Predicting Cardiovascular Events in Familial Hypercholesterolemia: The SAFEHEART Registry (Spanish Familial Hypercholesterolemia Cohort Study).

Abstract

Although risk factors for atherosclerotic cardiovascular disease (ASCVD) in familial hypercholesterolemia (FH) have been described, models for predicting incident ASCVD have not been reported. Our aim was to use the SAFEHEART registry (Spanish Familial Hypercholesterolemia Cohort Study) to define key risk factors for predicting incident ASCVD in patients with FH. SAFEHEART is a multicenter, nationwide, long-term prospective cohort study of a molecularly defined population with FH with or without previous ASCVD. Analyses to define risk factors and to build a risk prediction equation were developed, and the risk prediction equation was tested for its ability to discriminate patients who experience incident ASCVD from those who did not over time. We recruited 2404 adult patients with FH who were followed up for a mean of 5.5 years (SD, 3.2 years), during which 12 (0.5%) and 122 (5.1%) suffered fatal and nonfatal incident ASCVD, respectively. Age, male sex, history of previous ASCVD, high blood pressure, increased body mass index, active smoking, and low-density lipoprotein cholesterol and lipoprotein(a) levels were independent predictors of incident ASCVD from which a risk equation with a Harrell C index of 0.85 was derived. The bootstrap resampling (100 randomized samples) of the original set for internal validation showed a degree of overoptimism of 0.003. Individual risk was estimated for each person without an established diagnosis of ASCVD before enrollment in the registry by use of the SAFEHEART risk equation, the modified Framingham risk equation, and the American College of Cardiology/American Heart Association ASCVD Pooled Cohort Risk Equations. The Harrell C index for these models was 0.81, 0.78, and 0.8, respectively, and differences between the SAFEHEART risk equation and the other 2 were significant (P=0.023 and P=0.045). The risk of incident ASCVD may be estimated in patients with FH with simple clinical predictors. This finding may improve risk stratification and could be used to guide therapy in patients with FH.