Wits FIND-FH – EAS 2019 Cascade screening for Familial Hypercholesterolemia in South Africa reveals a significant number of subjects with more than one FH mutation: The Wits FIND-FH Program Frederick J Raal 1 , El Mustapha Bahassi 2 , Belinda Stevens 1 , Traci A Turner 2 and Evan A Stein 2 . 1 Stein Center for FH, Carbohydrate and Lipid Metabolism Research Unit, Department of Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg. South Africa; 2 Medpace and Medpace Reference Laboratories (MRL), Cincinnati. Ohio. USA.  Until recently no systematic program existed to detect subjects with FH or to test their family members. Furthermore, information regarding prevalence of FH in black South Africans is sparse. The Wits FIND- FH program was initiated in late 2016 with the goal of addressing both these issues. One full-time and 1 part-time research nurse were hired and trained in late 2016 and beginning in January 2017 follow up of family members commenced based on index patients identified from the Wits Lipid Clinic. To date 700 subjects have been screened with 479 clinically diagnosed as probable or definite FH using the Simon Broome criteria. Demographic and LDL-C levels are shown in table 2. Genetic analysis confirmed 285/479 (59.5%) as having mutations consistent with FH –table 3. CNVs in the LDLR gene were found in 16 of these subjects (5.6%). The program has identified a small but growing number of black South Africans with FH, including 2 subjects with genetically confirmed homozygous FH - table 4. * Includes both Caucasian and SA ‘coloured or mixed’ populations Location on LDLR gene Codon Change Amino acid change Common name Population group Exon 4 Exon 9 Exon 4 Exon 14 NM_000527.4:C.681c>G NM_000527.4:C.1285G>A NM_000527.4:c523G>A NM_000527.4:c.2054C>T NM_000527.4:c.654_656deTGG D227E V429M D154N P664L GLY219del FH Afrikaner 1 FH Afrikaner 2 FH Afrikaner 3 FH Gujarat FH Lithuania Afrikaner* Afrikaner* Afrikaner* South Asian Lithuanian TABLE 1: Founder FH LDLR mutations common in South Africa Total (285) Male (129) Female (156) Treated (182) Untreated (103) Variant identified LDL-C (mmol/L) Apolipoprotein B (mg/dL) Lipoprotein (a) (nmol/L) 4.9 ± 2.5 135 ± 47 64 (22-152) 4.8 ± 2.2 133 ± 43 56 (20-147) 5.0 ± 2.6 137 ± 50 71 (24-162) 4.3 ± 2.0 125 ± 41 83 (22-167) 6.0 ± 2.7 153 ± 52 45 (21-121) Total (194) Male (94) Female (100) Treated (91) Untreated (103) No variant identified LDL-C (mmol/L) Apolipoprotein B (mg/dL) Lipoprotein (a) (nmol/L) 3.6 ± 1.2 113 ± 31 39 (15-151) 3.5 ± 1.3 113 ± 31 29 (11-106) 3.7 ± 1.2 113 ± 32 53 (20-166) 3.0 ± 1.1 104 ± 28 57 (18-171) 4.2 ± 1.1 121 ± 31 32 (11-95) LDL-C and Apolipoprotein B presented as mean ± SD. Lipoprotein (a) as median (IQR) TABLE 3 Race Mutation +ve Mutation -ve Spectrum of common FH causing mutations LDR mutations (n=187; 89%) FH Afrikaner-1 (c.681C>G = 73 FH Afrikaner-2 (c.1285G>A) = 50 130 (63%) FH Afrikaner-3 (c.523G>A) = 7 ApoB mutations (n=22; 10%) PCSK9 mutations (n=2; 1%) LDR mutations (n=44; 85%) FH Gujarat (c.2054C>T = 30 (68%) ApoB mutations (n=7; 13%) PCSK9 mutations (n=1; 2%) LDR mutations (n=7; 70%) FH Afrikaner-2 (c.1285G>T = 4 ApoB mutations (n=1; 10%) PCSK9 mutations (n=67%) LDR mutations (n=8; 67%) FH Cape Town (c.137-142 del) = 3 ApoB mutations (3, 25%) PCSK9 mutations (n=1; 8%) 121 (36%) 63 (55%) 6 (37%) 6 (37%) 211 (64%) 52 (45%) 10 (63%) 12 (63%) White (332) Indian (115) Mixed Race (16) Black African (16)  After obtaining an IRB approved written informed consent from a known FH index case, 1st degree relatives were contacted and a home or clinic visit arranged where after individual informed consent was obtained a targeted medical, cardiovascular, family and medication history, physical (including for skin and tendon xanthoma and corneal arcus) and blood sample were obtained. Fasting blood samples obtained from subjects were analyzed at Medpace Reference Laboratories (MRL), Leuven, Belgium for lipids and apolipoproteins and select chemistries to exclude underlying metabolic conditions known to cause secondary elevations of LDL-C.  In patients with likely FH by clinical assessment, DNA analysis for LDLR, APOB, PCSK9, LDLRAP1 mutations were analyzed by Next Generation Sequencing (NGS). Briefly whole blood samples were drawn into a K2EDTA tube and Initially processed at MRL Belgium and genomic DNA isolated using the QIAamp DNA Blood kit (Germantown, MD).  The genomic DNA (gDNA) was sequenced at MRL in Cincinnati, Ohio. Sequencing was performed for the coding regions of four genes (LDLR, APOB, PCSK9, LDLRAP1) known to account for the majority of cases of FH.5ApoB exons 1-29, LDLR exons 1-18, LDLRAP1 exons 1-9 and PCSK9 exons 1-12 were captured, amplified by PCR and subjected to bidirectional DNA sequencing using the Illumina Mi Sequencing platform(San Diego, CA). The sequencing probes were designed based on the GRch37/hg19 reference genome. Secondary and tertiary analysis of DNA sequences were performed using commercial bioinformatics software.  The bioinformatics method used also identifies DNA copy number variations (CNVs) in the LDLR gene which are the cause of FH in up to 10% of cases. Variants identified were compared to the GRCh37/hg19 reference genome and the pathogenicity of reported variants were determined according to current guidelines. 6  Familial hypercholesterolemia (FH) is an autosomal co-dominant disorder usually resulting from mutations in the LDL receptor (LDLR) gene and less commonly from mutations in apoB100, PCSK9 or LDLRAP1. This condition is characterized by elevated levels of LDL-cholesterol (LDL-C) and premature cardiovascular disease, particularly coronary artery disease (CAD) 1.  The heterozygous phenotype (HeFH) is characterized by elevated LDL-C levels approximately twice the normal levels (4.9 – 10 mmol/L), tendon xanthoma and premature CAD. If untreated, the cumulative risk of a coronary event by the age of 60 years is at >50% in men and 30% in women. The homozygous phenotype (HoFH) is characterized by LDL-C levels >13 mmol/L, skin and tendon xanthoma beginning soon after birth and if untreated CAD prior to age 20 years of age. There is however some overlap in the clinical phenotype between HeFH and HoFH when assessed by genotype. 2  FH is one of the commonest inherited diseases in the world with an estimated frequency of 1:200 to 1:250 for Caucasian populations 1 . However in some countries such as South Africa, the prevalence in certain population groups such as the Afrikaner, Jewish and south-Asian Indians it is as high as 1:80probably due to a founder effect. 3,4  In South Africa 70 to 80% of subjects of Afrikaner, Jewish or Indian origin with clinical heterozygous FH identified to date have one of 5 founder mutations - table 1. However the vast majority of FH patients remain undiagnosed and untreated and have not been screened for other mutations in the LDLR, ApoB, PCSK9 or LDLRAP1 genes.  Using phenotype cascade screening The Wits FIND-FH program has averaged 30 subjects monthly, found a clinical FH diagnosis in 68% of whom ≈ 60% were genetically confirmed. In addition, a number of genetic HoFH or compound HeFH patients who do not meet the traditional clinical criteria, especially LDL-C, for HoFH have been found. 2  The South African multi-ethnic society and well described founder effects emphasize the need for differential approaches to diagnosis and management of FH. Cascade testing of index cases based on phenotype is an important start and has identified many family members who were previously unaware that they had FH. The program is identifying a small but growing number of black South Africans with FH. While the prevalence of FH in virtually all populations in the world demonstrates a gene frequency of 1 per 200 to 500 there are no similar studies in black Africans where even case reports of FH in the literature are few. Studies involving larger cohorts and inclusive of different ethnicities are paramount to establishing an accurate prevalence of FH in black South Africans. 1. Nordestgaard BG, Chapman MS, Raal FJ et al. Familial hypercholesterolaemia is underdiagnosed and undertreated in the general population: guidance for clinicians to prevent coronary heart disease. Consensus Statement of the European Atherosclerosis Society. European Heart Journal 2013;34:3478-3490. 2. Raal FJ, Sjouke B, Hovingh GK, Isaac BF. Phenotype diversity among patients with homozygous familial hypercholesterolaemia: A cohort study. Atherosclerosis 2016;248:238-244. 3. Steyn K, Goldberg YP, Kotze MJ, et al. Estimation of the prevalence of familial hypercholesterolaemia in a rural Afrikaner community by direct screening for three Afrikaner founder low density lipoprotein receptor gene mutations. Hum Genet 1996;98:479-84. 4. Rubinsztein DC, van der Westhuyzen DR, Coetzee GA. Monogenic primary hypercholesterolaemia in South Africa. S Afr Med J 1994;84:339- 344. 5. Iacocca MA and Hegele RA. (2017). Recent advances in genetic testing for familial hypercholesterolemia. Expert Rev Mol Diagn. 17(7): 641- 651. 6. Richards S, et al. Standards and guidelines for the interpretation of sequence variants: A joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med 2015;17(5):405-24. TABLE 2 Discussion References Results Methods Background TABLE 5 – Subjects with two or more FH gene mutations TABLE 4 – Black African subjects with FH Subjects with two or more FH gene mutations: Five subjects met the clinical diagnosis for homozygous FH but DNA analysis revealed a further 32 patients, including 4 black African subjects, with two or more FH mutations. HoFH based on either a clinical or genetic diagnosis was therefore found in 37 (7.7%) of subjects who underwent DNA testing - table 5. Nineteen of these subjects (48%) has two mutations which are considered definitely pathogenic and the remainder has one or more mutations of uncertain pathogenicity. NUMBER AGE UNTREATED LDL (mmol/L) Clinical markers of FH CVD/AGE ONSET MUTATIONS found on NGS FH GENE VARIANT PATH 1 80 5.3 arcus Nil - - NEGATIVE 2 43 5.2 arcus,tendons Nil NM_000527.4:c.del137-142 LDLR PATH 3 25 11.1 arcus, tendons, xanthelasma Nil NM_000527.4:c.del137-142 NM_000527.4:c.del137-142 LDLR PATH PATH * 4 82 8.2 arcus, tendon TIA age 71 NM_000527.4:c.del137-142 LDLR PATH 5 79 11.4 arcus, tendon Nil NM_000527.4:c.1222G>A NM_000527.4:c.1104C>T LDLR PATH UNCERT 6 55 9.5 arcus, tendons, other Nil NM_000527.4:c.313+1G>A NM_000527.4:c.757C>T LDLR PATH UNCERT 7 57 4.3 arcus Nil - - NEGATIVE 8 55 11.5 arcus, tendon, xanthelasma Nil NM_000383.2:c.7501A>T APOB UNCERT 9 70 9.4 arcus Nil NM_000384.2:c.8889C>T NM_000384.2:c.7242A>C APOB UNCERT UNCERT 10 52 7.4 arcus, tendon Nil NM_000527.4:c.414G>C LDLR PATH 11 28 5.2 tendon Nil NM_000527.4:c.1285G>A LDLR PATH 12 86 9 arcus, tendon, planar Nil NM_000527.4:c.829G>A NM_000527.4:c. 2441G>A LDLR PATH PATH * 13 48 4.7 NO Nil NM_174936.3:c.1658A>G PCSK9 UNCERT 14 50 6.7 NO Nil - - NEGATIVE 15 27 3.7 arcus, tendon Nil - - NEGATIVE 16 45 5.2 NO Nil NM_000384.2:c.G>A APOB UNCERT NM_000527.4:c = LDLR mutation * = genetically confirmed homoygous FH NM_000383/4.2:c = ApoB mutation NM_174936.3:c = PCSK9 mutation NUMBER ETHNICITY AGE (years) UNTREATED LDL (mmol/L) Clinical markers of FH CVD/AGE ONSET Mutations found on NGS FH GENE VARIANT PATH 1 W 9 12.3 tendon,arcus, planar NM_000527.4:c.681C>G NM_527.4:C.1285G>A LDLR/LDLR PATH X 2* 2 BA 25 11.1 tendon,arcus, planar NM_000527.4:c.137_142del x 2 LDLR/LDLR PATH X 2* 3 I 13 18.6 tendon,arcus,planar NM_000527.4:c.2054C>T NM_000527.4:c.401G>A LDLR/LDLR PATH X 2* 4 W 3 25.4 tendon,planar NM_000527.4:c.1285G>A NM_000527.4:c.1285G>A LDLR/LDLR PATH X 2* 5 W 12 11.5 tendon, planar NM_000527.4:c.1285G>A NM_000527.4:c.1285G>A LDLR/LDLR PATH X 2* 6 W 76 7.2 arcus, tendon CABG age 46 NM_000527.4:c.681C>G Duplication exons 1-2 LDLR/LDLR PATH X 2 7 W 22 5 tendon NM_000527.4:c.681C>G Deletions exons1-3, 7-11, 13-18 LDLR/LDLR PATH X 4 8 W 49 4.5 tendon NM_000527.4:c.681C>G deletion exons 91-19 duplication exons 1-2 LDLR/LDLR PATH X 3 9 W 68 4 tendon NM_000527.4:c.681C>G Duplication exons 1-2 LDLR/LDLR PATH X 2 10 BA 79 8.1 arcus, tendon NM_000527.4:C1222g>a NM_000527.4:c.1104C>T LDLR/LDLR PATH X 2 11 I 43 7.4 arcus, tendon, planar, other CABG age 29 and 42 NM_000527.4:C268G>T NM_000527.4:c.1951G>A LDLR/LDLR PATH X 2 12 I 50 6.9 tendon MI age 42/CABG age 46 NM_000527.4:c.1724T>G NM_000527.4:c.649C>T LDLR/LDLR PATH X 2 13 I 45 9.5 arcus/tendon NM_000527.4:c.1724T>G NM_000527.4:c.649C>T LDLR/LDLR PATH X 2 14 W 67 4.2 nil NM_000527.4:c.2359G>A deletion exons 13-15 LDLR/LDLR PATH X 2 15 BA 86 9 arcus, tendon, planar NM_000527.4:c.829G>A NM_000527.4:c.2441G>A LDLR/LDLR PATH X 2 16 W 32 9.3 tendon NM_000527.4:c.662A>G NM_000527.4:c.2359G>A LDLR/LDLR PATH/UNCERT 17 I 9 6.7 nil NM_000527.4:c.2054C>T NM_000527.4:c.2478C>G LDLR/LDLR PATH/UNCERT 18 I 60 11.5 arcus, tendon, planar Deletion exons 9-10 NM_174863.3:c.1563C>T LDLR/LDLR PATH/UNCERT 19 BA 55 4.7 arcus, tendon, other NM_000527.4:c.313+1G>A NM_000527.4:c.757C>T LDLR/LDLR PATH/UNCERT 20 W 30 6.4 nil NM_000527.4:c.681C>G NM_000384.2:c.3383G>A LDLR/APOB PATH x 2 21 W 53 8.1 tendon PCI age 44 NM_000527.4:c.681C>G NM_000384.2:c.C5599C>T LDLR/APOB PATH x 2 22 W 23 6.1 tendon NM_000527.4:c.681C>G NM_000384.2:c.C5599C>T LDLR/APOB PATH x 2 23 W 56 8.7 tendon PCI age 47 NM_000527.4:c.681C>G NM_000384.2:c.C5599C>T LDLR/APOB PATH x 2 24 A 64 5.9 arcus, tendon CABG age 35 NM_000527.4:C2043C>A NM_000384.2:c.3927T>A NM_000384.2:C5254G>A LDLR/APOB/APOB PATH/UNCERT X 2 25 W 62 9.2 xanthelasma, arcus CABG age 31 NM_000527.4:c.1048C>T NM_00384.2:c.152A>G LDLR/APOB PATH/UNCERT 26 W 55 8.6 arcus, tendon NM_000527.4:c.1285G>A NM_000384.2:c.3383G>A LDLR/APOB PATH/UNCERT 27 W 29 7.4 nil NM_000527.4:c.1285G>A NM_000384.2:c.3383G>A LDLR/APOB PATH/UNCERT 28 W 33 7.6 nil NM_000527.4:c.1285G>A NM_000384.2:c.3383G>A LDLR/APOB PATH/UNCERT 29 I 13 12.9 nil NM_000527.4:c.2054C>T NM_000384.2:c.8227C>A LDLR/APOB PATH/UNCERT 30 W 40 5.2 tendon NM_000527.4:c.1285G>A NM_000384.2:c.3383G>A LDLR/APOB PATH/UNCERT 31 I 22 6 nil NM_000527.4:c.2054C>T NM_000384.2:c.8227C>A LDLR/APOB PATH/UNCERT 32 W 61 7.3 arcus NM_000527.4:c.682C>G NM_000384.2:c.3383G>A LDLR/APOB PATH/UNCERT 33 W 44 7.2 nil Duplication exons 1-2 NM_000384.2:c.10131G>A LDLR/APOB PATH/UNCERT 34 W 44 5.2 nil Deletion exons 12-14 NM_000384.2:c.3426G>A LDLR/APOB PATH/UNCERT 35 A 49 4.7 nil NM_000527.4:c.1104C>T NM_174936.3:c.267G>A LDLR/PCSK9 UNCERT X 2 36 W 45 6.1 nil NM_000527.4:c.2043C>A NM_015627.2:c.284G>A LDLR/LDLRAP1 PATH/UNCERT 37 W 50 4.8 nil NM_000384.2:c.12382G>A NM_000384.2:c.10131G>A APOB/APOB UNCERT X 2 NM_000527.4:c = LDLR mutation * = Clinical homozygous FH PCI = per cutaneous coronary artery intervention NM_000384.2:c = ApoB mutation CABG = coronary artery bypass graft surgery NM_015627.2:c = LDLRAP1 mutation MI = myocardial infarction NM_174863.3:c = PCSK9 mutation