575-T.
|
Evolution of Phenotypic Drug Susceptibility and Viral Replication Capacity during Virologic Failure of Combination Antiretroviral Therapy
J. D. Barbour1, 3, T. Wrin2, R. M. Grant1, 3, M. R. Segal1, C. J. Petropoulos2, and S. G. Deeks*1
1Univ. of California, San Francisco; 2ViroLogic, Inc., South San Francisco; and 3Gladstone Inst. of Immunology and Virology, San Francisco, CA
|
Background: Continued use of HAART despite the emergence of drug-resistant HIV has been associated with sustained suppression of viral load below pre-therapy levels. We examined the virologic factors that may account for this persistent benefit.
Methods: This is a prospective observational study of 20 patients who failed to maintain complete viral suppression with a protease inhibitor (PI) based regimen. Longitudinal plasma samples were assayed for genotype, phenotype (fold-change IC50), and replication capacity (measured using a single cycle recombinant virus assay where replicative capacity = patient virus replication divided by reference virus replication).
Results: Data are based on 233 specimens. Patients were observed during continuous virologic failure for a mean of 29.9 months. The mean replicative capacity prior to PI treatment was 0.37 and was correlated with 3TC resistance (rho = -0.59, p = 0.02). Early virologic failure (< 6 months) was associated with emergence of virus with reduced replicative capacity (decrease of 0.19, p = 0.03). At month 6 of virologic failure, the decrease in viral load relative to the pre-therapy baseline was directly associated with the decrease in replicative capacity (0.017 log VL/RC unit, p = 0.04). Long-term virologic failure (> 6 months) was associated with increasing viral load (0.03 log/month, p < 0.0001), increasing PI resistance (0.026 log/month, p<0.0001) and increasing RT resistance (0.015 log/month, p < 0.0001). In a multivariable mixed effects model, PI resistance was the strongest predictor of change in viral load (0.26 log VL/log PI resistance, p = 0.03). There were associations between increasing viral load and decreasing CD4 (-29 cells/log VL, p = 0.005), and between increasing PI resistance and decreasing replicative capacity (-0.46 log/month, p < 0.01). After 48 months of continuous virologic failure, 11/20 had maintained a viral load of at least one log below baseline.
Conclusions: During virologic failure, the level of plasma viremia relative to pre-therapy levels appears to be predicted by the change in replicative capacity from pre-therapy levels. During long-term virologic failure, viral load slowly increased and was associated with increased PI resistance, but not change in replication capacity.
|
