Brian Hoyle

March 02, 2011

March 2, 2011 (Boston, Massachusetts) — Starting antiretroviral therapy (ART) within 2 weeks of commencing treatment for tuberculosis (TB) is dramatically effective, as opposed to starting ART 8 to 12 weeks after commencing TB therapy, according to the results of a large international trial presented here at the 18th Conference on Retroviruses and Opportunistic Infections.

"In patients with a CD4 count of less than 50 cells/mm3, the proportion of patients who had AIDS-defining conditions decreased from 27% to 16% over 48 weeks [with anti-TB therapy plus immediate ART]. This is really a dramatic reduction. We absolutely need to act on these data," Diane V. Havlir, MD, chief of the HIV/AIDS division at the University of California at San Francisco, told Medscape Medical News.

The call for action is based on findings from a study by the AIDS Clinical Trial Group. The 48-week phase 4 open-label randomized study involved 806 people from 26 sites on 4 continents, and was undertaken to optimally determine when to start ART.

"When should we start ART? This question is complex and is compounded because diagnostics can be limited, so clinicians can be sitting with their patients wondering, 'Should I start therapy or should I not?' in the absence of a diagnosis," Dr. Havlir explained.

Participants were HIV positive with a CD4 count below 250 cells/mm3, and were being treated for confirmed or suspected TB. ART with a standard rifampicin-containing country-approved regimen (involving trimethoprim/sulfamethoxazole prophylaxis and tenofovir/emtricitabine plus efavirenz) was started in 405 patients within 2 weeks of commencing TB treatment (median start time, 10 days) and in 401 patients 8 to 12 weeks after beginning TB therapy (median start time, 10 weeks).

The primary end points at 48 weeks were the proportion of patients who had died or who had AIDS-defining conditions. Treatment safety, TB outcome, and the occurrence of immune reconstitution inflammatory syndrome (IRIS) were also considered.

When therapy was started, TB had been confirmed in 46% of the patients.

Overall, AIDS or death occurred in 13.0% and 16.1% of the patients in the immediate and early treatment groups, respectively (confidence interval [CI], –1.8 to 8.1; = .45). Of the 521 patients whose CD4 counts exceeded 50 cells/mm3, AIDS or death occurred in 11.5% of the immediate group and 10.5% of those early group (CI, –6.7 to 4.3; = .67).

But, of the 285 patients whose CD4 count was 50 cells/mm3 or less, AIDS or death occurred in 26.6% of the early group and a far lower 15.5% of the immediate group (CI 1.5–20.5; = .02), Dr. Havlir reported.

TB IRIS was significantly greater in the 43 patients receiving immediate ART (11%) than in the 19 patients receiving early ART (5%; = .29). However, the events tended to occur within the first 2 weeks of ART and were relatively minor and treatable.

Both groups were similar in grade 3 or 4 toxicities, the suppression of viral RNA copy number at the end of the study (74% of patients), and the increase in CD4 count to 100 cells/mm3 or higher (60%).

These findings, according to Dr. Havlir, build on previously reported findings that collectively argue for the commencement of ART as early as possible in patients coinfected with HIV and TB.

"I think this is why we do clinical trials. There have now been 3 trials that have come up with the same thing. I think that these [collective] results should be a practice changer," Dr. Havlir told Medscape Medical News.

Although praising the study, William J. Burman, MD, from the Denver Health Hospital in Colorado, was concerned that, of the 21 deaths that occurred, 14 were in the immediate group, "possibly a manifestation of fatal IRIS." Although the data to this point are inconclusive, "given the strategy design of the study," indications are that the 14 deaths were due to a progression of TB that might well have occurred in the absence of IRIS, according to Dr. Havlir.

In addition, information about TB drug resistance in the patients is incomplete.

"In the CAMELIA and SAPIT studies, much of the benefit was seen beyond 48 weeks. While the data from this study look great, I'd be really interested in seeing data on patients after 48 weeks as well," said Richard E. Chaisson, MD, from Johns Hopkins University and director of the John Hopkins Center for Tuberculosis Research in Baltimore, Maryland.

Dr. Havlir reports a financial relationship with Abbott Laboratories. Dr. Burman and Dr. Chaisson have disclosed no relevant financial relationships.

18th Conference on Retroviruses and Opportunistic Infections (CROI): Abstract 38. Presented February 28, 2011.