When researcher David Ho, MD, unleashed a media frenzy in February by announcing that a gay New York City man had contracted a multi-drug resistant, fast-progressing HIV strain (see “Big, Bad Media Bugout"), few doubted that he'd grab headlines two weeks later in Boston at CROI, the year's top AIDS confab. The hoopla unfortunately overshadowed “the best [CROI] in years in terms of therapy progress,” says Project Inform's Marty Delaney. Check these future features:

PROTEASE PROMISE  In Phase III trials, long-anticipated protease inhibitor (PI) tipranavir—the FDA will likely approve it this summer—performed better in protease-resistant HIVers than PI rivals like Kaletra. Tibotec's PI contender TMC-114 may prove even more promising. In a large, six-month Phase II study of multidrug-resistant HIVers, TMC-114 combos smashed viral load on average by nearly two logs (say, from 60,000 to 600)—even to undetectable at the highest dose.

NON-NUKE NOTES  Over a year, adding Pfizer's non-nuke capravirine to a PI-plus-two-nukes combo didn't suppress non-nuke-resistant HIVers' viral loads better than the combo without capravirine. But Tibotec scored again with non-nuke TMC-278, plunging viral loads in seven days. Larger TMC-278 trials start this spring.

NEW CLASSES  Drugs that stop HIV in novel ways are a-brewin'. For nine days, a single dose of Panacos' maturation inhibitor PA-457—which blocks the same HIV step as PIs, but differently—dropped viral loads significantly. And in a tiny trial, Merck's integrase inhibitor L-870810—which keeps HIV from stuffing its own DNA into human DNA—cut HIVers' viral loads over 10 days of twice-daily dosing. Too bad toxicity in dog studies has shelved it, though Merck plans to proceed with a similar compound. Woof!