Abstract Body

Background: There is increasing evidence that cardiovascular disease (CVD) is more common among persons living with HIV (PLWH) than among HIV-negative individuals. However, most previous studies have been limited by small sample size, lack of well-defined clinical events, incomplete case ascertainment, and the use of composite endpoints. We examined which traditional CVD risk factors and HIV-specific factors predict myocardial infarction (MI) in PLWH.

Methodology: CNICS is a multi-center cohort comprised of 8 clinical sites contributing comprehensive clinical data on >27,000 HIV patients followed during routine care. We developed a state-of-the-art screening algorithm and central adjudication protocol for the validation of incident MIs. Among 18,115 eligible participants, we found 1360 individuals with potential incident MIs, of which 314 were confirmed as incident MIs. We used logistic regression to estimate risk of MI associated with baseline Framingham risk score (FRS) covariates (total and HDL cholesterol, systolic blood pressure (SBP), age, sex, diabetes, and smoking) and HIV factors (current CD4, HIV viral load, and transmission risk of injection drug use).

Results: In multivariate analysis the Framingham covariates smoking, age, total and HDL cholesterol, and SBP were significant predictors of incident MI. Diabetes and IDU were not significant. After adjusting for traditional CVD risk factors, each 300 cell/mm3 increase in current CD4 count was associated with a 25% decrease in the risk of MI, and higher HIV viral load was associated with a significant increase in MI risk. Median time between last measured CD4 count and end of follow-up due to event or administrative censoring was 84 days (IQR: 30-172 days).

Conclusions: While FRS covariates remain important, HIV-specific factors are also significantly associated with incident MI in PLWH. Higher prevalence of IDU among females likely confounds the effect of male sex. After adjusting for traditional CVD risk factors, the most recent CD4 count and HIV viral load prior to the event were independent predictors of MI incidence. Our findings highlight the importance of early and effective treatment for PLWH and provide additional evidence, based on a rigorously adjudicated clinical endpoint (MI), that immunodeficiency and active viral replication may be associated with the excess CVD risk in PLWH.