Analysis of routine molecular genetic diagnostics of familial hypercholesterolemia at tertiary care lipid clinics

Abstract

Abstract Background Familial hypercholesterolemia (FH) is an autosomal dominant genetic disease, causing dyslipidemia and premature atherosclerotic cardiovascular disease (ASCVD), with currently ∼2,000,000 unidentified adult cases as well as ∼500,000 children cases in Europe. Purpose Aim of this study is to analyse the current yield of routine molecular genetic diagnostics for FH and the spectrum of mutations in two specialist lipid clinics. Methods We investigated all genetic tests performed for FH in our clinics over a 4-year period. Reports of causative mutations in the main FH-causing genes including LDLR, APOB, PCSK9 were collected. Variants were classified based on HGVS nomenclature using the mutation database ClinVar. For clinical classification, the Dutch Lipid Clinic Network Score (DLCNS) was used, which is based on off-treatment low-density lipoprotein cholesterol (LDL-C) ranges, family and patient history as well as pathognomonic signs. Results Of 473 patients tested, 102 (21.6%) had a causative mutation (LDLR=84%, APOB=9%, PCSK9=5%), three of which were novel. 14 further, previously unreported variants of unknown significance were identified. The detection rate was 64% in patients with clinically definite FH i.e. DLCNS >8 (10.7% of the cohort), 30% in those with a DLCNS between 6 and 8 (20.6%), 16% in those with a DLCNS between 3 and 5 (47.2%) and 4% in DLCNS <3 (21.5%). The detection rate among those with an off-treatment LDL-C ≥190 mg/dL was 30.5%, and 7.5% in those <190 mg/dL. Median DLCNS at testing was 4, in positively tested 6. Mean off-treatment LDL-C levels were significantly higher in positively compared to negatively tested (266±67 mg/dl vs 198±61 mg/dl; p<0.001), and in those with DLCNS ≥6 compared to <6 (284±67 mg/dl vs 190±53 mg/dl p<0.001). The currently established LDL-C cut-off of 190 mg/dL showed a sensitivity of 90.6% [80.1 - 96.1] and a specificity of 35.9% [29.5 - 42.9], while a DLCNS cut-off of ≥6 shows a sensitivity of 60.3% [48.1 – 71.3], a specificity of 76.7% [71.0 – 81.6]. Both performed well in ROC analyses with an AUC of 0.632 for the LDL-C cut-off of 190 mg/dL vs 0.707 for DLCNS≥6, with an insignificant difference in AUC between the two (-0.075; p=0.053). Conclusions Practically, a simple off-treatment LDL-C cut-off of 190 mg/dL, performing comparably to the more time-consuming DLCNS might enhance the identification of suspected cases especially by non-specialists and in primary care, allowing for a much needed surge in referrals to specialist centers for diagnostic workup, thereby possibly improving diagnosis rates in this severely underdiagnosed disease.Mutation detection within DLCNS groupsOff-treatment LDL-C by mutation status