A simplified diagnosis algorithm for dysbetalipoproteinemia

Background Dysbetalipoproteinemia (DBL) is a disease of remnant lipoprotein accumulation caused by a defective apolipoprotein (apo) E and is associated with a considerable atherogenic burden. However, there exists confusion concerning the diagnosis of this disorder, and as a consequence, misdiagnosis is frequent. Objective The objective of the present study is to propose an algorithm for the diagnosis of DBL using simple clinical variables. Methods In a large cohort of 12,434 dyslipidemic patients, 4891 patients presented with mixed dyslipidemia (total cholesterol ≥ 5.2 mmol/L [200 mg/dL] and triglycerides ≥ 2.0 mmol/L [175 mg/dL]), and 188 DBL patients were identified based on the presence of an elevated very-low-density lipoprotein cholesterol/triglyceride ratio and were carriers of apoE2/E2. The APOE genotype or phenotype as well as the lipoprotein ultracentrifugation results were available for all patients. Results Among the laboratory variables associated with the lipid profile, the non-high-density lipoprotein cholesterol (HDL-C)/apoB ratio was the best predictor of DBL diagnosis based on the C-statistic. Previous proposed criteria had either low sensitivity or low specificity for the diagnosis of DBL. Using a non-HDL-C/apoB cut point of 3.69 mmol/g (1.43 in conventional units) followed by the presence of apoE2/E2 resulted in a good sensitivity (94.8%), negative predictive value (99.8%), specificity (99.6%), positive predictive value (88.5%), accuracy (99.4%), and area under the curve (0.97 [0.95–0.99]) for the prediction of DBL. Conclusion We therefore propose a 3-step algorithm for the diagnosis of DBL using total cholesterol and triglycerides as a first step, the non-HDL-C/apoB ratio as a second screening criterion and finally the APOE genotype, lipoprotein ultracentrifugation, or electrophoresis as a confirmatory test.