Clinical characterization and mutation spectrum of German patients with familial hypercholesterolemia

Highlights

• 206 patients were screened for mutations in the genes LDLR, PCSK9 and APOB.

• 98 FH causing variants were found in 92 patients.

• 9 new “pathogenic” or “likely pathogenic” LDLR variants were identified.

• The mutation detection rate was 77.1% in definite FH patients (DLCN criteria).

• DLCNC score and LDL-C levels showed equal capacity to predict mutation carriership.

Abstract

Background and aims

Autosomal-dominant familial hypercholesterolemia (FH) is characterized by elevated plasma levels of low-density lipoprotein cholesterol (LDL-C) and a dramatically increased risk to develop cardiovascular disease (CVD). Mutations in three major genes have been associated with FH: the LDL receptor gene (LDLR), the apolipoprotein B gene (APOB), and the proprotein convertase subtilisin/kexin 9 gene (PCSK9). Here we investigated the frequency and the spectrum of FH causing mutations in Germany.

Methods

We screened 206 hypercholesterolemic patients, of whom 192 were apparently unrelated, for mutations in the coding region of the genes LDLR, PCSK9 and the APOB [c.10580G > A (p.Arg3527Gln)]. We also categorized the patients according to the Dutch Lipid Clinic Network Criteria (DLCNC) in order to allow a comparison between the mutations identified and the clinical phenotypes observed. Including data from previous studies on German FH patients enabled us to analyse data from 479 individuals.

Results

Ninety-eight FH causing variants were found in 92 patients (nine in related patients and 6 patients with two variants and likely two affected alleles), of which 90 were located in the LDLR gene and eight mutations were identified in the APOB gene (c.10580G > A). No mutation was found in the PCSK9 gene. While 48 of the LDLR mutations were previously described as disease causing, we found 9 new LDLR variants which were rated as “pathogenic” or “likely pathogenic” based on the predicted effect on the corresponding protein. The proportions of different types of LDLR mutations and their localization within the gene was similar in the group of patients screened for mutations here and in the combined analysis of 479 patients (current study/cases from the literature) and also to other studies on the LDLR mutation spectrum, with about half of the variants being of the missense type and clustering of mutations in exons 4, 5 and 9.

The mutation detection rate in the 35 definite and 45 probable FH patients (according to DLCNC) was 77.1% and 68.9%, respectively. The data show a similar discriminatory power between the DLCNC score (AUC = 0.789 (95% CI 0.721–0,857)) and baseline LDL-C levels (AUC = 0.799 (95% CI = 0.732–0.866)).

Conclusions

This study further substantiates the mutation spectrum for FH in German patients and confirms the clinical and genetic heterogeneity of the disease.

Introduction

Familial hypercholesterolemia (FH, OMIM 143890) is an inherited autosomal-dominant disorder, which is characterized by a reduced hepatic capacity to clear atherogenic cholesterol-rich low-density lipoproteins (LDL) from the circulation, leading to a marked elevation of serum LDL-cholesterol (LDL-C). Consequently, FH causes early morbidity and mortality from cardiovascular disease (CVD) (reviewed in Ref. [1]). Common tools for the clinical FH diagnosis are the Dutch Lipid Clinic network (DLCN) or the Simon Broome criteria, both representing a weighted score of the clinical FH characteristics including plasma LDL-C concentration, presence of tendon xanthoma, premature corneal arcus lipoides, cardiovascular disease and also the clinical family history because of the high likelihood of close relatives to be also affected from the disease. The prevalence of FH is estimated to be as high as one in 200–500 and might be even higher in populations with founder effects [2], [3], [4].

FH is caused by defects in one of at least three different genes, LDLR, APOB or PCSK9, coding either for the low density lipoprotein (LDL) receptor, its ligand apolipoprotein B, or for proprotein convertase subtilisin/kexin type 9, a protein involved in LDL receptor turnover [2], [3], [5], [6]. Autosomal-recessive hypercholesterolemia (ARH) is a rare disorder, which is clinically indistinguishable from FH and caused by mutations in the LDLRAP1 gene, leading to a complete loss of function of an adaptor protein required for receptor-mediated hepatic uptake of LDL-C.

About 60% of patients with the clinical diagnosis of FH may be mutation negative, i.e. no mutation is identified in the three established FH genes. This might partly be explained by an accumulation of common small-effect LDL-C raising alleles [7]. Another possibility is that variants in other genes, which have not yet been associated with FH, may cause the disease in patients negative for mutations in the known FH genes. Indeed, Fouchier and colleagues recently identified STAP1 as a possible fourth gene causing FH [8].

FH therapy includes lifestyle modification and treatment with a statin, first line therapy choice. International guidelines for prevention of CVD recommend rigorous lipid lowering treatment especially in FH patients. Early effective treatment is associated with a 75% reduction in lifetime risk of death from CVD [9]. Some countries, such as Netherlands and UK, already implement FH cascade family screening as a cost-effective tool to deal with FH and initiate lipid lowering treatment to prevent early CVD [10]. In most parts of Europe, however, FH is still underdiagnosed, with a detection rate below 15% [11]. In Germany, the diagnosis rate is not known.

Early diagnosis is essential for prevention of cardiovascular events in FH-patients. In the case of a clinically definite or probable FH (DLCN or Simon Broome criteria), genetic testing remains the gold standard in diagnostics and enables cost-effective cascade screening [12]. In the current study we report on the clinical characteristics of over 200 patients with hypercholesterolemia from Germany and the genetic variations identified in these patients. Additional consideration of some 200 published cases on German FH patients [13], [14], [15], [16] allowed us to describe the FH mutation spectrum and phenotype-genotype correlations based on data from >400 patients.