Viral load at 6 months correlates with baseline CD4 but not baseline viral load Gail Skowron reported on an analysis of the

Viral load at 6 months correlates with baseline CD4 but not baseline viral load

Gail Skowron reported on an analysis of the predictability of baseline CD4 and viral load for treatment outcome (viral load at 6 months below 200/400/500 copies/ml). Skowron conducted a meta-analysis of 26 treatment arms in 23 studies for a total of 1,469 patients. The studies were of HAART regimens of a potent NNRTI or PI in primarily naļve populations. She did a sub-analysis of data for indinavir and nevirapine based regimens. The studies included 10 with nevirapine, 2 with efavirenz, 10 with indinavir, and 4 with nelfinavir. The total number of patients on treatment at 6 months was 1,022 (39 per study). These studies included: AVANTI 2 & 3, Merck 035 and 020, EFV 006, CHEESE-1, START 1, Atlantic, EFV 043, VIRGO, INCAS, and others.

Some of the study limitations Skowron mentioned were—

No adjustment was made for the number of patients in each study. The impact of a small study of say 15 individuals is the same as that for a study with 150 individuals.
Variables such as adherence and disease stage were not included in analysis
ITT analysis would have been preferable but were not available for most trials
Mean and median baseline values for CD4 and viral load were used in this analysis. Although the mean or median does represent accurately the average for the study there were individuals in each study whose baseline CD4 or viral load were not accurately represented by these values. The results of this meta-analysis may be biased by high or low baseline values

Results

Skowron reported that in this analysis baseline CD4 was more predictive of outcome (significant correlation) of HAART (below 200/400/500 copies/ml) at 6 months than was baseline viral load. She said, patient groups with lower CD4s were, therefore, less likely to have undetectable viral loads at 6 months than those with higher CD4s.
There was no significant difference in outcome for patient groups with high or low baseline plasma viral load. There was no correlation between mean/median baseline viral load and outcome at 6 months.
She suggests that the poor response of patient groups with high viral loads seen in some studies may be related to low CD4s. In addition, she suggests comparisons of studies with similar viral loads but widely disparate baseline CD4s are not appropriate.
The sub-analyses of NVP and IDV based HAART regimens also showed a significant correlation between CD4 and virological outcome.