Abacavir May Be Less Effective in HIV Patients With High Viral Load

Bob Roehr

August 13, 2008

August 13, 2008 (Mexico City) — The controversy about whether use of abacavir (ABC) increases the risk for virologic failure in patients with a high viral load was given center stage in back-to-back late-breaker presentations here at the XVII International AIDS Conference.

Paul Sax, MD, from the Brigham and Women's Hospital, Boston, Massachusetts, presented interim findings from part of the AIDS Clinical Trials Group (ACTG) 5202 trial. The 96-week phase 3B study randomly assigned treatment-naive patients to double-blinded regimens of ABC/lamivudine (3TC) or tenofovir (TDF)/emtricitabine (FTC), with open-label use of efavirenz or atazanavir/ritonavir, and stratified them at entry by HIV RNA levels either lower or higher than 100,000 copies/mL.

In January 2008, the data safety monitoring board (DSMB), during a scheduled review of data, identified "highly significant difference between ABC/3TC and TDF/FTC, with excess virologic failures on the ABC/3TC arm occurring in the >100,000 stratum," Dr. Sax said. The ACTG followed the DSMB recommendation and unblinded the stratum with the higher viral load. Dr. Sax presented an analysis of that group.

This portion of the study included 797 study patients (of a total of 1858 patients enrolled) with a median follow-up of 60 weeks. Baseline genotyping was performed on 43% of patients, and if relevant point mutations were present, the patients were excluded from the study.

Virologic failure was more common in the ABC/3TC group than in the TDF/FTC group (57 vs 26 events; P = .0003; hazard ratio for failure of 2.33). Dr. Sax said, "There is a consistent excess of failures in the ABC/3TC arm regardless of the criteria you use for virologic failure"; either early failure to suppress viral load below 200 copies/mL or later failure in rebounding above that level.

Combining time to virologic failure and time to modification of the nucleoside component of the regimen, the researchers found 114 and 68 events, respectively (P = .0001).

Suspected drug hypersensitivity events occurred in 27 patients (7%) in each group. "Screening for HLA 5701 was only done in a small fraction of the patients in the study, and generally not until the end of it," Dr. Sax said.

One death likely was a result of a hypersensitivity reaction on reinitiation of ABC, despite sites being "extensively educated" about recognition and management of that risk.

Grade 3 and 4 adverse events occurred in 33% and 19% of the ABC/3TC and TDF/FTC groups, respectively. Changes in lipid levels were similar in both groups and were probably not clinically significant.

Dr. Sax concluded, "ABC/3TC was associated with a significantly shorter time to virologic failure...and a shorter time to the first adverse event" in the group of patients with HIV RNA levels higher than 100,000 copies/mL.

These findings differ from what has been seen in earlier clinical trials. Dr. Sax suggested that the larger sample size in ACTG 5202 "may have allowed us to describe differences between the groups with greater precision than with a smaller sample size." The use of a different comparator regimen might also help to explain the differences.

During discussion, Dr. Sax acknowledged that different rates of adherence might have affected outcomes in this study; adherence has not yet been analyzed.

GlaxoSmithKline's Analysis

"The ACTG 5202 endpoints are unique, [and] the interim results are unexpected and inconsistent with earlier data," said Keith Pappa, MD, who presented GlaxoSmithKline's review of 6 recent trials of abacavir involving 2940 treatment-nave patients, using the ACTG 5202 endpoints. "The range of response is between 87% and 95%," and there was no significant difference by baseline viral load, he said.

Dr. Pappa focused on the HEAT trial, which included 688 patients in a 96-weekcomparison of the same nucleoside regimens as the ACTG study. At 48 and 96 weeks, stratifying patients by viral loads above and below HIV RNA levels of 100,000 copies/mL, there were no statistically significant differences between the groups in terms of suppressing HIV RNA to levels lower than 50 copies/mL. Nor were there significant differences in terms of adverse events.

Dr. Pappa concluded, "The ABC/3TC regimen demonstrates [that] robust results, irrespective of baseline viral loads," are safe and well tolerated.

University of North Carolina, Chapel Hill, researcher Joseph Eron, MD, took the company to task for insinuating that the ACTG study, as the outlier, was any less valid, saying, "that is a bit disingenuous."

The 6 studies the company presented showed ABC/3TC to be marginally better in patients with lower rather than higher viral loads, whereas the comparator drugs were the reverse. Dr. Eron asked, "If you did cumulative analysis across all 6 studies, would you see a significant difference" between the higher and lower viral load groups?

Dr. Pappa replied that it is difficult to combine trials that had different entry criteria and used different regimens. "What we have to look at is the consistency of response and the percentage point difference."

Cornell University researcher Roy Gulick, MD, raised the point of clinical trials being "generalizable to the patients we take care of." He asked about the sex and racial/ethnic breakdown of participants in the company trials and whether the researchers has conducted an analysis of the HEAT trial along those lines. The company has not done so., Dr. Pappa said.

Dr. Gulick later acknowledged to Medscape HIV/AIDS that as yet there is no analysis of the ACTG 5202 data by race or sex. Still, "It is remarkable that more than half of [the study participants] were people of color, and we have to generalize those results to the population we're serving," he said.

"Race and ethnicity may be a marker for other things," such as diabetes, hypertension, and higher body mass index. "Other studies have shown that black race may be correlated with not as good virologic responses," Dr. Gulick said.


"I take the findings in the [ACTG] study as serious. They represent a concern, and we need to be vigilant," Edwin De Jesus, MD, an HIV physician with a private practice in Orlando, Florida, told Medscape HIV/AIDS. He cautioned that only 1 group of the study has been stopped, whereas the others continue on toward their 96-week conclusion. The full data have not been unblinded, so the possible contribution of other drugs in the regimens is not known.

"The fact that some of the patients were not screened for hypersensitivity reactions, for HLA 5701," may also play a role, he said, noting that he looks forward to seeing further data from the study.

Bob Huff, from the AIDS think tank Treatment Action Group (TAG), thought it was important to remember that the DSMB did not stop the lower viral load strata of the trial. That presumes it did not have a similar concern with excess negative events in that patient population. Perhaps, therefore, events in the high viral load group were a fluke, he said.

Mr. Huff was shocked to hear that a patient in the ACTG trial had died of a hypersensitivity reaction on restarting abacavir treatment. He told Medscape HIV/AIDS, "If it is happening among that group of highly sophisticated clinicians, then it is happening elsewhere."

He speculated that the greatest risk for mortality in using ABC may not be enhanced cardiovascular risk but, rather, the hypersensitivity reaction to the drug, even though use of the HLA screening test for sensitivity should be standard of care.

The ACTG 5202 trial was funded by the National Institutes of Health. GlaxoSmithKline supported an internal review of data. Dr. Pappa is an employee of GlaxoSmithKline. Dr. Sax, Dr. Gulick, Dr. Eron, Dr. De Jesus, and Mr. Huff have disclosed no relevant financial relationships.

XVII International AIDS Conference: Abstracts THAB0303 and THAB0304. Presented August 7, 2008.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.