Improved diagnosis of familial hypercholesterolemia by correcting LDL-C for lipoprotein(a) in a German cohort.

BACKGROUND

Familial hypercholesterolemia (FH) is an autosomal-dominant disorder with elevated low-density lipoprotein cholesterol (LDL-C), contributing to premature atherosclerotic cardiovascular disease. The Dutch Lipid Clinic Network (DLCN) score and an LDL-C cutoff of 190 mg/dL are used in FH screening. However, LDL-C can be overestimated because of cholesterol content from lipoprotein(a) (Lp(a)-C), which may affect diagnostic accuracy.

OBJECTIVE

We examined how correcting LDL-C for Lp(a) influenced the ability of the DLCN score to discriminate between genetically confirmed (FH/M+) and genetically unconfirmed (FH/M-) patients with FH.

METHODS

We analyzed 384 German patients with suspected FH (237 women, 147 men) who underwent genetic testing. LDL-C and DLCN scores were corrected by 17.3%, 30%, and 45% of measured Lp(a). For both, we evaluated and compared model discrimination and calibration using receiver operating characteristic statistics and calibration plots.

RESULTS

LDL-C correction mainly reclassified patients with FH/M- (23-46), while FH/M+ reclassification was significantly less (7-20), depending on the correction level (17.3%, 30%, or 45%). 18 to 36 patients with FH/M- were reclassified from "Possible" to "Unlikely," whereas FH/M+ reclassification was rare (0-4). LDL-C correction improved specificity with minimal sensitivity loss. Areas under the curve for DLCN (0.782-0.803) and LDL-C (0.791-0.797) were similar, with no significant differences in calibration, indicating comparable performance for FH/M+ prediction.

CONCLUSION

Correcting LDL-C for Lp(a) improves diagnostic accuracy, especially by reducing false positives. A 17.3% correction effectively improved the specificity of DLCN and LDL-C while minimizing FH/M+ misclassification. LDL-C alone showed comparable performance to the DLCN score, supporting its potential as a practical diagnostic tool in clinical FH assessment.