icon-folder.gif   Conference Reports for NATAP  
  International AIDS Conference
July 13-16, 2003, Paris
Back grey_arrow_rt.gif
ABC/AZT/3TC vs Efavirenz Based Regimens: ACTG 5095
Reported by Jules Levin
  Before launching into a report of the study presented today I have to tell you about my experience last night in Paris. I got into a taxi at my hotel to go to the Eiffel Tower. The taxi driver did not know where the Eiffel tower was. Would you believe this? He had to take out a map and still got lost. After getting to the Eiffel Tower, it was closed, perhaps because it was Sunday night.
Trip Gulick (Cornell University, NYC) reported preliminary results from ACTG Study 5095 which compared the triple nuke regimen of AZT/3TC/abacavir to two efavirenz based regimens: AZT/3TC/efavirenz (3 drugs) and AZT/3TC/abacavir plus efavirenz (EFV) (4 drugs). This is a randomized, double-blind study in treatment-naive patients which examines and compares the safety and viral efficacy of these three regimens. This study has been the subject of much fanfare and controversy.
The two efavrienz regimens were more potent in this study, but observers at the conference commented at the microphone after the oral presentation by Gulick that Trizivir (abacavir/3TC/AZT) has usefulness in certain circumstances: potency is one of several factors considered in selecting a regimen; ease of use and tolerability are considered for some patients. The decision of which regimen to initiate a patient with is one based on the circumstances of the patient.
Viral failure was higher for patients in this study taking ABC/3TC/AZT than either of the two EFV regimens (21% vs 11%). 74% of patients in this study achieved <200 copies/ml using AZT/3TC/abacavir vs 89% using the two EFV regimens. Using a time to viral failure evaluation, the triple nuke regimen demonstrated higher rates of viral failure. The adverse event profile appeared similar between the treatment regimens. One observer mentioned that if you fail the EFV regimen you can have NNRTI resistance plus nuke resistance, while only nuke resistance occurs if you fail the triple nuke regimen. But with viral failure of ABC/3TC/AZT you may develop more extensive nuke resistance.
The Data and Safety Monitoring Board stopped this study based on viral failure rates being higher and occurring sooner in the triple nuke arm compared to the two EFV-based regimens, not due to safety concerns. The DSMB recommended that the AZT/3TC/ABC arm be discontinued and unblinded due to demonstrated viral failure rate inferior to the two EFV regimens. So the results from the two EFV regimens have not been revealed yet.
1147 treatment-naive patients were enrolled in this study. 81% men, 40% white, 36% black, 21% latino, and 11% IDUs. Average baseline viral load was 71,000 copies/ml. 57% had <100,000 copies/ml and the results are stratified by whether a patient had below or above 100,000 copies/ml of baseline viral load. Average CD4 count was 238. The median follow-up for patients in this analysis is 32 weeks.
10% of patients discontinued initial treatment and there were no differences between the two treatment groups. 8% of patients had protocol-permitted substitutions of d4T for AZT, 5% substituted ddI for ABC, and 6% substituted nevirapine for EFV.
The primary study endpoint is viral failure: 167 patients reached this study endpoint of virologic failure-
--82 (21%) of 382 in the AZT/3TC/ABC arm
--85 (11%) 765 in the pooled EFV arms (EFV/AZT/3TC and EFV/ABC/AZT/3TC)
The difference is significant (p<0.001).
--Patients taking AZT/3TC/ABC were more likely to experience viral failure whether they had below or above 100,000 copies/ml at the start of the study.
Proportion of Patients with <200 and <50 cps/ml
--EFV arms: 89% had <200 and 83% had <50 copies/ml at week 48
--Abacavir: 74% had <200 and 61% had <50 copies/ml at week 48
In a post-hoc analysis (an analysis that was not planned until after the study started) viral failure rates were higher for patient taking AZT/3TC/ABC if they had at least one viral load <200 copies/ml (p<0.001). In other words, if a patient achieved a viral load <200 copies/ml at least once they were more likely to experience viral load rebound if they were taking AZT/3TC/ABC than if they were taking the EFV-based regimens. If a patient achieved <50 copies/ml at least once there appeared to be a trend towards higher rate of viral rebound but this was not significant (P=0.083).
CD$ responses were about the same for the triple nuke regimen and the two EFV regimens.
The adverse events profile were similar between the triple nuke regimen and the EFV regimens: grade 3/4 signs and symptoms, 12% for AZT/3TC/ABV and 15% for pooled EFV data; grade 3/4 lab abnormalities: 27% for triple nuke and 28% for pooled EFV regimens. Suspected ABC hypersensitivity reactions: 7% triple nuke, 8% pooled EFV regimens.
Resistance and Adherence
There were 82 viral failures on the triple nuke regimen: 22% had wild-type virus; 34% had the 3TC M184 mutation; 11% had M184V plus other NRTI mutations; 27% had <500 copies/ml so resistance test could not be performed. Gulick said that by patient self-report adherence was 100%, but we suspect that patient self-report is not always reliable. The fact that 22% of study patients had wild-type virus is suggestive that perhaps non-adherence may have occurred. All patients in the study took I think 7 pills per day. Gulick reported there was no difference in adherence between the treatment groups in the study. In real life, Trizivir is one pill twice daily, so this may effect the performance of Trizivir.
Gulick concluded that AZT/3TC/ABC is inferior to a regimen of EFV with AZT/3TC with or without ABC with respect to viral failure in treatment-naive patients.