 |
Evolution of drug resistance among patients failing highly active antiretroviral therapy in a supervised treatment setting
Presented by Delivette Castor, United States.
Castor D.1, Hammer S.2, Hoover D.3, Vlahov D.4
1Rockefeller University, Aaron Diamond AIDS Research Center, NY, United States, 2Columbia University, College of Physicians and Surgeons, NY, United States, 3Rutgers, The State University of New Jersey, Statistics, Piscataway, United States, 4New York Academy of Medicine, Center for Urban Epidemiologic Studies, NY, United States
Objective: We examined the predictiveness of resistance testing HAART regimens informed by specific HIV drug resistance (HIVDR) mutations from the IAS-USA panel and HIV-1 variability using greater sequence information among treatment failures in a supervised treatment setting.
Methods: Resistance testing was conducted at baseline on 225 newly admitted patients. Genotypic sensitivity scores (GSS) of the resistance testing informed regimens based on the IAS-USA mutation list was tabulated. HIV-1 variability was defined using amino acid substitutions in the full length of protease and the first 400 codons of reverse transcriptase regions. Baseline viral sequence was compared to HXB2 reference strain, and sequence differences between baseline and treatment failure were calculated using unweighted (total amino acid substitutions) and weighted (protein accepted mutation 250 [PAM250] scores matrix to weight amino acid substitutions) methods. Negative Binomial regression was used to predict evolution rate and linear regression for predicted mean mutations. Classification and regression tree (CART) analyses were used to classify HIVDR mutations predictive of virologic failure and viral evolutionary rate among treatment failures.
Results: Baseline substitutions at positions 82 in protease and 181, 190, 215 and 219 in reverse transcriptase were significantly higher among virologic failures than among non-failures (X2 P <0.05). GSS did not predict viral evolution. After controlling for selected demographic and clinical factors, higher baseline PAM 250 scores indicating less HIV-1 heterogeneity, independently predicted a 1% decrease in viral mutational rate (OR=0.99; 95% CI 0.99, 0.99). A higher number of total amino acid substitutions at baseline was independently associated with a 0.67 (P=0.04) decrease in PAM 250 scores.
Conclusions: Though not associated with treatment failure in this analysis, variability in baseline HIV sequences were associated with higher viral evolution rate among treatment failures. Future studies are needed to better characterize the clinical and pathogenic significance HIVDR.
Back to the session -
Back to the Programme-at-a-Glance
|
|