Progression of Subclinical Vascular Damage in People Living With HIV Is Not Predicted by Current Cardiovascular Risk Scores: A Prospective 3-Year Study

Abstract

People living with HIV (PLWH) are at high cardiovascular disease (CVD) risk. Traditional CVD risk scores do not accurately reflect their CVD risk. Noninvasive subclinical vascular damage (SVD) biomarkers are valid surrogates of CVD and able to stratify CVD risk. We tested whether 4 widely applied CVD risk scores [Framingham (FRS), Atherosclerotic CVD, Data Collection on Adverse Effects of Anti-HIV Drugs Study (D:A:D), and Greek-specific European Society of Cardiology (ESC) risk scores] are associated with or detect the presence, incidence, and progression of arteriosclerosis, atheromatosis, and arterial hypertrophy in PLWH and uninfected individuals. We prospectively examined (at baseline and 3-year follow-up) 10 different arterial sites applying 5 different noninvasive vascular biomarkers and measured all 4 CVD risk scores at baseline. In both PLWH (n = 138) and uninfected (n = 664) individuals, the CVD risk scores (except the ESC) performed differently but reasonably well in identifying the presence of SVD, but all scores failed to predict the incidence/progression of overall SVD. The most clinically useful biomarkers (carotid plaque/atheromatosis) revealed that in PLWH, only the FRS was able to stratify the progression (11% of the low-risk, 33.3% of the medium-risk, and 0% of the high-risk group). This extensive vascular phenotyping study demonstrated the clear need to incorporate vascular imaging in CVD risk stratification, in addition to designing more accurate HIV-specific CVD risk models. The use of FRS would further enable treatment optimization and CVD prevention strategies in PLWH at medium CVD risk because one-third of carotid atheromatosis progresses within 3 years.