Abstract

E-mail Abstract Author Add To Itinerary Session Search Abstracts Program

Session 24 Oral Abstract Session Antiretroviral Chemotherapy: Pathogenesis of Primary HIV Infection Session Time: Wednesday, 10 am - 12:30 pm Room 6A-B

10:45   94. Viral Blip Dynamics during HAART M. Di Mascio*1, M. Markowitz2, M. Louie2, A. Hurley2, D. Ho2, and A. S. Perelson1 1Los Alamos Natl. Lab., NM and 2Aaron Diamond AIDS Res. Ctr., Rockefeller Univ., New York, NY Background: Intermittent episodes of low-level viremia (blips) are often observed in well-suppressed, HAART-treated patients. It has been reported that viral blips do not correlate with the emergence of new HAART-related mutations; however, increased frequency of blips correlates with slower decay of latently infected cells. Since blips are transient and unpredictable, detailed knowledge about them is difficult to obtain. We present an analysis of the dynamics of viral blips from viral load (VL) measurements on 123 patients for a period of 809±480 days (21-1817 days) and sampled every 31±12 days for a total of 26±15 samples per patient. Methods: 123 patients naïve to HAART and treated with 8 different potent regimes were analyzed. A viral blip is defined as any VL > 50 and < 1000 copies/mL during the period of suppression. Results: 77.5% of patients showed intermittent viral blips. The mean (median) frequency of such blips was 0.09/sample (0.06/sample). There was no evidence of an increase in blip frequency with longer periods of observation. Bootstrap-based comparisons of the distribution of the number of consecutive blips, as well as of clusters of blips, show that viral blips are generated substantially independently from each other (i.e., appear randomly). Thus, 2 observed consecutive blips are likely to belong to 2 different independent episodes of incomplete viral suppression. The mean (median) amplitude of viral blips was 165 (110) copies/mL. The distribution of blip amplitudes appears to decrease exponentially, so most blips have low amplitude and very few have high amplitude. No evidence was observed of an increase in blip amplitude for longer periods of observation. The observed distribution of blip amplitudes can be reproduced by a model that generates viral blips randomly, and that assumes a viral blip rises rapidly and then decreases with exponential kinetics. Conclusions: These observations suggest that viral blips are independent of the duration of HAART; thus, decrease adherence to therapy with time does not appear to drive the occurrence of blips. Instead, blips appear to reflect a random process. It may be caused by unexplained decreases in drug concentration or unrecognized increases in T-cell activation. As for the latter, one is tempted to speculate that an intermittent viremic episode might reflect a random, massive release of virions from one compartment, followed by transport to another where they are more quickly cleared.

©2002 9th Conference on Retroviruses and Opportunistic Infections