Lipid changes associated with antiretroviral therapy: results differ by reporting approach.

In order to evaluate consistency and comparability of lipid reporting in antiretroviral treatment (ART) studies, 25 ART trials cited in the Department of Health and Human Services Guidelines were analyzed. Changes in lipids from a separate clinical cohort initiating ART were presented using different approaches to illustrate impact of various methods. Most trials presented lipids as the proportion of graded laboratory toxicities (n=20). Studies of protease inhibitor (PI)-based regimens reported lipids more completely than non-nucleoside reverse transcriptase inhibitor (NNRTI)-based studies. Trials comparing PI-based regimens to NNRTI therapy generally reported graded toxicities or did not include lipids. Within the cohort group, lipid changes appeared lower when presented as median values compared to mean values. Use of grade 3 or higher lipid elevations as the reporting method minimized impact: only 1/64 (2%) reached this threshold. In contrast, reporting outcomes as proportion not at goal per National Cholesterol Education Program (NCEP) Guidelines revealed rates of 23% for triglycerides and 39% for HDL cholesterol. The Framingham risk-prediction model predicted an elevated 10 year risk of cardiovascular disease in 12%. Lipid reporting approaches vary considerably among published ART trials. Implementing a standard approach to reporting lipids including use of NCEP Guidelines and the Framingham risk-prediction model may provide more useful data for clinicians.