Diagnostic accuracy of cardiovascular and imaging biomarkers to identify index patients with familial hypercholesterolaemia.

Abstract

Objective: To determine the utility of secondary stratification measures in ascertainment of index cases for monogenic familial hypercholesterolaemia (FH).

Methods: Referrals from primary care were screened by methods for the potential diagnosis of FH, including Simon Broome (SB) or Dutch Lipid Clinic Network score (DLCN) criteria, initial LDL-C, lipoprotein (a) (Lp(a)) > 125 nM, troponin-T (hsTnT), imaging using carotid intima-media thickness and plaque assessment and a single nucleotide polymorphism (SNP) polygenic hypercholesterolaemia panel (12 loci).

Results: The population comprised 793 patients aged 55 ± 17 years, of whom 3% had tendon xanthomata, 7% coronary artery disease, and with pre-treatment LDL-C 5.84 ± 1.47 mmol/L. Genotyping was performed in 793 patients and 36% had monogenic FH. Dutch lipid score assessment was associated with a positive likelihood ratio (PLR) for FH 3.91 with a net reclassification index (NRI) of 8% while addition of negative modification for triglycerides (Welsh lipid score) had a PLR 6.88 (NRI 30%). In the whole cohort, the SNP12 score had a negative LR (NLR) of 1.32 (NRI -16%) above the 75^th centile while Lp(a) > 125nmol/L had a NLR of 1.18 (NRI -29%) and raised hsTnT a PLR of 1.08 (NRI -16%). In a non-pre-stratified primary care cohort (n = 236), imaging had a PLR 1.70 (NRI 14%) for identifying patients with FH.

Conclusions: A clinical algorithm based on Welsh Lipid score criteria modifying DLCN score for triglycerides allied with stratification for the presence of tendon xanthomata, highly elevated LDL-C (>7 mmol/L) or positive imaging provides an efficient system to raise the yield of diagnosis of FH with a low chance of missing cases.