|
Efficacy and safety of etravirine (TMC125) in patients with highly resistant HIV-1: primary 24-week analysis
|
|
|
AIDS: Volume 21(6) 30 March 2007 p F1-F10
The TMC125-C223 Writing Group
Received 7 October, 2006
Revised 17 November, 2006
Accepted 27 November, 2006
".....Etravirine is the first NNRTI to show clinical efficacy over 24 weeks, with or without enfuvirtide or a PI, after previous nevirapine or efavirenz failure, combined with PI failure, in patients with very limited treatment options.
....A mean viral load reduction of 1 log10 copies/ml or greater with etravirine 800 mg was demonstrated in patients with less than 3 NNRTI resistance-associated mutations at screening (n = 50). Although the 0.66 log10 viral load reduction observed in patients with three or more NNRTI mutations was lower than with etravirine overall (-1.18 log10 copies/ml), it was greater than with control (-0.19 log10 copies/ml)..... At week 24, the proportion of patients with a viral load decrease of 1.0 log10 copies/ml or greater from baseline was statistically higher with etravirine (400 mg, 36.3%, P = 0.005; 800 mg, 41.8%, P < 0.001) compared with control (7.5%). The proportion of patients with HIV-1-RNA levels less than 400 copies/ml at week 24 was also significantly higher with etravirine 400 and 800 mg than controls (30.0%, P = 0.018 and 38.0%, P = 0.002 versus 7.5%, respectively). More etravirine-treated patients achieved HIV-1-RNA levels of less than 50 copies/ml at week 24, but this was not statistically significant (400 mg, 21.3%, P = 0.133 and 800 mg, 17.7%, P = 0.218 versus 7.5% for controls)....
.... Overall, there was no significant difference in efficacy between the two etravirine doses. However, in patients not using enfuvirtide, however, those receiving etravirine 800 mg had a greater viral load reduction at week 24 than those who received etravirine 400 mg, although more etravirine 400 mg patients had one or less active antiretroviral drugs in the OBR than with 800 mg (77.6 versus 51.7%, respectively). Furthermore, subanalyses indicated a trend for 800 mg etravirine to perform better than 400 mg when not as well supported by the OBR. In addition, there were no relevant safety differences between the doses. Therefore, the etravirine 800 mg twice-daily dose was selected for further investigation. A new formulation of etravirine has been developed, providing comparable exposure to the 800 mg twice-daily dose used here when given as 200 mg twice a day in healthy volunteers [12]. This new etravirine formulation is being used in phase III clinical studies....
.... Rashes occurring with etravirine were generally early onset, mild to moderate in severity and lasted less than a week on average. Although characteristic of rashes associated with other NNRTI, the incidence was less than that cited for efavirenz [13]. There was no mucosal or hepatic involvement or Stevens-Johnson syndrome in patients treated with etravirine, and no relationship was observed between the CD4 cell count and rash.... Overall, the rate of permanent discontinuation of etravirine because of rash was 3.1% in this 24-week analysis compared.... As NNRTI-related rash usually occurs early during treatment, this is likely to provide a reasonable indication of the relative frequency of discontinuations caused by rash. Neuropsychiatric events (insomnia and headache) were predominantly mild to moderate...." See actual data on adverse events reported below
Abstract
Objective: TMC125-C223 is an open-label, partially blinded, randomized clinical trial to evaluate the efficacy and safety of two dosages of etravirine (TMC125), a non-nucleoside reverse transcriptase inhibitor (NNRTI) with activity against wild-type and NNRTI-resistant HIV-1.
Design: A total of 199 patients were randomly assigned 2: 2: 1 to twice-daily etravirine 400 mg, 800 mg and control groups, respectively. The primary endpoint was a change in viral load from baseline at week 24 in the intention-to-treat population.
Methods: Patients had HIV-1 with genotypic resistance to approved NNRTIs and at least three primary protease inhibitor (PI) mutations. Etravirine groups received an optimized background of at least two approved antiretroviral agents [nucleoside reverse transcriptase inhibitors (NRTI) and/or lopinavir/ritonavir and/or enfuvirtide]. Control patients received optimized regimens of at least three antiretroviral agents (NRTIs or PIs and/or enfuvirtide).
Results:
The mean change from baseline in HIV-1 RNA at week 24 was -1.04, -1.18 and -0.19 log10 copies/ml for etravirine 400 mg twice a day, 800 mg twice a day and the control group, respectively (P < 0.05 for both etravirine groups versus control).
Etravirine showed no dose-related effects on safety and tolerability. No consistent pattern of neuropsychiatric symptoms was observed. There were few hepatic adverse events, and rashes were predominantly early onset and mild to moderate in severity.
Conclusion: Etravirine plus an optimized background significantly reduced HIV-1-RNA levels from baseline after 24 weeks in patients with substantial NNRTI and PI resistance, and demonstrated a favorable safety profile compared with control.
Resistance
The impact of baseline NNRTI resistance on the virological outcome at week 24 was analyzed, and data are presented for patients who received etravirine 800 mg (the selected dose from phase II). Mean viral load changes from baseline at week 24 for patients with no (historically documented NNRTI resistance; n = 15), one (n = 18), two (n = 17) and three or more (n = 29) baseline NNRTI-resistance-associated mutations were -1.82, -1.65, -1.00 and -0.66 log10 copies/ml, respectively, compared with a -0.19 log10 copies/ml reduction for controls.
At baseline, mutations associated with an etravirine fold-change greater than 10 (arbitrary threshold), and always seen in combination with up to four other NNRTI-resistance-associated mutations, were K101P, V179E, V179F, Y181V, G190S and M230L. With mutations at positions 179, 190 and 230, when the fold-change was greater than 10, additional mutations always included Y181C. All these mutations, including Y181C, were previously found to be associated with increased etravirine fold-change in vitro [7]. In this study, K103N and Y181C were most prevalent and were generally associated with additional NNRTI resistance mutations. In an analysis adjusted for baseline viral load, baseline CD4 cell count, the number of accompanying NNRTI mutations and naive use of enfuvirtide, viral load reductions were not statistically different for those with and without K103N or Y181C.
Introduction
Non-nucleoside reverse transcriptase inhibitors (NNRTI), combined with a nucleoside/nucleotide reverse transcriptase inhibitor (NRTI) backbone, are used extensively to treat HIV-1 infection. The prevalence of resistance to approved NNRTI is substantial [1], and there is evidence of the increasing incidence of primary or transmitted NNRTI resistance [2]. NNRTI resistance can be measured for up to 6 years [3], and resistance to one NNRTI currently signifies the loss of this class for the patient [4]. A potent NNRTI with activity against HIV-1 resistant to current NNRTI and a favorable safety and tolerability profile would provide a much needed treatment option for many patients [5].
Etravirine (TMC125), is a diarylpyrimidine NNRTI, selected for activity against both wild-type and NNRTI-resistant HIV-1 [6], with a high genetic barrier to the development of resistance in vitro. This is probably because of its molecular flexibility, allowing multiple binding conformations in reverse transcriptase [6,7].
Etravirine has demonstrated safety and efficacy in short-term monotherapy studies in both treatment-naive and NNRTI-resistant HIV-1-infected patients [8,9]. The TMC125-C223 study is a multicenter, open-label, partially-blinded, randomized phase IIb clinical trial evaluating the efficacy and safety of etravirine 400 or 800 mg administered twice a day compared with a control group (standard-of-care) for 48 weeks in highly treatment-experienced HIV-1-infected patients with multidrug resistance. The 24-week planned primary analysis is presented here.
Methods
Study design
Patients were randomly assigned 2: 2: 1 to receive etravirine 400 or 800 mg twice a day (TF035 formulation), or an active control regimen (Fig. 1). Using a telephone-based interactive voice response system, patients were stratified by intended enfuvirtide use, CD4 cell count less than 100 cells/μl, CD4 cell count 100 cells/μl or greater, and treatment interruption at screening. Subjects and investigators were blinded to the etravirine dose, but not randomization to etravirine or control. Both etravirine groups received an optimized background regimen (OBR), consisting of two or more approved antiretroviral drugs: NRTI and/or lopinavir/ritonavir or enfuvirtide, in any combination. The control group received an optimized regimen consisting of three or more approved antiretroviral drugs: NRTIs, and/or protease inhibitors (PIs) and/or enfuvirtide. All antiretroviral agents, except etravirine, were investigator-selected and guided by VirtualPhenotype (Virco, Mechelen, Belgium) and treatment history. Low-dose ritonavir (< 800 mg/day) was not counted as a separate PI. Lopinavir/ritonavir was the only PI allowed for use with etravirine, because of the low likelihood of interactions between this PI regimen and etravirine. The protocol did not restrict the continuation of antiretroviral drugs used during screening or require a minimum number of active agents in the OBR. Dose adjustments and substitutions (restricted to NRTI for etravirine groups) were allowed for tolerability reasons only.
Study visits were at screening, baseline, weeks 1, 2 and 4, then every 4 weeks until week 24, and thereafter every 8 weeks until study termination. Follow-up visits were performed one and 4 weeks after study termination or premature discontinuation. Samples were taken for virology (including resistance testing), immunology, hematology, biochemistry and pharmacokinetic analysis.
The research protocol received ethical approval at all 50 study centers in the United States, and written informed consent was obtained from all participants (enrolled between April and November 2004). Details of the study were registered with ClinicalTrials.gov ( http://clinicaltrials.gov ), trial identifier NCT00081978.
Study population
Eligible subjects were HIV-1-infected adults with resistance to currently approved NNRTI (one or more NNRTI resistance-associated mutations in previous genotype or at screening), three or more primary PI mutations at screening [10], three or more months of previous NRTI experience, and a plasma viral load greater than 1000 HIV-1-RNA copies/ml. Subjects received either a stable antiretroviral regimen for eight or more weeks before screening or a treatment interruption for four or more weeks. Exclusion criteria included chronic hepatitis B/C with aminotransferases greater than three times the upper limit of normal, evidence of significantly decreased hepatic function, serum creatinine greater than two times the upper limit of normal, permanent discontinuation of any NNRTI as a result of cutaneous events, and any grade 3/4 toxicity (excluding glucose, lipid and γ-glutamyl-transferase elevations).
Efficacy and safety analysis
The primary endpoint was a change in HIV-1 log10 copies/ml from baseline at week 24. Secondary virological response parameters were the proportion of patients showing a 1.0 log10 or greater reduction in viral load, and with less than 400 or less than 50 HIV-1-RNA copies/ml. Virological failure was defined as a less than 1.0 log10 reduction in viral load by week 12, or two consecutive viral load measurements greater than 0.5 log10 above nadir after week 12. Plasma samples were collected in ethylenediamine tetraacetic acid tubes and measurements of viral load were by quantitative polymerase chain reaction (Amplicor HIV-1 monitor, v. 1.5; Roche Diagnostics, Branchburg, New Jersey, USA). The change in the CD4 cell count from baseline was assessed. Subgroup analyses of the primary efficacy parameter were performed after stratifying patients by actual enfuvirtide use, treatment interruption, lopinavir use, tenofovir/didanosine use, baseline viral load, baseline CD4 cell count and the number of active drugs in the OBR. For the pharmacokinetic analyses, etravirine plasma concentrations were determined by validated liquid chromatography tandem mass spectrometry assay. Safety analyses included laboratory tests, electrocardiogram changes, and the incidence of all adverse events and HIV-related events. Adverse events were graded using the AIDS Clinical Trials Group severity list [11]. For laboratory tests with no defined AIDS Clinical Trials Group grades, the World Health Organization toxicity grades were applied.
Resistance analysis
Resistance testing was performed by VirtualPhenotype (Virco) and phenotype assay (Antivirogram; Virco). The VirtualPhenotype was used to guide treatment decisions, whereas resistance data reported here were from Antivirogram. NNRTI resistance mutations were those of the International AIDS Society (IAS) - USA Drug Resistance Mutations Group [10], plus other important known NNRTI resistance mutations (A98G, K101E, K101P, K101Q, K103H, K103S, K103T, Y181V, G190E, K238N, K238T and Y318F). The number of antiretroviral drugs to which a patient isolate was sensitive was determined from the phenotype result (excluding etravirine). Virus was assumed to be enfuvirtide sensitive if the patient had no previous exposure.
Statistical methods
The study aimed to enroll 150 patients. Sixty patients per etravirine group was predicted to detect at least a 0.5 log10 difference between them in viral load change at week 24 between the two etravirine groups (primary study endpoint), with a common standard deviation of 0.8, 90% power and 5% significance level (two-sided). Thirty control patients were predicted to provide 80% power to detect the same viral load change between the control and etravirine groups.
Efficacy and safety analyses used the intention-to-treat population, defined as all patients who had received one or more dose of study medication. For patients prematurely discontinuing, viral load data were imputed from discontinuation onward with the baseline value (non-completer equals failure). The viral load and CD4 cell count change from baseline to week 24, and the relationship between pharmacokinetics and antiviral activity, were analysed by analysis of covariance with the factors treatment group, treatment interruption at screening, enfuvirtide use (not used, re-used or naive), and the covariates baseline CD4 cell count and viral load. The subgroup analyses of the primary endpoint were adjusted for the same factors (including enfuvirtide usage). Virological response rates were calculated according to the time to the loss of virological response algorithm, and analyzed by logistic regression with the covariates above.
Safety parameters were analyzed by descriptive statistics and frequency tabulations. Clinical laboratory tabulations included the most serious toxicity grading after baseline.
Results
Patient disposition
The study included 199 heavily treatment-experienced subjects. Participant characteristics are presented in Table 1. The median duration of HIV infection was 15 years and the median number of previous antiretroviral agents was 12. Most patients had one or more NNRTI resistance mutations detected at screening and the remainder had previous genotypes documenting NNRTI resistance mutation(s). The median number of primary PI, NNRTI and NRTI mutations at screening was four, two and six, respectively.
![Patient-1.gif](../images/032807/Patient-1.gif)
There was an average of 2 (0-5) NNRTI mutations for each of the 3 treatment groups. The K103 N was present in 40%, 36%, and 42% on the 400, 800, & Control groups. The Y181C was in 40%, 36%, & 37% on the 400, 800 and Control groups.
![Median-2.gif](../images/032807/Median-2.gif)
![reg-3.gif](../images/032807/reg-3.gif)
![Fig2-4.gif](../images/032807/Fig2-4.gif)
Treatment groups were generally well balanced, although control patients more commonly used an active PI compared with etravirine patients (13.9 versus 4.3%, respectively). Whereas some control patients used dual-boosted PI, only one received amprenavir/lopinavir, the combination with complex interactions that could potentially jeopardize therapy efficacy. There was comparable NRTI use, including tenofovir/didanosine (23.3% in etravirine-treated patients versus 27.5% for controls). Despite stratification by intended enfuvirtide use, there was less use of enfuvirtide among control patients than with etravirine 400 or 800 mg (52.5 versus 62.5% and 64.6%, respectively). In addition, more control patients were enfuvirtide-experienced (72.5 versus 53.7% and 58.2%, respectively). Overall, 83.5% of etravirine-treated patients used at least one drug (excluding etravirine) in the regimen that was scored as active compared with 74.4% of control patients (Table 1).
Overall, 26.3, 24.1 and 95.0% of patients discontinued in the etravirine 400, 800 mg and control groups, respectively, and the median treatment duration was 34.2, 34.7 and 17.9 weeks, respectively. Of the control patients, 75.0% discontinued because of virological failure compared with 6.3 and 5.1% of the etravirine 400 and 800 mg groups, respectively.
Efficacy: changes in plasma viral load
The mean change from baseline in viral load at week 24 was -1.04 and -1.18 log10 copies/ml with etravirine 400 and 800 mg twice a day, respectively, significantly greater than the -0.19 log10 copies/ml observed with control patients (P = 0.005 and P < 0.001, respectively; intention-to-treat analysis; Fig. 2a). There was no significant difference in viral load change from baseline at week 24 between etravirine doses, except in the subgroups not using enfuvirtide and not using lopinavir/ritonavir, in which this was greater for patients receiving etravirine 800 than 400 mg (Fig. 2b; note that subgroups were categorized by actual enfuvirtide use rather than intended use at screening). The use of enfuvirtide in enfuvirtide-naive patients was a significant predictor of virological response (P < 0.001), associated with considerably enhanced viral load reduction. Other predictors included the baseline CD4 cell count (P = 0.043), and baseline fold-change in antiretroviral 50% effective concentration for etravirine (P = 0.025).
At week 24, the proportion of patients with a viral load decrease of 1.0 log10 copies/ml or greater from baseline was statistically higher with etravirine (400 mg, 36.3%, P = 0.005; 800 mg, 41.8%, P < 0.001) compared with control (7.5%). The proportion of patients with HIV-1-RNA levels less than 400 copies/ml at week 24 was also significantly higher with etravirine 400 and 800 mg than controls (30.0%, P = 0.018 and 38.0%, P = 0.002 versus 7.5%, respectively). More etravirine-treated patients achieved HIV-1-RNA levels of less than 50 copies/ml at week 24, but this was not statistically significant (400 mg, 21.3%, P = 0.133 and 800 mg, 17.7%, P = 0.218 versus 7.5% for controls). Virological responses were not statistically different between etravirine doses.
Fig. 2. Efficacy to week 24.
Patients taking Fuzeon who were naive to Fuzeon and patients using more active agents had better viral responses. This is displayed in these graphs below.
![ACT-5.gif](../images/032807/ACT-5.gif)
Etravirine pharmacokinetics (area under the curve over 12 h, or predose concentration) were not statistically associated with viral loads at week 24.
Efficacy: changes in CD4 cell count
Although not statistically significant, the mean change in CD4 cell count from baseline at week 24 with etravirine 400 and 800 mg was numerically greater than controls, being +47, +48 and +10 cells/μl, respectively (Fig. 2d).
Safety and tolerability
The safety analysis compares etravirine doses (Table 2). Control data are presented for completeness, although the significant difference in treatment duration confounds comparisons with the etravirine groups. For adverse events in which the incidence was at least 5% greater in the combined etravirine group than controls, incidence corrected for duration of exposure was not substantially higher with etravirine, with the exception of infections (21.8 versus 14.4%) and psychiatric disorders (16.9 versus 7.2%).
Overall, the incidence of psychiatric adverse events with etravirine was 10.7% (2.5% in Control group). Psychiatric events considered to be possibly etravirine related were mood swings and nightmares (one subject each) in the 400 mg group and abnormal dreams, anxiety and depression (one subject each) in the 800 mg group. There were no grade 3/4 psychiatric events or discontinuations because of psychiatric events. The most common nervous system events with etravirine were headache and insomnia. Neither of these events led to discontinuation.
![TableOver-6.gif](../images/032807/TableOver-6.gif)
The most common adverse events with etravirine, irrespective of causality, were diarrhea, enfuvirtide-related injection site reactions, pyrexia, fatigue, nausea, headache and insomnia (Table 2). There was no etravirine dose effect on grade 3/4 adverse events.
There were four deaths during the study period, one in the control and three in the etravirine 400 mg group, of which one (cardiopulmonary failure and myocardial infarction) was considered to be possibly etravirine related (multiple risk factors for cardiovascular disease, including PI-containing HAART, dyslipidemia and obesity were present). The other three deaths were caused by central nervous system lymphoma, pseudomonal sepsis and cardiac arrest (control), all of which were considered to be not, or doubtfully, related to study medication. No single adverse event led to discontinuation in more than two patients, except drug-related rash in five patients (3.1%), of which four received etravirine 800 mg. There was no other etravirine dose differentiation for adverse events leading to discontinuation.
The incidence of grade 3/4 laboratory abnormalities was similar between etravirine groups (Table 2). The incidence of hepatic abnormalities was low. In additon, there was no consistent, clinically relevant or dose-associated increase in median total cholesterol or triglyceride levels. The most common treatment-emergent grade 3/4 laboratory abnormality with etravirine was elevated pancreatic amylase, occurring in 13 patients (8.2%). There was no consistent, clinically relevant or dose-associated increase in median pancreatic amylase levels; 34 patients (21.5%) had grade 1/2 elevations. Three etravirine patients had clinical pancreatitis but all had other risk factors (didanosine with and/or without tenofovir or previous pancreatitis). One patient permanently discontinued because of clinical pancreatitis and one because of elevated pancreatic amylase and lipase levels, both of whom received 400 mg etravirine.
The incidence of rash (all forms) was 19.5% in etravirine patients (7.5% in Controls). The incidence of rash considered to be possibly etravirine related was 15.1%, with no significant difference between doses. The median time of rash onset and duration was 13 and 5 days, respectively, with no new rash events reported after week 12. Two cases of grade 3 drug-related rash considered to be at least possibly etravirine related were diffuse rash with fever without hepatic or mucosal involvement. No association was apparent between rash events and CD4 cell counts.
There was no consistent, clinically relevant or dose-associated increase in the median total cholesterol or triglyceride levels.
Discussion
This study demonstrates the short-term virological efficacy of etravirine 400 or 800 mg administered twice a day with an OBR for patients with advanced HIV-1 disease and multidrug-resistant virus. Statistically significant reductions in plasma viral load with etravirine were sustained over 24 weeks.
Pronounced virologic responses were seen in the etravirine groups compared with controls in all subgroups analyzed, including with or without enfuvirtide or lopinavir/ritonavir, and in patients with baseline CD4 cell counts less than 100 cells/μl and 100 cells/μl or greater. The use of one or more active antiretroviral drug with etravirine was a significant predictor of a viral load change at week 24, and adds to a growing body of data showing the benefits of two or more active drugs for salvage regimens.
Although randomization was stratified by intended enfuvirtide use, discrepancies between this and actual use produced an imbalance between the groups. As a result of the open-label design of this trial, some investigators may have decided against using enfuvirtide in patients randomly assigned to control, to preserve it as a future treatment option when an additional active agent was available. According to a pre-planned analysis, the change in viral load from baseline to week 24 was analyzed using an analysis of covariance model that considered enfuvirtide use as a factor. Therefore the changes in viral load reported in this publication were adjusted for the observed imbalances of enfuvirtide across the groups.
Overall, there was no significant difference in efficacy between the two etravirine doses. However, in patients not using enfuvirtide, however, those receiving etravirine 800 mg had a greater viral load reduction at week 24 than those who received etravirine 400 mg, although more etravirine 400 mg patients had one or less active antiretroviral drugs in the OBR than with 800 mg (77.6 versus 51.7%, respectively). Furthermore, subanalyses indicated a trend for 800 mg etravirine to perform better than 400 mg when not as well supported by the OBR. In addition, there were no relevant safety differences between the doses. Therefore, the etravirine 800 mg twice-daily dose was selected for further investigation. A new formulation of etravirine has been developed, providing comparable exposure to the 800 mg twice-daily dose used here when given as 200 mg twice a day in healthy volunteers [12]. This new etravirine formulation is being used in phase III clinical studies.
A mean viral load reduction of 1 log10 copies/ml or greater with etravirine 800 mg was demonstrated in patients with less than 3 NNRTI resistance-associated mutations at screening (n = 50). Although the 0.66 log10 viral load reduction observed in patients with three or more NNRTI mutations was lower than with etravirine overall (-1.18 log10 copies/ml), it was greater than with control (-0.19 log10 copies/ml).
The overall safety profiles of the two etravirine doses were similar. Comparisons between the etravirine and control groups were confounded by differences in treatment exposure as a result of the high discontinuation of control patients and the unblinded nature of the active controls. Rashes occurring with etravirine were generally early onset, mild to moderate in severity and lasted less than a week on average. Although characteristic of rashes associated with other NNRTI, the incidence was less than that cited for efavirenz [13]. There was no mucosal or hepatic involvement or Stevens-Johnson syndrome in patients treated with etravirine, and no relationship was observed between the CD4 cell count and rash, in contrast to nevirapine [14,15]. Overall, the rate of permanent discontinuation of etravirine because of rash was 3.1% in this 24-week analysis compared with 7.0 and 1.7% at 48 weeks for nevirapine and efavirenz, respectively [13,15]. As NNRTI-related rash usually occurs early during treatment, this is likely to provide a reasonable indication of the relative frequency of discontinuations caused by rash. Neuropsychiatric events (insomnia and headache) were predominantly mild to moderate. There was no pattern of psychiatric events.
The low genetic barrier to resistance and the extensive cross-resistance of currently available NNRTI has limited their use to a single line of therapy. Etravirine is the first NNRTI to show clinical efficacy over 24 weeks, with or without enfuvirtide or a PI, after previous nevirapine or efavirenz failure, combined with PI failure, in patients with very limited treatment options. Furthermore, it was generally safe and well tolerated in the short-term, with an acceptable safety profile for adverse events of specific interest to NNRTI, namely rash (a class-related side effect), hepatotoxicities (associated with nevirapine) [16] and neuropsychiatric symptoms and lipid elevations (associated with efavirenz) [13].
|
|
|
|
|
|
|