Carotid atherosclerosis in virologically suppressed HIV patients: comparison with a healthy sample and prediction by cardiovascular risk equations

OBJECTIVES

To compare the prevalence of carotid atherosclerosis in virologically suppressed HIV patients with that of a community sample, and to evaluate the capacity of various cardiovascular risk (CVR) equations for predicting carotid atherosclerosis.

METHODS

This was a cross-sectional study with two randomly selected groups: HIV patients from an HIV unit and a control group drawn from the community. Participants were matched by age (30-80 years) and sex without history of cardiovascular disease. Carotid plaque, common carotid intima-media thickness (cc-IMT) and subclinical atherosclerosis (carotid plaque and/or cc-IMT > 75^th percentile) were assessed by carotid ultrasound. The Systematic Coronary Risk Evaluation (SCORE), Framingham, REGICOR, reduced Data Collection on Adverse Effects of Anti-HIV Drugs (D:A:D), and COMVIH equations were applied, and their abilities to predict carotid plaque were compared using the area under the curve (AUC).

RESULTS

Each group included 379 subjects (77.8% men, age 49.7 years). Duration of antiretroviral therapy was 15.5 years. There were no differences between the groups for carotid plaque (HIV, 33.2%; control, 31.3%), mean cc-IMT (HIV, 0.63 mm; control, 0.61 mm) or subclinical atherosclerosis (HIV, 42.9%; control, 47.9%). Thymidine analogues were independently associated with subclinical atherosclerosis in HIV-infected patients. CVR equations revealed AUCs between 0.715 and 0.807 for prediction of carotid plaque; prediction was better in the control group and did not improve when HIV-adapted scales were used.

CONCLUSIONS

The features of carotid atherosclerosis did not differ between the HIV-infected and the control group, although CVR equations were more predictive for carotid plaque in controls than in HIV-infected patients. HIV-specific equations did not improve prediction.