This paper presents the interim results of AIDS Clinical Trials Group (ACTG) Study 5095, a comparison of the triple-nucleoside regimen abacavir + lamivudine + zidovudine (coformulated as Trizivir) with two efavirenz-containing regimens in initial treatment of HIV infection. This double-blind trial randomized a diverse group of 1147 treatment-naive subjects to one of three treatment arms: abacavir + lamivudine + zidovudine, abacavir + lamivudine + zidovudine + efavirenz, and abacavir + lamivudine + efavirenz.

A planned interim analysis revealed significant differences in rates of virologic failure (VF) between the triple-nucleoside regimen and the two efavirenz-containing regimens, leading to termination of the triple-nucleoside study arm. At a median of 32 weeks on treatment, 21% of the triple-nucleoside group versus 11% of the combined efavirenz groups had virologic failure (defined as viral load >200 copies/mL after 16 weeks on treatment). For those with data at 48 weeks, 61% of triple-nucleoside recipients had an HIV RNA level <50 copies/mL, compared to 83% of those on efavirenz. Higher rates of VF were seen in the triple-nucleoside arm regardless of parameters such as pretreatment viral load, CD4 cell count, and self-reported adherence. In post hoc analysis, among those who achieved viral suppression to <200 copies/mL, the time to VF was significantly shorter in the triple-nucleoside group (p <0.001).

Genotype testing performed on 57/82 of triple-nucleoside recipients with VF revealed wild-type virus in 22%, the M184V mutation alone in 34%, and M184V plus reverse transcriptase mutations (primarily thymidine analogue associated mutations (TAMS)) in 11%.

ACTG 5095 demonstrates that abacavir + lamivudine + zidovudine is inferior to efavirenz-containing regimens in the initial treatment of HIV infection. The findings of this study are reflected in current U.S. Department of Health and Human Services (DHHS) treatment guidelines, which state that this regimen should be used "only when a preferred or alternative NNRTI- or PI-based regimen cannot or should not be used as first line therapy." Unfortunately, the ACTG 5095 data do not identify characteristics that would help to predict which patients may be treated successfully with this regimen.

Over the past year, this study has turned a spotlight on the subject of nucleoside-only regimens, as trials of several triple-nucleoside regimens have noted high rates of virologic failure. ACTG 5095 was the first of these to be reported (initially in a NIAID/NIH announcement in the spring of 2003). Subsequently, small noncomparative studies of other triple-nucleoside combinations showed markedly high rates of early virologic failure: 33-52% in 3 studies of abacavir + lamivudine + tenofovir and 91% in 1 study of didanosine + lamivudine + tenofovir.(1-4) These thymidine analogue-sparing regimens appeared to fail earlier and at higher rates than did the abacavir + lamivudine + zidovudine combination studied in ACTG 5095. In them, the rapid emergence of the K65R and M184V mutations appeared to be responsible for the major portion of treatment failure. Interestingly, a 4-nucleoside regimen consisting of abacavir + lamivudine + tenofovir + zidovudine, administered 1 time per day, showed a lower rate of virologic nonresponse at 24 weeks than did the thymidine-sparing combinations, and the resistance profiles of those with virologic failure showed a predominance of M184V and TAMS rather than the K65R.(5)

The differences in failure rates between the combinations that contained zidovudine and those that did not probably are based on differences in the genetic barrier to resistance and on the resistance-modulating effects of zidovudine. Thus, it appears that, while the various nucleoside-only regimens may in general be inferior to those that contain a protease inhibitor (PI) or a nonnucleoside reverse transcriptase inhibitor (NNRTI), they are not equally ineffective. For patients in whom the use of a PI or an NNRTI is not appropriate, it is important to recognize that abacavir + lamivudine + zidovudine may be superior to thymidine-sparing triple-nucleoside combinations, and may have a role in treatment of these select patients.