Pharma's PR machine scored a coup in its slant on the news out of EuroSIDA, the 5,000-strong pan-European observational cohort of HIVers "Drug Stoppers Six Times As Likely to Develop AIDS or Die," trumpeted the Los Angeles Times about the data's debut in March. The Washington Post ("Studies Find Danger in Waiting, Interrupting") fanned the flames: "Even if they [later] restarted treatment," David Brown wrote, "they continued to have nearly twice the risk of death." But readers able to muster the fortitude to read on discovered that this frightening statistic is true only for those few HAART starters with CD4s less than 200.

London doc Ian Williams' take was that the risk of HIV progression in those with 250 or more CD4s who interrupted therapy for three months or more is "practically zero." But for interrupters with CD4s below 50, those "drug failures" who stayed on therapy fared much better than those who went off. "There must be [as yet unidentified] mechanisms [causing] this effect, which confer the clinical benefit," Williams said.

Break it down? Even though the drugs are not "working" -- that is, suppressing virus and elevating CD4s -- they are still forcing mutations on the virus, weakening it and making it less lethal than it would be in its natural HAART-less state.