Correlates of Time-to-Loss-of-Viral-Response in CSF and Plasma in the CHARTER Cohort
Scott Letendre, MD
17th Conference on Retroviruses and Opportunistic Infections (CROI, February 2010)
(this will launch CROI 2010 site)
Scott Letendre1, R Ellis1, R Deutsch1, D Clifford2, C Marra3, A McCutchan1, S Morgello4, D Simpson4, R Heaton1, I Grant1, and the CHARTER Group
1Univ of California, San Diego, US; 2Washington Univ in St Louis, MO, US; 3Univ of Washington, Seattle, US; and 4Mt Sinai Sch of Med, New York, NY, US
Background: No studies have determined the correlates of the time-to-loss-of-HIV suppression in CSF. To address this, we determined the time-to-loss-of HIV suppression in a large, longitudinal cohort and compared the results in CSF and plasma.
Methods: Longitudinal analysis of HIV+ pts of CHARTER, a 6-center, US-based cohort, who were taking antiretroviral therapy (ART), were suppressed below the lower limit of quantitation (LLQ, 50 copies/mL) in CSF (n = 346) or plasma (n = 225) at their initial visit, and had at least 2 semiannual visits. Kaplan-Meier estimates of time-to-loss-of-viral response (TLVR), defined as a VL >50 copies/mL, were calculated for dichotomized demographic, clinical, and neuropsychological (NP) variables. Baseline variables that were significant at the 10% significance level (Wilcoxon) were analyzed by multivariable Cox proportional hazards models and likelihood ratio tests.
Results: Median duration of ART at the initial visit was 11.1 months in the CSF analysis and 13.0 months in the plasma analysis. In the CSF analysis, 67 events (19%) occurred over a median 9.3 months and 279 (81%) maintained suppression over a median 20.3 months. Baseline predictors of shorter TLVR in CSF were detectable HIV VL in plasma, CD4+ counts <200, age ≤43 years, black ethnicity, protease inhibitor use, and absence of antidepressant use. In addition, patients who were NP impaired and reported non-adherence had shorter TLVR in CSF. Multivariable analysis confirmed that each variable - other than plasma VL and antidepressant use - contributed to the duration of TLVR (model P <0.0001). In the plasma analysis, 82 events (36%) occurred over a median of 9.3 months and 143 (64%) maintained suppression over a median of 18.0 months. Baseline predictors of TLVR in plasma were CD4+ counts < 200, age ≤43 years, black ethnicity, and global NP impairment. Multivariable analysis confirmed that each of these variables were associated with shorter TLVR in plasma (model P <0.0001).
Conclusions: A substantial minority lost their viral response within a year of entering this cohort study. Predictors of shorter TLVR were similar in CSF and blood except that use of protease inhibitors and non-adherence were only associated with shorter TLVR in CSF. Although this cohort-based analysis has an important limitation – initiation of ART was not observed as in a clinical trial – the findings identify that NP impairment is associated with shorter viral responses in CSF and plasma.
