• Incidence and Predictors of Clinical Progression Among HIV-Infected Patients Experiencing Virologic Failure of Protease Inhibitor-Based Regimens (Poster 428)
  Authored by S. Deeks, J. Barbour, R. Grant, and J. Martin
  View the original abstract

Despite the progress of the treatment of HIV infection, we are still encountering treatment failures. These "failures" can be defined in different ways, beginning with either an intolerance to HIV medications or the development of widespread resistance and the subsequent failure of the medications to abolish viral replication. Deeks in earlier work described how the common clinical picture of patients with persistent viral replication, intermediate viral loads ranging as high as 70,000 copies though the upper limit, and stable CD4 counts is not thoroughly understood. Rather, there was no obvious immunologic decline despite viral replication. This became known as a "disconnect" or "discordancy" between the viral control and immune decline. There are numerous theories for this discordancy ranging from viral mutations that are resistant to our medications to different viruses with different shapes and activity that are less antagonistic or virulent to our immune system. Chaisson et al. (abstract 429) reported from their cohort in Baltimore that continuing treatment despite virologic failure resulted in higher CD4 counts than in patients who discontinued therapy. We are creating an unsteady truce between viral suppression, the selection of different viruses, and immunologic control.

This trial was a retrospective observational study of 485 patients at San Francisco General Hospital, 302 of which had confirmed virologic failure. These patients had median CD4 counts of 124 and median HIV RNA of 4.81 log (approximately 75,000 copies/mL) before the initiation of a protease inhibitor. At the onset of subsequent virologic failure there was a probability of clinical progression of 18% at two years and 41% at four years. This progression was independently associated with a return of viral load to the vicinity of the original viral set point before the initiation of therapy and with a low CD4 nadir prior to therapy as well. Apparently, the progression may have been related to the strength of patients' original responses as measured by the level of viral load or CD4 count at the time of therapy initiation, and interestingly, was not related to the highest point of viral load achieved.

Deeks et al. concluded that clinical progression is unfortunately common after four years of continuous virologic failure and could be predicted by the original set point of virologic control and the low point of immune decline as measured by the nadir of the CD4 count. However, 59% of the patients had still not progressed after four years of clinical failure, giving hope to the notion that maintaining a failing regimen as reflected in persistent viral replication may be a viable strategy -- as long as some measure of virological suppression can be maintained. How much suppression is enough has not been clearly determined; this will drive future development of novel strategies for these long-term, slowly progressing viral failures. We may be creating persistently replicating but less-virulent viral strains. It is hopeful to know that we are buying time, but it is not clear how much time we are buying. There is hope that a combination of new approaches involving treatment interruptions (see abstract 292) and new therapies could create new options that would result in greater clinical stability.