August 15, 2006
A three-year study of 765 patients has found no advantage in adding a fourth drug to the standard three-drug cocktail for treating newly diagnosed HIV patients.
"Triple drug treatment has been the standard approach to treatment of HIV infection for a decade or so, but there's always been a question about whether we could do better with more drugs," said Dr. Dan Kuritzkes of Harvard's Brigham and Women's Hospital, a co-author of the report. "This reaffirms the potency of the current standard of care."
The results of the study bring new clarity to an old question. Some smaller studies had found better results when more drugs were added to the cocktail, while others saw no benefit.
In the current study, researchers gave two groups of treatment-naïve patients each a regimen containing five pills daily. Neither the patients nor their doctors knew who was actually getting the four-drug regimen. The researchers found the four-drug cocktail, which added abacavir, offered no benefit over the three-drug cocktail either in terms of reduced viral load or increased CD4 cell count. "Over the entire course of the study, at no point did there seem to be an advantage of the four-drug regimen," Kuritzkes said.
The research was supported by the National Institutes of Health and conducted at more than 40 U.S. sites. Almost 20 percent of subjects were female, and more than half were black or Hispanic. Black patients who adhered to their regimen did as well as white patients. However, non-compliant black patients showed a quicker return to high viral levels than non-compliant white patients. Kuritzkes said the reason for this is not clear.
"These results support current guidelines recommending 2 nucleosides plus efavirenz for initial treatment of HIV-1 infection; adding abacavir as a fourth drug provided no additional benefit," the researchers concluded.
The full report, "Three- vs Four-Drug Antiretroviral Regimens for the Initial Treatment of HIV-1 Infection," was published in the Journal of the American Medical Association (2006;296(7):769-781).