Comparison of the mutation spectrum and association with pre and post treatment lipid measures of children with heterozygous familial hypercholesterolaemia (FH) from eight European countries

Highlights

• LDLR mutations are the most common cause of familial hypercholesterolaemia (FH) in children from 8 European countries.

• Overall, 279 different LDLR mutations were found in 2531 FH children.

• The frequency of APOB p.(Arg3527Gln) varied significantly over the 8 countries.

• APOB-FH was less severe than LDLR-FH for low density lipoprotein-cholesterol (LDL-C) concentration and family onset of coronary heart disease (CHD).

Abstract

Background and aims

Familial hypercholesterolaemia (FH) is commonly caused by mutations in the LDLR, APOB or PCSK9 genes, with untreated mean low density lipoprotein-cholesterol (LDL-C) concentrations being elevated in APOB mutation carriers, even higher in LDLR mutation and highest in those with a PCSK9 mutation. Here we examine this in children with FH from Norway, UK, The Netherlands, Belgium, Czech Republic, Austria, Portugal and Greece.

Methods

Differences in characteristics and pre- and post-treatment lipid concentrations in those with different molecular causes were compared by standard statistical tests.

Results

Data were obtained from 2866 children, of whom 2531 (88%) carried a reported LDLR/APOB/PCSK9 variant. In all countries, the most common cause of FH was an LDLR mutation (79% of children, 297 different), but the prevalence of the APOB p.(Arg3527Gln) mutation varied significantly (ranging from 0% in Greece to 39% in Czech Republic, p < 2.2 × 10⁻¹⁶). The prevalence of a family history of premature CHD was significantly higher in children with an LDLR vs APOB mutation (16% vs 7% p=0.0005). Compared to the LDLR mutation group, mean (±SD) concentrations of pre-treatment LDL-C were significantly lower in those with an APOB mutation (n = 2260 vs n = 264, 4.96 (1.08)mmol/l vs 5.88 (1.41)mmol/l, p < 2.2 × 10⁻¹⁶) and lowest in those with a PCSK9 mutation (n = 7, 4.71 (1.22)mmol/l).

Conclusions

The most common cause of FH in children from eight European countries was an LDLR mutation, with the prevalence of the APOB p.(Arg3527Gln) mutation varying significantly across countries. In children, LDLR-FH is associated with higher concentrations of LDL-C and family history of CHD compared to those with APOB-FH.

Introduction

Familial hypercholesterolaemia (FH) is a monogenic autosomal dominant inherited disorder characterised by elevated low-density lipoprotein cholesterol (LDL-C) concentrations from birth and a very high risk of developing coronary heart disease (CHD) at a young age [1], with a prevalence in many countries of around 1 in 250 [2]. Mutations in one of four genes involved in clearance of LDL-C from the blood are known to cause FH, most commonly in the LDLR gene, which encodes the low-density lipoprotein receptor (LDL-R), but mutations in apolipoprotein B (APOB), and gain-of-function (GoF) mutations in proprotein convertase subtilisin/kexin type 9 (PCSK9) can produce the phenotype [3]. Recently, it has been reported that a single mutation in the gene for APOE can also cause the FH phenotype [4,5].

Not all identified variants affect the gene-product and cause hypercholesterolemia. The ClinVar database has used criteria published by the American College of Medical Genetics (ACMG) [6] to determine the likely pathogenicity of published variants in LDLR/APOB/PCSK9 reported in patients with clinical FH [7]. Classifications are “definitely not” and “likely not pathogenic”, “variants of unknown significance” (VUS) and “likely” and “definitely pathogenic”. While more than 70% of the 2314 published LDLR variants are likely or definitely pathogenic, only 10% of the APOB and 13% of PCSK9 variants are classified as such [7]. Mutations in the LDLR gene can also be grouped into 5 classes based on results of functional studies using patient-specific cell culture [8]. Although there is a very large spectrum of different LDLR mutations causing FH [7], only one APOB mutation is common in Europeans, p.(Arg3527Gln), with a carrier frequency in gnomAD (https://gnomad.broadinstitute.org/) in non-Finnish Europeans of roughly 1/900. The frequency of this variant varies over Europe, being absent in Greece [9] and at a carrier frequency of roughly 1 in 200 in Switzerland [10]. In clinical FH patients where no causative mutation can be found, a polygenic cause of their hyperlipidaemia is most likely [11,12].

In the last 10 years, many National and European guidelines have been published for the identification and management of children with FH [[13], [14], [15], [16], [17], [18], [19], [20]], with lipid-lowering therapy using a statin as well as other agents being the key treatment recommendation. In the UK, the 2008/2017 NICE Guideline (CG71) recommends the diagnostic threshold for children under the age of 16 years should be a total cholesterol >6.7 mmol/l and/or LDL-C >4.0 mmol/l, and recommends statin therapy should be considered by the age of 10 years [13,18], while the European Atherosclerosis Society 2015 consensus statement [19] use a diagnostic threshold of LDL-C ≥5 mmol/l, or an LDL-C ≥4 mmol/l with family history of premature CHD and/or high baseline cholesterol in one parent, to make the phenotypic diagnosis. If a parent has a genetic defect, the LDL-C cut-off for the child is ≥ 3.5 mmol/l. This guideline recommends that statin use should be considered by the age of 8 years, and LDL-C be lowered below 3.5 mmol/l, if possible [19]. Both recommend use of Ezetimibe as an adjunct to statin therapy in those over the age of 10 years who are statin intolerant or who have not achieved the LDL-C target on a maximal tolerated statin dose. Children (and adults) with FH are also recommended to adopt a healthy life style to decrease their elevated cardiovascular risk (e.g. avoiding or stopping smoking, healthy eating, exercise).

In a study funded by the International Atherosclerosis Society (IAS), we have recently reported on the characteristics at diagnosis and the prevalence, age of initiation and the use of lipid-lowering treatment in FH children from eight countries across Europe [21]. In the current paper, we analyse the mutation spectrum in these children and examine the association between the gene mutation and predicted class of LDLR gene mutation and selected characteristics at diagnosis as recorded at registration as well as pre and post-treatment lipid concentrations. In adults with FH, compared to those with an LDLR mutation, those with the APOB mutation tend to have lower LDL-C concentrations and a better response to statin therapy [3]. This is due to the fact that VLDL remnants can be cleared by their intact LDL-receptors using apoE as a ligand, and that their intact LDL-R will be upregulated by statin therapy [22,23]. We wished to examine if this difference was also seen in children with either an LDLR mutation or the APOB mutation. A recent study on adults with FH showed better statin response in patients with a monogenic cause of FH vs mutation negative FH patients (the polygenic cause), another example of a genotype-phenotype correlation in FH [24].