A-201 Deriving Information from the Lipid Panel as Spherical Coordinates

Abstract

Abstract Background Lipid panel assays are used to calculate LDL-C as a proxy for cardiovascular disease risk. Further investigations are required to classify the patient’s dyslipidemia into a phenotype to assist with management. We explored the information contained within the lipid panel with two aims: First, to estimate cardiovascular risk directly, bypassing LDL-C calculation. Second, to provide a novel phenotypic classification system. Methods Various transformations of lipid panel data were explored as inputs to a logistic regression model of cardiovascular risk, and to produce a spherical coordinate system for phenotyping. Area under the receiver operating characteristic (AUROC) curve was used to assess the predictive ability of the model. Clinical utility metrics were calculated and Kaplan-Meier survival curves were used to determine the prognostic value of the model. Results Transforming the lipid panel data into indices allowed derivation of a spherical coordinate system. The spherical coordinates proved to be the best input values for logistic regression to provide the probability of high cardiovascular risk. When we added age, sex and race as variables in the model, the AUROC for predicting cardiovascular events approached that of the American Heart Association’s pooled cohort equation (PCE). The final model was more sensitive than the PCE, and provided greater prognostic ability than LDL-C. The spherical coordinate system classified individuals into “normal” or one of eight phenotypes, which were differentially associated with clinical features such as elevated CRP or metabolic syndrome. Conclusion The new model uses all three lipid panel parameters and other routinely available laboratory data to directly predict cardiovascular risk with similar accuracy to the PCE. Conversion of the parameters to a spherical coordinate system allows classification into one of 9 phenotypic groups, which may alert clinicians to underlying or consequent pathological processes to be managed.