gms | German Medical Science

Background: Traditional and HIV-specific cardiovascular risk factors are associated with cardiovascular events (CVE) in people living with HIV (PLH). We investigated the impact of these risk factors on the risk estimation of incident events comparing two endpoint definitions.

Methods: The HIV HEART Aging study (HIVH) is an ongoing prospective observational cohort study to assess the risk of CVE in PLH in the German Ruhr Area since 2004. We included male and female PLH with at least 5 years of follow-up. CD4 cell count, AIDS, viral load, and HIV diagnosis before 1996 were identified as HIV-specific cardiovascular risk factors. Traditional cardiovascular risk factors used in the analysis were smoking, diabetes mellitus, systolic blood pressure, cholesterol, and body mass index (BMI). We used two endpoint definitions for CVE:

1.

a more extensive cardiovascular endpoint (ECE) definition, including all cardiovascular events (myocardial infarction, coronary revasculation, bypass operation, heart valve operation, reanimation/defibrillation, pacemaker, stroke) and cardiac death;

2.

an endpoint definition of hard cardiac events (HCE), including only cardiac death and myocardial infarction.

For the longitudinal analysis, we used the time-dependent proportional hazards model using inverse probability-of-censoring weights (IPCW) with a cumulative/dynamic receiver operating characteristic (ROC) curve approach to illustrate the influence of the baseline parameters on the endpoint occurrence after 5 years. All results were adjusted for age and sex.

Results: For the analysis of the ECE 1258 patients (m 1058, f 205) with 122 CVE, and accordingly for the HCE 1292 patients (m 1081, f 211) with 63 events, were reviewed. Both analysis groups showed a mean age of 43.5±10.1 years and 29% of clinical HIV stage C/AIDS with a mean duration of 7±6 years since the first diagnosis of HIV infection. For the more extensive ECE, traditional cardiovascular risk factors predicted incident CVE better than using only HIV-specific cardiovascular risk factors, with an AUC of 0.84 (95%Cl 0.76-0.91) compared to 0.78 (95%Cl 0.68-0.88). The combination of traditional and HIV-specific cardiovascular risk factors showed no improvement in risk prediction with an AUC of 0.85 (95%Cl 0.79-0.91). For the HCE, traditional cardiovascular risk factors predicted incident CVE only slightly better than HIV-specific cardiovascular risk factors, with an AUC of 0.80 (95%Cl 0.74-0.86) compared to 0.77 (95%Cl 0.71-0.83). Here, combining the traditional and HIV-specific cardiovascular risk factors resulted in a slightly better prediction compared to the traditional risk factors with an AUC of 0.83 (95%CI 0.73-0.92).

Conclusion: Combining traditional and HIV-specific cardiovascular risk factors may improve the prediction of incident CVE. This improvement seems to be related to the definition of the endpoint used in the analysis. Selecting a more extensive endpoint led to no improvement in the study presented here. There is an unmet medical need to develop a CVE risk prediction model for PLH that includes HIV-specific cardiovascular risk factors, but the endpoint selection has to be chosen by care.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.