Introduction: Traditional global risk assessment for cardiovascular disease (CVD) fails to identify a significant percentage of the population at high risk for acute coronary syndrome (ACS). We examined the most important biomarkers and risk factors responsible for identifying patients at discordantly high risk.
Methods: We studied patients without known CVD from a clinical registry who had measures of 5-year risk for CVD using a modified Framingham risk (mFR) score and 5-year risk of ACS using the PULS test (GD Biosciences, Irvine, CA), a clinical algorithm of 9 biomarkers (HGF, sFAS, FasLigand, Eotaxin, CTACK, MCP3, Il-16, HbA1c, HDL-C), in addition to age, sex, diabetes, and family history. Each was divided into low (≤3.5%), intermediate (3.5-7.49%), and high (≥7.5%) 5-year risk. Those who scored low to intermediate on mFR and high risk on PULS were defined as being at discordantly high risk. Stepwise logistic regression with PULS variables, as well as non-PULS biomarkers (ApoA1, ApoB, LDL, Lp(a), triglycerides, SBP, DBP, and hsCRP ) and obesity, BMI, smoking, hypertension medication, and statin medication were included in the model.
Results: Among our sample of 2362 patients (mean age 62 years, 57% male), 30%, 26%, and 45% were classified as low, intermediate, and high risk, respectively, of ACS based on the PULS score. Of 34% of patients (33% of men and 36% of women) classified at low risk and 72% of patients (75% of men and 51% of women) classified as intermediate risk by mFR were classified as high risk by PULS (discordantly high risk patients). Stepwise logistic regression identified 8 PULS variables in the final 14 variable model (Table). HDL-C did not enter into the model.
Conclusions: More than 40% of patients classified as low to intermediate 5-year risk of CVD using mFR are at high risk, with this discordance disproportionately higher in males. Age, Il-16, low FasLigand, and HGF best predicted those at discordantly high risk.