428   Incidence and Predictors of Clinical Progression Among HIV-Infected Patients Experiencing Virologic Failure of Protease Inhibitor-Based Regimens.

S. Deeks*, J. Barbour, R. Grant, and J. Martin.
San Francisco Gen. Hosp., Univ. of California San Francisco, and Gladstone Inst. of Virology and Immunology, San Francisco, CA.

Background:Virologic failure of a protease inhibitor-based regimen is often associated with a durable CD4 T cell benefit, but the duration and predictors of clinical benefit have not been well-defined.

Methods:This is a clinic-based observational study of HIV-infected patients who initiated a protease inhibitor- based regimen prior to December, 1997. Virologic failure was defined as two consecutive plasma HIV RNA levels of >500 copies/mL. Clinical progression was defined as a new AIDS-defining clinical event or death. Patients experiencing virologic failure were censored at the time they switched to a successful salvage regimen. Patients discontinuing therapy are included in the analysis. Factors associated with clinical progression were determined using proportional hazards regression with time-dependent covariates.

Results:Of the 485 patients studied, 302 patients entered a state of confirmed virologic failure. Prior to the initiation of a protease inhibitor, the median CD4 cell count of those who developed virologic failure was 124 cells/mm³and the median HIV RNA was 4.81 log copies/mL. After the onset of virologic failure, there were 55 events in 567.3 person-years of observation. The probability of clinical progression was 18% at 2 years and 41% at 4 years. Clinical progression was independently associated with return of plasma HIV RNA levels to within 0.4 log¹⁰of the pre-therapy "set point" (RR = 3.8, p = 0.07) and with a low pre-therapy CD4 nadir (RR = 4, p = 0.001) but not the absolute level of viremia achieved. Return of viral load to pre-therapy levels was due both to treatment failure and to treatment interruption.

Conclusions:After 4 years of continuous virologic failure, clinical progression appears to be common. Disease progression was predicted by return of HIV RNA to "set-point" levels and by nadir CD4. These data suggest that maintenance of a virologically failing regimen is a clinically viable strategy as long as partial suppression can be maintained. Novel therapeutic agents and/or strategies are needed for those unable to achieve this goal.