submitted for marketing approval as a fixed-dose combination for antiretroviral-naive patients with HIV-1 infection. If this fixed-dose single-tablet is approved, it would be an alternative to the currently licensed once-daily single tablet, ATRIPLA combining emtricitabine and tenofovir-disoproxil-fumarate to efavirenz.
THRIVE assessed the safety and efficacy of rilpivirine with three different background regimens (tenofovir-disoproxil-fumarate plus emtricitabine, zidovudine plus lamivudine, or abacavir plus lamivudine) versus efavirenz. Based on these data and the findings in ECHO, rilpivirine 25 mg tablets have been approved in the USA in combination with other antiretroviral drugs for the first-line treatment of HIV infection.
Taken together, these data suggest that once-daily rilpivirine is likely to be a valuable treatment option for antiretroviral-naive patients with HIV-1 infection.
Although our study was not powered to assess within-group significance, response rates seemed highest in the rilpivirine group for patients with lowest baseline viral loads, and background N(t)RTI regimen seemed to have no significant effect on responses. The slightly higher virological failure rate noted with rilpivirine than with efavirenz might be explained by a greater effect of suboptimum adherence on virological failure with rilpivirine than with efavirenz. Because of statistical power limitations in the separate ECHO and THRIVE studies, however, results of additional exploratory analyses of effects of factors on response and virological failure will be reported separately for pooled data analyses. Furthermore, pharmacokinetic and pharmacodynamic relationships will be reported elsewhere for the pooled data.
The proportion of virological failures (according to the resistance analysis criteria) with at least one treatment-emergent NNRTI RAM was much the same in both groups, whereas the proportion with at least one treatment-emergent international AIDS society-USA N(t)RTI RAM23 was higher in the rilpivirine group than it was in the efavirenz group. Consistent with reports from the phase 2b TMC278-C204 trial,7, 8 E138K was the most prevalent NNRTI RAM in the rilpivirine group and K103N was in the efavirenz group, whereas M184I/V were the most prevalent N(t)RTI RAMs in both groups. Phenotypic testing in the pooled analysis showed that 28 of 31 (90%) patients who had virological failure in the rilpivirine group and were phenotypically resistant to rilpivirine were cross-resistant to etravirine.30 A pooled analysis of sensitivity to NNRTIs will be presented separately.
Rilpivirine was well tolerated, with a more favourable overall profile than efavirenz considering grade 2-4 adverse events at least possibly related to treatment, rash, dizziness, and smaller increases in some proatherogenic lipid parameters, but there was no significant difference in the ratio of total cholesterol to HDL-cholesterol between groups. These data support the more favourable safety profile for rilpivirine compared with efavirenz that was reported in the phase 2b TMC278-C204 trial7, 8 and in the phase 3 ECHO trial.10 By use of cystatin C, which is an alternative indication of renal function,31 rilpivirine did not have a clinically relevant effect.
One limitation of our trial was that it was not powered to assess comparisons of efficacy in the various subsets of patients. Moreover, a comprehensive NNRTI RAM list was used to screen out patients potentially resistant to NNRTIs because the trial was double-blinded. Response rates might have been higher in this trial than they would be in the clinic where patients might harbour transmitted resistance. However, E138K has a low prevalence in routine clinical resistance testing (<1%).32 Our trial had a double-blind, double-dummy design, meaning that patients had to take study treatment twice daily, rather than once daily, although the effect of this design feature on response rates is not known. A further limitation of the study was that, in common with most clinical HIV studies,33 some populations of patients were under-represented (eg, women). Nevertheless, subgroup analyses of the combined ECHO and THRIVE populations by sex, region, race, clade, and co-infection with hepatitis B and C,30, 34, 35 show that the efficacy of rilpivirine and efavirenz are equivalent, suggesting broader applicability of our data.
Thus, on the basis of our data and those from the companion phase 3 trial, ECHO,10 rilpivirine is expected to be a valuable treatment option for antiretroviral-naive patients infected with HIV-1.
------------------------------------------
The Lancet 16 July 2011
Rilpivirine versus efavirenz with tenofovir and emtricitabine in treatment-naive adults infected with HIV-1 (ECHO): a phase 3 randomised double-blind active-controlled trial - pdf attached
Download the PDF here
Jean-Michel Molina, Pedro Cahn, Beatriz Grinsztejn, Adriano Lazzarin, Anthony Mills, Michael Saag, Khuanchai Supparatpinyo, Sharon Walmsley,
Herta Crauwels, Laurence T Rimsky, Simon Vanveggel, Katia Boven, on behalf of the ECHO study group
Summary
Background
Efavirenz with tenofovir-disoproxil-fumarate and emtricitabine is a preferred antiretroviral regimen for treatment-naive patients infected with HIV-1. Rilpivirine, a new non-nucleoside reverse transcriptase inhibitor, has shown similar antiviral efficacy to efavirenz in a phase 2b trial with two nucleoside/nucleotide reverse transcriptase inhibitors. We aimed to assess the efficacy, safety, and tolerability of rilpivirine versus efavirenz, each combined with tenofovir-disoproxil-fumarate and emtricitabine.
Methods
We did a phase 3, randomised, double-blind, double-dummy, active-controlled trial, in patients infected with HIV-1 who were treatment-naive. The patients were aged 18 years or older with a plasma viral load at screening of 5000 copies per mL or greater, and viral sensitivity to all study drugs. Our trial was done at 112 sites across 21 countries. Patients were randomly assigned by a computer-generated interactive web response system to receive either once-daily 25 mg rilpivirine or once-daily 600 mg efavirenz, each with tenofovir-disoproxil-fumarate and emtricitabine. Our primary objective was to show non-inferiority (12% margin) of rilpivirine to efavirenz in terms of the percentage of patients with confirmed response (viral load <50 copies per mL intention-to-treat time-to-loss-of-virological-response [ITT-TLOVR] algorithm) at week 48. Our primary analysis was by intention-to-treat. We also used logistic regression to adjust for baseline viral load.
Findings
346 patients were randomly assigned to receive rilpivirine and 344 to receive efavirenz and received at least one dose of study drug, with 287 (83%) and 285 (83%) in the respective groups having a confirmed response at week 48. The point estimate from a logistic regression model for the percentage difference in response was -0·4 (95% CI -5·9 to 5·2), confirming non-inferiority with a 12% margin (primary endpoint). The incidence of virological failures was 13% (rilpivirine) versus 6% (efavirenz; 11% vs 4% by ITT-TLOVR). Grade 2-4 adverse events (55 [16%] on rilpivirine vs 108 [31%] on efavirenz, p<0·0001), discontinuations due to adverse events (eight [2%] on rilpivirine vs 27 [8%] on efavirenz), rash, dizziness, and abnormal dreams or nightmares were more common with efavirenz. Increases in plasma lipids were significantly lower with rilpivirine.
Interpretation
Rilpivirine showed non-inferior efficacy compared with efavirenz, with a higher virological-failure rate, but a more favourable safety and tolerability profile.
Funding
Tibotec.
Introduction
Many antiretroviral regimens with similar antiviral activity are available for treatment-naive individuals infected with HIV-1.1-4 Treatment selection is increasingly based on the tolerability profile and convenience of the regimen. Several treatment guidelines recommend the non-nucleoside reverse transcriptase inhibitor (NNRTI), efavirenz, in combination with tenofovir-disoproxil-fumarate and emtricitabine.1-4 However, efavirenz is associated with several adverse events.5, 6 Furthermore, efavirenz is contraindicated for pregnant women because of concerns over potential teratogenicity.1, 5, 6
Rilpivirine (Tibotec Pharmaceuticals, Co Cork, Ireland) is an NNRTI that has been approved for use in the USA.7 It is a potential alternative to efavirenz for treatment-naive patients infected with HIV-1. This drug does not seem teratogenic in non-primate animals.8 In a large, phase 2b, randomised trial (TMC278-C204)9 in 368 treatment-naive patients, all once-daily doses of rilpivirine (25 mg, 75 mg, and 150 mg) showed antiviral efficacy similar to that recorded with once-daily 600 mg efavirenz, when either drug was given in combination with two nucleoside/nucleotide reverse transcriptase inhibitors (N[t]RTIs). Both rilpivirine and efavirenz had similar, sustained efficacy over 192 weeks.10 Rash and neurological and psychiatric adverse events were reported less commonly with rilpivirine than with efavirenz, and lipid increases were smaller. The once-daily 25 mg dose of rilpivirine was selected for further development, because it had the best benefit-risk balance, with the lowest incidence of discontinuations due to adverse events and rashes.9 Further, once-daily 25 mg rilpivirine had no effect on QTc interval in a thorough QT trial.11, 12 In our trial, Efficacy Comparison in Treatment-naive, HIV-infected Subjects of TMC278 and Efavirenz (ECHO), our aim was to assess the efficacy, safety, and tolerability of rilpivirine versus efavirenz, each combined with a background regimen of tenofovir-disoproxil-fumarate and emtricitabine. We present data from our primary 48-week analysis.
Methods
Participants
We did a 96-week (April 21, 2008, to Jan 4, 2011), phase 3, double-blind, double-dummy, active-controlled randomised trial, to assess the efficacy, safety, and tolerability of once-daily 25 mg rilpivirine versus once-daily 600 mg efavirenz with a background regimen of tenofovir-disoproxil-fumarate and emtricitabine. Our trial was done at 112 sites across 21 countries (USA, Canada, Australia, South Africa, ten countries in Europe, three in Asia, and four in Latin America).
Our main inclusion criteria were patients aged 18 years or older, who had not been previously treated with antiretroviral drugs, a plasma viral load at screening of 5000 copies per mL or greater, and viral sensitivity to tenofovir-disoproxil-fumarate and emtricitabine (assessed with the resistance genotype virco TYPE HIV-1 assay; Virco BVBA, Beerse, Belgium). Further exclusion criteria included infection with HIV-2, documented evidence of at least one NNRTI resistance-associated mutation (RAM) from a list of 39 (A98G, L100I, K101E/P/Q, K103H/N/S/T, V106A/M, V108I, E138A/G/K/Q/R, V179D/E, Y181C/I/V, Y188C/H/L, G190A/C/E/Q/S/T, P225H, F227C, M230I/L, P236L, K238N/T, and Y318F),13 any active clinically significant disease (eg, pancreatitis, cardiac dysfunction, active and significant psychiatric disorder, adrenal insufficiency, hepatic impairment), renal impairment (estimated glomerular filtration rate based on creatinine <50 mL per min), and, for women, pregnancy or breastfeeding.
All patients gave written consent before any trial-related procedure. Our protocol was reviewed and approved by independent ethics committees and institutional review boards, and our trial was done in accordance with the principles of good clinical practice and the Declaration of Helsinki.
Randomisation and masking
Patients infected with HIV-1 were randomly assigned (1:1) by a computer-generated interactive web response system to receive either once-daily 25 mg rilpivirine or once-daily 600 mg efavirenz, both given in combination with a fixed-dose background regimen of once-daily 300 mg tenofovir-disoproxil-fumarate and once-daily 200 mg emtricitabine. The investigator, sponsor, and patient did not know which NNRTI treatment the patient was assigned to receive. Randomisation was stratified by screening viral load (≤100 000 copies per mL, >100 000 to ≤500 000 copies per mL, and >500 000 copies per mL). Our double-dummy design required that rilpivirine (or matching placebo) be taken with a meal, whereas efavirenz (or matching placebo) was required to be taken on an empty stomach in the evening.
Procedures
Our trial consisted of a 6-week screening period, a 96-week treatment period, and a 4-week follow-up period. The results of a companion phase 3 trial, TMC278 against HIV, in a once-daily regimen versus efavirenz (THRIVE),14 are reported separately.
Disallowed drugs included those which could reduce exposure to rilpivirine (ie, potent cytochrome 3A4-inducers and proton-pump inhibitors); drugs disallowed for efavirenz or tenofovir-disoproxil-fumarate and emtricitabine, as per the package inserts; any anti-HIV treatment other than drugs used in our trial; and all investigational drugs. Antacids (given at least 2 h before or at least 4 h after rilpivirine) and histamine H2-receptor antagonists (given at least 12 h before or at least 4 h after rilpivirine) were permitted. Switches of N(t)RTIs were only allowed if there was N(t)RTI intolerance, and in accordance with the drug susceptibility profile; treatment was judged to have not failed in these patients.
Our primary objective was to show non-inferiority of treatment with once-daily 25 mg rilpivirine compared with once-daily 600 mg efavirenz in terms of the percentage of patients with confirmed response (according to the intention-to-treat time-to-loss-of-virological-response [ITT-TLOVR] algorithm) at week 48, with a non-inferiority margin of rilpivirine versus efavirenz of 12% (based on the lower limit of the two-sided 95% CI). Our selected non-inferiority margin was chosen in accordance with US Food and Drug Administration guidelines for HIV drug development that suggest a margin ranging from 10% to 12%.15
Our secondary endpoints were non-inferiority at a 10% margin, superiority (if non-inferiority was shown), durability of antiviral activity, changes from baseline in CD4 cell count, safety, tolerability, HIV genotypic and phenotypic characteristics (in virological failures), adherence (measured with the Modified Medication Adherence Self-Report Inventory [M-MASRI]), pharmacokinetics, and pharmacokinetic and pharmacodynamic relations.
Patients attended scheduled trial visits at weeks 2 and 4, every 4 weeks until week 16, and then every 8 weeks. We collected urine and blood samples for urinalysis, haematology and biochemistry, immunology, plasma viral load, and viral genotype and phenotype determinations. We established plasma viral load (HIV-1 RNA concentration) with the Amplicor HIV-1 Monitor Test version 1·5 (Roche, Basel, Switzerland).
In our ITT-TLOVR analysis, non-responders were patients who discontinued the trial prematurely for any reason, or who had virological failure. We categorised patients with virological failure as either never suppressed (never achieving viral load <50 copies per mL before week 48) or as a rebounder (after having achieved two consecutive viral load values of <50 copies per mL but then having viral load ≥50 copies per mL at two consecutive timepoints).
We established virological failure according to our resistance analysis in our intention-to-treat (ITT) population and included all treatment failures in the database, irrespective of time of failure (whether at, before, or after week 48), treatment status, or reason for discontinuation, provided these criteria were met: never achieved two consecutive viral-load values less than 50 copies per mL and had an increase in viral load of 0·5 log10 copies per mL or greater above the nadir (never suppressed) or first achieved two consecutive viral-load values less than 50 copies per mL with two subsequent consecutive (or single, when last available) viral-load values of 50 copies per mL or greater (rebounder). Viral phenotypic and genotypic assessments were done by Virco BVBA (Mechelen, Belgium), with Antivirogram and Virco TYPE HIV-1 assays, respectively.
An independent data and safety monitoring board monitored the safety of patients during the trial. The Medical Dictionary for Regulatory Activities (version 11·0) was used to code adverse events, and adverse-event severity was assessed with the Division of Acquired Immunodeficiency Syndrome grading scale.16 Included in our assessment of adverse events were those whose association with NNRTIs are well described: neurological events of interest and psychiatric events of interest. Neurological events of interest are defined as cluster headache, cranial neuropathy, disturbance in attention, dizziness, facial palsy, headache, lethargy, memory impairment, mononeuropathy, paraesthesia circumoral, photophobia, restlessness, sensation of pressure in ear, somnolence, uveitis, vertigo, or blurred vision. Psychiatric events of interest are defined as abnormal dreams, affective disorder, aggression, agitation, anxiety, confusional state, depressed mood, depression, euphoric mood, homicidal ideation, insomnia, irritability, libido decreased, major depression, mood swings, nervousness, nightmare, panic attack, phobia, post-traumatic stress disorder, sleep disorder, social phobia, sopor, stress symptoms, or suicide attempt.
We estimated glomerular filtration rate (GFR), based on serum creatinine, with the Modification of Diet in Renal Disease trial formula (eGFRcreat).17 An electrocardiograph was recorded at screening and at weeks 2, 12, 24, and 48.
Statistical analysis
The population for primary analysis was our ITT population (ie, all who had received a study drug). We did an additional analysis on our per-protocol population (ie, as ITT but excluding major protocol violations). Our sample-size calculations took into account response rates in previous trials with efavirenz.18-24 Given an expected response rate of 75% at week 48, we needed 340 patients in each treatment group to establish non-inferiority of rilpivirine to efavirenz with a maximum allowable difference of 12% at 95% power. We also did a logistic-regression analysis for our primary efficacy endpoint adjusted for the stratification factor, baseline log10 plasma viral load (≤100 000, >100 000 to ≤500 000, and >500 000 copies per mL). We did a sensitivity analysis on the subpopulation censored for non-virological failure according to our resistance analysis, excluding patients who discontinued for reasons other than virological failure according to our resistance analysis.
In our analysis of mean change in absolute CD4 cell count from baseline, for premature discontinuations data were assigned the baseline value (non-completer equalled failure). For other missing values, our last observation was carried forward.
We preplanned all presented statistical analyses. We used Fisher's exact test (5% significance level) to compare prespecified adverse events for which a significant difference had been recorded in the phase 2b trial.9 We applied no adjustment for multiple comparisons between groups. We did a non-parametric Wilcoxon rank-sum test to compare lipid changes between our two treatment groups.
Role of the funding source
The study sponsor was involved in the design and conduct of the trial, and in the collection and analysis of the data. All authors had full access to the 48-week clinical trial report. The corresponding author had final responsibility to submit the manuscript for publication.
Results
Figure 1 shows the trial profile. Our ITT analyses included the 690 patients who received at least one dose of study drug. 106 patients (15%) discontinued treatment before week 48, and reasons for discontinuations, as reported by the investigators, were balanced between groups, with the exception of adverse events (eight in the rilpivirine group vs 28 in the efavirenz group) and investigator-reported virological failure (23 vs six, respectively). The incidence of major protocol violations was similar in each group (11 of 346 patients in the rilpivirine group vs 14 of 344 in the efavirenz group) and included use of a disallowed drug in the treatment period (six vs nine), deviation of background treatment (three vs five), selection criteria not met (two vs none), and non-compliance with study drug intake (one vs two). One patient assigned to receive efavirenz switched background regimen to abacavir plus lamivudine because of renal impairment.
At week 48, 83% of patients from both groups had confirmed response (ITT-TLOVR algorithm; table 2). There were proportionally more virological failures in the rilpivirine group than in the efavirenz group (table 2). Our model-predicted responses, with the covariate log10 baseline plasma viral load, were similar to the ITT-TLOVR responses (table 2). The estimated difference in ITT-TLOVR response from our logistic-regression model was -0·4% (95% CI -5·9% to 5·2%). Since the lower limit of the 95% CI for the difference between rilpivirine and efavirenz was greater than both -12% (p<0·0001) and -10% (p=0·0007), we established non-inferiority at the 12% (primary endpoint) and 10% margins. However, we did not show superiority at the 5% significance level. Analysis of our per-protocol population confirmed that rilpivirine was non-inferior to efavirenz in confirmed response (table 2). In a sensitivity analysis excluding patients who discontinued for reasons other than virological failure according to our resistance analysis (ITT-TLOVR, population censored for non-virological failure), response rates were 86% (287 of 333) in our rilpivirine group and 94% (285 of 303) in our efavirenz group (difference -7·9%, 95% CI -12·5% to -3·2%). The percentage of responders for the two treatments increased over time, with no notable differences between the two groups (figure 2). Figure 2 also shows a steady increase from baseline in mean CD4 cell count.
The proportion of responders in the group of patients who self-reported greater than 95% adherence (M-MASRI; although data were not available for all patients) was 86% (236 of 275) for rilpivirine and 87% (229 of 262) for efavirenz. For patients who were 95% adherent or less, the proportion of responders was 68% (30 of 44) versus 73% (41 of 56), respectively. The median adherence of patients in the 95% adherent or less category who were treated with rilpivirine was 91% (n=44) and in patients treated with efavirenz was 92% (n=56).
The proportion of responders for patients with baseline viral load of 100 000 copies per mL or less was 90% (162 of 181) for rilpivirine versus 83% (136 of 163) for efavirenz. The proportion of responders for baseline viral load of 100 000 copies per mL to 500 000 or less copies per mL, was 79% (104 of 131) versus 83% (111 of 134), respectively. For baseline viral load of greater than 500 000 copies per mL, the proportion of responders was 62% (21 of 34) versus 81% (38 of 47). However, since some of the numbers of patients in these categories are small, the results in patients with 95% or less M-MASRI adherence and high baseline viral load should be interpreted with caution.
There was a greater proportion of virological failures according to our resistance analysis in the rilpivirine group versus the efavirenz group (table 3). At the time of failure, a similar high proportion of virological failures in each treatment group developed at least one treatment-emergent NNRTI RAM, whereas the proportion of virological failures with one or more treatment-emergent International AIDS Society-USA (IAS-USA) N(t)RTI RAM25 was higher in the rilpivirine group (table 3). The most common treatment-emergent NNRTI RAM in the rilpivirine group was E138K; in the efavirenz group K103N was the principal NNRTI RAM. M184I, V, or both were the most common IAS-USA N(t)RTI RAMs in both groups.
Our safety analysis included data from patients treated beyond week 48 (table 4). Adverse events were generally mild-to-moderate (grade 1 or 2). The incidence of grade 2 or greater adverse events possibly related to treatment was greater for efavirenz (p<0·0001 rilpivirine vs efavirenz; Fisher's exact test, preplanned analysis). The most commonly reported grade 2 or greater adverse events possibly related to treatment in 2% or greater of patients in either group were dizziness, abnormal dreams and nightmares, insomnia, nausea, and any rash (table 4). The incidences of serious adverse events, irrespective of relatedness, were similar between groups. There was one death in the efavirenz group due to Burkitt's lymphoma, which was unrelated to treatment (table 4). The incidence of discontinuations due to adverse events was greater for efavirenz (table 4).
Neurological events of interest (55 [16%] of 346 patients in the rilpivirine group vs 126 [37%] of 344 in the efavirenz group; p<0·0001) and psychiatric events of interest (50 [15%] vs 86 [25%], respectively; p=0·0006) possibly related to treatment (any grade) were at a significantly lower incidence with rilpivirine than with efavirenz. Individual neurological adverse events in 2% or greater of patients were dizziness (22 of 346 vs 85 of 344; p<0·0001), headache (22 vs 15), somnolence (12 vs 21), and disturbance in attention (two vs ten). Psychiatric adverse events in 2% or greater of patients were abnormal dreams and nightmares (32 of 346 vs 49 of 344; p=0·045), insomnia (14 vs 23), depression (six vs nine), anxiety (two vs eight), and sleep disorder (two vs 11). Most neurological and psychiatric adverse events of interest were grade 1 or 2 in severity, and the prevalence of these adverse events declined after the first 4-8 weeks of treatment in both groups.
The incidence of any rash possibly related to treatment (any grade) was lower (p<0·0001) for rilpivirine (4%; 12 of 346) than for efavirenz (15%; 50 of 344); most rashes were grade 1 or 2. Grade 3 rash was reported in one patient treated with rilpivirine and two patients treated with efavirenz, and no grade 4 rash was reported. One patient on rilpivirine discontinued because of a treatment-related rash versus three on efavirenz. Rash resolved with continuous dosing in both treatment groups in those who remained on study treatment.
Treatment-emergent grade 3 or 4 laboratory abnormalities happened at an incidence of 10% with rilpivirine compared with 16% for efavirenz. With the exception of hypophosphataemia, individual grade 3 or 4 laboratory abnormalities happened in a lower proportion of patients treated with rilpivirine versus those treated with efavirenz (table 4).
There were no relevant increases from baseline at week 48 in mean low-density lipoprotein-cholesterol (LDL-C) and triglyceride concentrations for rilpivirine (table 5). However, LDL-C and triglycerides increased with efavirenz (table 5). Rilpivirine was associated with lower increases than efavirenz in total cholesterol and high-density lipoprotein-cholesterol (HDL-C; table 5). There was no difference in the change from baseline at week 48 in the total cholesterol over HDL-C between groups.
There was a small increase from baseline in mean serum creatinine concentration for rilpivirine at our first on-treatment assessment, but then the concentration remained stable over the 48-week treatment period (range 5·69-9·07 μmol/L), whereas values remained around baseline for efavirenz (range 0·10-2·38 μmol/L). Consequently, eGFRcreat remained slightly below baseline levels with rilpivirine, but within normal limits (mean decreases were 8-11 mL/min per 1·73 m2), and at about baseline levels with efavirenz. We did not record grade 3 or 4 abnormalities in creatinine, and no abnormalities were reported as adverse events. There were no discontinuations due to renal adverse events.
Overall, QT interval corrected according to Fridericia's formula (QTcF) increased over time up to week 48 for both rilpivirine and efavirenz, with no relevant difference between groups. The mean changes from baseline were 10·9 ms (95% CI 9·0-12·8) versus 12·0 ms (10·1-13·7), respectively. There were few adverse events potentially related to conduction abnormalities or to rate and rhythm disturbances (one of 346 vs two of 344; these were grade 1 in severity).
Discussion
Once-daily oral 25 mg rilpivirine showed non-inferior efficacy compared with once-daily 600 mg efavirenz at 48 weeks in our trial. Both study drugs achieved response rates (table 2) that are among the highest compared with earlier studies of efavirenz in combination with tenofovir-disoproxil-fumarate and emtricitabine in treatment-naive patients.24, 26 More patients discontinued treatment because of intolerability with efavirenz, although more discontinued because of virological failure with rilpivirine. The CD4 cell count increased to a similar extent in both groups. These efficacy results are consistent with those reported in the companion phase 3 trial, THRIVE,14 although virological failure rates for rilpivirine were lower in THRIVE, and the difference in virological failures with efavirenz was smaller.
Our finding of more virological failures in the rilpivirine group than in the efavirenz group differs from the phase 2b study.9 Although self-reported, suboptimum adherence (<95% by M-MASRI) was associated with lower responses in both treatment groups, this alone might not fully explain the difference in virological failure rates between patients on the two study drugs. The lower response in patients on rilpivirine with baseline viral load greater than 500 000 copies per mL compared with efavirenz probably contributed to these findings. Analyses are ongoing to better understand the role of factors such as adherence, drug exposure, and baseline viral load in virological failure. Because of the statistical power limitations for the individual trials, results of additional exploratory analyses of response factors will be reported in pooled analyses of data from ECHO and THRIVE. Additionally, pharmacokinetics and pharmacokinetic and pharmacodynamic relations will be presented elsewhere for the pooled data.
Consistent with other NNRTI regimens,27 at the time of failure, the proportion of patients with at least one NNRTI RAM was high and similar in both groups. The most prevalent treatment-emergent NNRTI RAMs were consistent with data from TMC278-C2049, 10 and THRIVE.14 Consequently, cross-resistance exists between rilpivirine and etravirine-initial findings from the pooled ECHO and THRIVE resistance analysis have shown that, of the 31 patients on rilpivirine who experienced virological failure and were phenotypically resistant to rilpivirine, 28 (90%) were cross-resistant to etravirine, 27 (87%) to efavirenz, and 14 (45%) to nevirapine.28 Studies are in progress to further characterise the effect of specific combinations of NNRTI RAMs that are found in viruses emerging from selection with rilpivirine, on the antiviral activity of etravirine. Of the 12 patients on efavirenz who experienced virological failure and were phenotypically resistant to efavirenz, all were resistant to nevirapine but remained sensitive to etravirine.
Treatment-emergent IAS-USA N(t)RTI RAMs were more common in the rilpivirine group than in the efavirenz group among virological failures. The most common treatment-emergent N(t)RTI RAMs in both groups were M184I and V, which are associated with reduced susceptibility to emtricitabine and lamivudine.29 An analysis of phenotypic sensitivity to NNRTIs will be presented separately for the pooled data.
As we anticipated from the phase 2b data (panel),9, 10 the safety and tolerability profile of rilpivirine was better than that of efavirenz. Discontinuations due to adverse events were less common with rilpivirine than with efavirenz (table 4). The incidence of any grade 2 or higher adverse events possibly related to treatment in the rilpivirine group was lower than in the efavirenz group (table 4). Additionally, incidences of rash, neurological adverse events of interest and psychiatric adverse events of interest were lower for rilpivirine than efavirenz. Increases in some proatherogenic lipid variables were also smaller with rilpivirine, but the difference was not substantial in the total cholesterol to HDL-C ratio. Increases in creatinine for rilpivirine were small, and might be related to a rilpivirine effect on the disposition of creatinine, rather than to renal toxicity. Indeed, with the use of cystatin C, thought to be a better indicator of GFR than creatinine,31 rilpivirine did not decrease GFR in THRIVE.14 Rilpivirine might increase the exposure to tenofovir after tenofovir-disoproxil-fumarate, but this increase is not thought clinically relevant.32 Long-term follow-up and assessment is available from the phase 2b trial and will be provided by the week-96 analysis of these phase 3 trials.
Panel
Research in context
Systematic review
A recent Cochrane review30 stated that the non-nucleoside reverse transcriptase inhibitors (NNRTIs) efavirenz and nevirapine are equally effective in suppressing infection with HIV but cause side-effects that can limit their use, highlighting a need for additional first-line NNRTIs with a better safety profile. We searched PubMed for supporting evidence up to May, 2011, with the search term "rilpivirine"; we selected randomised controlled trials and clinical trials published in English. Rilpivirine has only been assessed in one phase 2b study (NCT NCT00110305; TMC278-C204). This study showed that rilpivirine provided long-term (longer than 96 weeks) efficacy and tolerability in treatment-naive adults infected with HIV-1, with response rates similar to efavirenz. In this study, all rilpivirine doses resulted in similar response rates. Grade 2-4 adverse events at least possibly related to study medication, including nausea, dizziness, abnormal dreams/nightmare, rash, somnolence, and vertigo, were less common with TMC278 than with efavirenz in the context of an open-label trial (only the dose of rilpivirine was blinded).9
Interpretation
In combination with the THRIVE trial,14 which assessed the safety and efficacy of rilpivirine with three different background regimens (tenofovir-disoproxil-fumarate plus emtricitabine, zidovudine plus lamivudine, or abacavir plus lamivudine) versus efavirenz, our data suggest that once-daily rilpivirine is probably a valuable treatment option for antiretroviral-naive patients infected with HIV-1.
There are several limitations to our trial. First, our study used only one N(t)RTI background regimen (tenofovir-disoproxil-fumarate and emtricitabine). However, the THRIVE trial also assessed zidovudine with lamivudine or abacavir with lamivudine, as well as tenofovir-disoproxil-fumarate and emtricitabine selected by the investigator. The consistent findings, irrespective of N(t)RTI background, show the possibility of combining rilpivirine with other antiretroviral drugs. Second, our trial was not powered to assess comparisons of efficacy in various subsets of patients, and therefore it is difficult to generalise our findings to the overall population of patients. However, subgroup analyses of the combined ECHO and THRIVE populations by sex, region, ethnic origin, clade, and hepatitis B and C co-infection28, 33, 34 show that the efficacy of rilpivirine and efavirenz are similar, suggesting broader applicability of these data. Furthermore, in routine clinical practice, patients might harbour transmitted resistance, which could reduce response rates. For our trial, we used a comprehensive NNRTI RAM list to screen out patients potentially resistant to NNRTIs. The present prevalence of E138K in routine clinical resistance testing is low (<1%).35 Finally, since our trial had a double-dummy design, patients had to take their study medication twice daily, rather than the normal once-daily dosing for both NNRTIs, although we did not know what effect this design feature had on response rates. Also, patients were required to take rilpivirine (or matching placebo) with a meal. This recommendation might have been overlooked in our double-blind trial, resulting in some patients taking rilpivirine on an empty stomach and a lower rilpivirine exposure than expected in some cases.
The selection of rilpivirine will rely on assessing individual benefits and risks for individual patients. These data suggest that once-daily rilpivirine, perhaps as a single-tablet regimen in combination with tenofovir-disoproxil-fumarate and emtricitabine,36 is expected to be a valuable treatment option for patients infected with HIV who have not been previously treated with antiretroviral drugs.
|
|
| |
| |
|
|
|
