Key result
The PCE risk score showed moderate overall discrimination (C-statistic 0.72) but under-predicted cardiovascular events in women and Black/African American men in high-income countries and over-predicted events in low- and middle-income countries.
Why the study?
Does the PCE ASCVD risk score accurately predict cardiovascular events in a global cohort of people with HIV?
Cohort (n=3,893)
Yes
Does the PCE ASCVD risk score accurately predict cardiovascular events in a global cohort of people with HIV?
Effect estimate: C-statistic 0.72 (95% CI 0.68-0.76)
The PCE risk score requires recalibration for people with HIV, as it significantly under-predicts cardiovascular risk in women and Black/African American individuals in high-income countries.
Background: Risk estimation is an essential component of cardiovascular (CV) disease prevention among people with HIV (PWH). We aimed to characterize how well atherosclerotic CV disease (ASCVD) risk scores used in clinical guidelines perform among PWH globally. Methods: In this prospective cohort study leveraging the Randomized Trial to Prevent Vascular Events in HIV (REPRIEVE), we assessed among participants not taking statin therapy the performance of the Pooled Cohorts Equations (PCE) ASCVD risk score and the Data-collection on Adverse Effects of Anti-HIV Drugs (D:A:D) risk score. The C-statistic, the observed:expected (O:E) event ratio, and the Greenwood-Nam-D'Agostino goodness-of-fit (GND) statistic were calculated overall and in subgroups by race, sex-at-birth, and Global Burden of Disease regions (clustering low- and middle-income countries (LMIC) and high-income countries (HIC)). Recalibration was performed for the PCE risk score among PWH in HIC. Findings: Participants (N=3,893) were median age 50 years (IQR 45, 55), with 31% reporting female sex-at-birth. Overall, discrimination of the PCE risk score was moderate (C-statistic [95% CI], 0.72 [0.68, 0.76]) and calibration was good (O:E event ratio 1.11, GND P=0.87). However, calibration suggested over-prediction of risk in LMIC and corresponding under-prediction in HIC. When restricted to HIC, under-prediction (O:E ratio>1.0) was suggested among women (2.39) and Black or African American (Black/AA) participants (1.64). Analogous patterns were noted for the D:A:D risk score. Improved calibration of the PCE risk score in HIC was achieved by multiplying the original score by 2.8 in Black/AA women, 2.6 in women who are not Black/AA, and 1.25 in Black/AA men. Interpretation: Among the global cohort of PWH in REPRIEVE, the PCE risk score under-predicted CV events in women and Black/AA men in HIC and over-predicted CV events in LMIC. Under-prediction in subgroups should be considered when using the PCE risk score to guide statin prescribing for CV prevention among PWH in HIC. Additional research is needed to develop risk scores accurate in predicting ASCVD among PWH in LMIC. Funding: National Institutes of Health, Kowa Pharmaceuticals America, Gilead Sciences, ViiV Healthcare
No takes yet. Share an insight, caveat, or question.
Grinspoon et al. (2025) conducted a cohort in HIV with low-to-moderate traditional ASCVD risk (n=3,893). Pooled Cohort Equations (PCE) ASCVD risk score was evaluated on Discrimination (C-statistic) of the PCE risk score for hard MACE (C-statistic 0.72, 95% CI 0.68-0.76). The PCE risk score showed moderate overall discrimination (C-statistic 0.72) but under-predicted cardiovascular events in women and Black/African American men in high-income countries and over-predicted events in low- and middle-income countries.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: