Genetic Diagnosis of Familial Hypercholesterolemia in Asia

Familial hypercholesterolemia (FH) is a common genetic disease with an incidence of about 1 in 200–500 individuals. Genetic mutations markedly elevate low-density lipoprotein cholesterol and atherosclerotic cardiovascular disease (ASCVD) in FH patients. With advances in clinical diagnosis and genetic testing, more genetic mutations have been detected, including those in low-density lipoprotein receptor (LDLR), apolipoprotein B (APOB), proprotein convertase subtilisin/kexin type 9 (PCSK9), and so on. Globally, most FH patients remain undiagnosed, untreated, or inappropriately treated. Recently, there was a Global Call to Action by the Global Familial Hypercholesterolemia Community to reduce the health burden of FH. Asia, despite being the most populous continent with half of the global population, has low FH detection rates compared to Western countries. Therefore, we aimed to review the current status of FH genetic diagnosis in Asia to understand the gaps in FH diagnosis and management in this region.

However, globally, the majority of FH patients remain undiagnosed, untreated, or insufficiently treated (Nordestgaard et al., 2013;Khera et al., 2016, Cao et al., 2018Wang et al., 2019). Based on clinical criteria, FH is not rare in the Asian population . However, around 42.5% to 62.5% of the FH patients remain underdiagnosed, according to definitions by the Dutch Lipid Clinic Network (DLCN) and Simon Broome (SB). This finding supports the importance of genetic testing for the diagnosis of FH (Cao et al., 2018).
Furthermore, there are great advances in diagnostic technologies, such as next generation sequencing (NGS) (Di Resta and Ferrari, 2018), to improve the genetic diagnosis of FH. In the era of targeted and personalized medicine (Prodan Žitnik et al., 2018), it will be useful to identify different patient types to aid the selection of safer and more effective treatments.
Recently, there was a Global Call to Action by the Global Familial Hypercholesterolemia Community to reduce the health burden of FH (Wilemon et al., 2020). Asia is the most populous continent in the world, with more than half of the world's population. It is therefore important to better understand the current status of FH diagnosis and management in Asia. Here, we aimed to review the current status of FH genetic diagnosis in Asia.

FH DETECTION IN ASIA
There are two related studies on FH detection in Asia (Nordestgaard et al., 2013;Pang et al., 2019). Nordestgaard et al. (2013) reported the estimated number and percentage of individuals diagnosed with FH globally. Among roughly 200 countries/territories worldwide, data were only available in 22 countries/territories, including 3 in Asia. The data from Asian countries were derived from the second WHO consultation on FH in 1998. The estimated percentage of individuals diagnosed with FH was only 1% (n = 14,100) in Hong Kong, <1% (n = 46,300) in Taiwan, and <1% (n = 254,800) in Japan (Nordestgaard et al., 2013).
Another study in 2019 compared FH health care in 12 countries, including 9 countries selected in the Asia-Pacific region, and 2 in the Southern Hemisphere and United Kingdom (UK). In the United Kingdom, which was the benchmark, the estimated percentage of individuals diagnosed with FH was 10-20%. In non-Asian countries, the estimated percentage was 5.3% in South Africa, 4% in Australia, 2.9% in Brazil, and 2.5% in New Zealand. In Asian countries, however, the estimated percentage was only 3.8% in Taiwan, 2.2% in Hong Kong, 1.4% in Malaysia, 1% in Japan, <1% in China, <1% in the Philippines, and <1% in Vietnam. Although there is a continuous improvement in FH detection rates, compared to the 2013 data, there are still important gaps in comparison to other non-Asian countries and the United Kingdom (Pang et al., 2019).

AWARENESS AND KNOWLEDGE OF FH IN ASIA
The awareness and knowledge about FH have been surveyed among physicians across Asia (Pang et al., 2015(Pang et al., , 2017. One anonymous internet-based survey involving 230 physicians was reported in 2015 (Pang et al., 2015). These physicians were selected from Japan, South Korea, and Taiwan. Only a small proportion of them was aware of the heritability (47%), prevalence (27%), risk of FH-related cardiovascular disease (13%), and FH specialists in their geographic area (35%) (Pang et al., 2015).
Another survey was done with 1,078 physicians and reported in 2017 (Pang et al., 2017). These physicians were from 10 different countries/regions, including Australia, Japan, Malaysia, South Korea, Philippines, Hong Kong, China, Vietnam, Taiwan, and the United Kingdom. Self-perceived FH familiarity was only 34% and there were no significant differences among most countries (except in Japan and China, where the value was lower) and the United Kingdom. Physicians from the Asia-Pacific region were significantly worse at selecting the correct FH descriptions (72%), compared to those from the United Kingdom (89%) (P = 0.001). FH guideline awareness was significantly lower in physicians from the Asia-Pacific region (35%), compared to those from the United Kingdom (61%) (P < 0.001). The awareness of lipid specialists for referral or medical advice was also significantly lower in physicians from the Asia-Pacific region (35%), compared to those from the United Kingdom (50%) (P = 0.003) (Pang et al., 2017).
These studies showed that the awareness, knowledge, and perception about FH were still low among Asian physicians, particularly in developing countries (Pang et al., 2015(Pang et al., , 2017. More efforts are needed for improvement in these parameters, where extensive education and training programs could be of help.

GENETIC DIAGNOSIS OF FH IN ASIA
Taiwan FH genetic studies in Taiwan are summarized in Table 1.
Some studies have investigated genetic diagnosis techniques in Taiwan. A custom DNA resequencing array was designed to detect mutations on all three FH-causing genes (LDLR, APOB, and PCSK9) and 290 known LDLR insertion/deletion mutations in the Taiwanese patients. In previously sequenced patients, the average call rate was 98.45% and the agreement between microarray and capillary sequencing was 99.99%. In screening patients blindly, the FH array detected at least 1 mutation in 77.5% of patients clinically diagnosed with definite FH and in 52.9% of patients with probable FH. The high-throughput FH resequencing array detected LDLR, APOB, and PCSK9 with high efficiency and accuracy and identifies disease-causing mutations (Chiou et al., 2011).
A mass spectrometric assay (Agena iPLEX assay) was designed to simultaneously diagnose 68 point mutations in FH-causing genes in Taiwanese patients. In the first part of the study, only 1 discrepancy was found between the mass spectrometry and Sanger sequencing data of 180 previously sequenced patients. In the second part, 62 probands with mutations were identified by both techniques out of 185 FH probands, with only 5 mutations detected by Sanger sequencing. The sensitivity and specificity of mass spectrometry were 92.5% and 100%, respectively. This study found great potential for this assay due to its low cost, speed (around 1 day), and flexibility for FH genetic screening in Taiwanese patients (Chiou and Charng, 2017).
Probes for NGS were designed to capture the whole LDLR gene and all coding sequences of APOB and PCSK9 in Taiwanese patients. These probes correctly identified all the variants in 13 DNA samples, including 3 large duplications and 2 large deletions. In a new cohort of 28 unrelated FH patients with a Dutch Lipid Clinic Network score ≥ 5, they identified the causative variants in 21 unrelated probands. Five of them carried the novel splice site variant c.1186 + 2 T > G in LDLR. This study showed that this panel can comprehensively detect disease-causing variants in LDLR, APOB, and PCSK9 in FH patients (Hsiung et al., 2018).
With the advances in molecular diagnostic methods, the detection rate of genetic mutations in clinically diagnosed FH has increased in Taiwan in recent years. However, there were still large proportions of FH patients that did not have genetic diagnoses. There were multiple LDLR and APOB mutations identified in Taiwan, including some novel mutations. However, PCSK9 mutations, which had been reported in China and Japan, have not been identified in Taiwan.

China
FH genetic studies in China are summarized in Table 2.
In a multi-center study in China, 285 unrelated index cases with clinical FH were analyzed by NGS for LDLR, APOB, and PCSK9 mutations. Genetic mutations were detected in 148 cases (51.9%), including 84 (29.4%) in LDLR, 31 (10.9%) in APOB, and 4 (1.4%) in PCSK9 genes. LDLR c.1448 G > A (p.W483X) was detected in 9 patients, which was the most frequent mutation. There were 8 novel mutations (7 LDLR and 1 APOB), which were considered as pathogenic by in silico analysis. There were 5 true LDLR homozygotes, 16 compound heterozygotes, and 13 double heterozygotes identified. The most severe phenotypes were noted in true LDLR homozygotes .
There was a review of 163 case reports about Chinese FH patients published before 2018. The mutation frequency was 82% in LDLR, 9% in APOB, and very rare in PCSK9. The study concluded that the epidemiological investigation of FH was not large-scale, FH recognition remained rudimentary, and guidelines for diagnosis and management of FH patients were incomplete in China (Peng et al., 2019).
In a previous systematic review of LDLR mutations in Chinese FH patients, a total of 74 studies were included. Among the 295 probands studied, 131 LDLR mutations were identified. Most of them were located in exon 4, and approximately 60% of them were missense mutations. There were 30 novel mutations and most of them were found pathogenic by in silico analysis. The major LDLR mutations were LDLR c.986 G > A (p.C329Y), LDLR c.1747 C > T (p.H583Y), and LDLR c.1879 G > A (p.A627T) (Jiang et al., 2015).
The majority of FH mutations found in Chinese were in LDLR, there was one in APOB, and were rare in PCSK9. The functionality of these novel mutations was predicted by in silico analysis. Functional testing or familial co-segregation analysis may be required to check the pathogenicity of these mutations.

Hong Kong
FH genetic studies in the Hong Kong are summarized in Table 3.
Hu et al. reported the epidemiologic features of 252 HeFH patients from the Chinese population in Hong Kong. For patients aged ≥ 18 years, xanthomata was noted in 40.6% of males and 54.8% of females, while corneal arcus was noted in 81.2% of males and 66.9% of females. The incidence of coronary artery disease (CAD) was 9.9% and 8.5% in males and females, respectively. This study found that age and the presence of xanthelasmata were associated with the risk of CAD (Hu et al., 2013(Hu et al., , 2016. In one genetic study from Hong Kong, DNA sequencing was performed for the promoter and coding regions of the 18 exons of the LDLR gene in 30 Chinese FH patients. Overall, LDLR mutations were detected in 21 patients (70.0%). There were 18 mutations in the promoter and 10 exons. Besides, there were 6 polymorphisms with genotypic distributions similar to those in the Japanese population but different from those in the Western populations. For the 9 clinically diagnosed FH patients with no detectable LDLR gene mutations, there was also no APOB R3500Q mutations detected (Mak et al., 1998).
In another genetic study from Hong Kong, a total of 94 unrelated probands with clinical FH received genetic testing for LDLR, APOB, and PCSK9 by Sanger sequencing. Single-gene mutations were detected in 54 probands (57.4%), including 51 (54.2%) in LDLR and 3 (3.2%) in APOB. Heterozygous LDLR mutations were the most common. Compound heterozygous mutations were detected in five cases. Double heterozygous mutations were detected in three cases. A homozygous mutation in exon 10 of LDLR [c.1474G > A (p.D492N)] was detected in one case (Tan et al., 2018).
In a recent study, Chinese subjects with clinical FH received family based cascade screening and genetic testing. Among 96 index cases, 42 distinct mutations were identified in 64 cases (67%). The majority of mutations were in LDLR, and the three most common mutations were LDLR c.1241 T > G (p.L414R), LDLR c.1474G > A (p.D492N), and LDLR c.682G > A (p.E228X). Nine novel LDLR mutations were identified. APOB c.10579 C > T (p.R3527W) was identified in 5% of the total cases .

Japan
FH genetic studies in Japan are summarized in Table 4.  In a previous report, 385 clinically diagnosed HeFH patients received genetic analysis for APOB 3500 mutations to estimate the frequency of familial defective apolipoprotein B-100 (FDB) in Japan. The study detected no APOB 3500 mutation in this cohort (Nohara et al., 1995).
In another study conducted in the Hokuriku district of Japan, 25 clinical HoFH patients received genetic analysis reports. There were 15 true homozygotes and 10 compound heterozygotes for LDLR mutations. Five of these patients had PCSK9 mutation [PCSK9 c.94G > A (p.E32K)], including 2 true homozygotes and 3 compound heterozygotes (LDLR and PCSK9). APOB mutations were not detected in any FH patient. The study also confirmed a high incidence of HoFH (1/171,167) and HeFH (1/208) in the Hokuriku district of Japan (Mabuchi et al., 2011).
Korea FH genetic studies in Korea are summarized in Table 5.
In 28 clinically diagnosed HeFH patients, genetic analysis was performed for LDLR gene by Southern blot hybridization. Three of them (10.7%) were diagnosed with two large LDLR deletion mutations (FH29 and FH110) (Chae et al., 1997).
In 80 clinically diagnosed FH patients, genetic analysis was performed to detect small structural rearrangements of the LDLR gene, which were undetected by southern blot hybridization. Three novel small deletions were identified (FH 2, FH 34 and FH 400) .
In 45 unrelated HeFH patients, genetic analysis was performed to assess the LDLR gene mutations by long-distance PCR. Two different large deletion mutations were identified (FH6 and FH 32) .
In 20 unrelated FH patients, genetic analysis was performed for the LDLR gene by single-strand conformation polymorphism. LDLR mutations were detected in five patients (25%). There were four novel point mutations, including one non-sense mutation and three missense mutations (Shin et al., 2000).
To study the role of the LDLR gene in polygenic hypercholesterolemia, a study was conducted in 244 members of 43 different pedigrees (15 normal and 28 FH pedigrees) and 245 individuals (187 normal and 58 FH) for 9 different restriction fragment length polymorphisms (RFLPs) (TaqI, StuI, HincII, BstEII, AvaII, PvuII, MspIA, MspIB, and NcoI) and a sequence variation at Arg450. Frequencies of these polymorphisms were similar between FH patients and controls. Among the 512 possible combinations for the nine polymorphic sites, 39 unique haplotypes were detected. The four most common haplotypes accounted for 59.4% of the total patients sampled.
The study found that the haplotypes indicated marked linkage disequilibrium for these 10 sites and throughout the region containing the LDLR gene (Chae et al., 2001). In 31 clinically diagnosed HeFH patients, genetic analysis was performed for the LDLR gene by single-strand conformation and heteroduplex formation. LDLR gene mutations were detected in 25 patients (80.6%). There were sixteen different LDLR mutations. The study identified one mutation that had only been reported in a Korean FH patient, the in-frame 36-bp deletion (1591del36) in exon 11. Five novel mutations were identified, including LDLR c.311G > A (C104Y), c.661del17, c.1705insCTAG, c.2088C > A (C696X), and c.941-1G > A ( Kim et al., 2004).
In 69 clinically diagnosed FH cases, whole-exome sequencing analysis was applied to identify genetic mutations in the 3 known FH-related genes (LDLR, APOB, and PCSK9 genes). Genetic mutations were detected in 23 patients (33.3%). The most common genetic mutation was in the LDLR gene, which was identified in 19 patients (82.6%), included 2 copy number deletions and 17 mutations. There was also one APOB gene mutation and one PCSK9 gene mutation. Two frameshift deletions in the LDLR gene and one PCSK9 mutation were novel causative mutations. By co-segregation in their relatives, one copy number deletion and novel mutation were validated .
One study enrolled patients with >190 mg/dL LDL-C and xanthoma or a family history of CAD or hypercholesterolemia. Among 97 patients, putative pathogenic mutations in LDLR, APOB, or PCSK9 genes were identified in 31 patients (32%). When comparing the four clinical diagnosis criteria of FH (Simon Broome, Dutch, MEDPED, and Japanese), mutation detection rates were higher using the MEDPED criteria (67-75%) and lower in the Simon Broome or Dutch criteria (35%-37%). The study found that 225 mg/dL was the best LDL-C threshold for putative mutations in Korean FH patients (Shin et al., 2015).
The polygenic cause of FH in Korean patients was investigated in 97 FH patients and 2,274 controls from the Korean Health Examinee (HEXA) shared control study. These patients received genotyping for 12 single nucleotide polymorphisms (SNPs) used in prior studies to assess the polygenic causes of FH in Caucasian patients and 4 SNPs associated with LDL-C levels in East Asians. The study found that mutation-negative FH patients had higher SNP scores than controls. The study also showed that SNP score analysis could be used for the identification of the polygenic cause of FH in Korean patients (Kwon et al., 2015).

Malaysia
FH genetic studies in Malaysia are summarized in Table 6.
A total of 140 clinically diagnosed autosomal dominant hypercholesterolemia (ADH) and 111 healthy control subjects were analyzed for selected SNPs in LDLR, APOB, PCSK9, and other genes. The ADH patients included Malays (60.0%), Indians (25.0%), and Chinese (15.0%). A total of 310 markers were examined, including 73 from LDLR, 130 from APOB, and 107 from PCSK9. Significant differences in allele frequency among Malaysians and other populations (European Whites, Han Chinese, Yoruba, and Gujarati Indians) were noted in 23 markers when compared to publicly available data. These markers included 2 on LDLR, 17 on APOB, and 4 on PCSK9 (Alex et al., 2012).
In another study, 141 patients with clinically diagnosed FH and 111 unrelated control subjects were genetically analyzed via a high throughput microarray genotyping platform. Among the FH cases, 108 (76.60%) had at least one SNP that was associated with FH risk. Eleven of the 14 SNPs were significantly associated with an increased risk of FH, including 1 SNP in LDLR, 7 in APOB, and 3 in PCSK9. The study showed that APOB rs12720762 was associated with the highest risk of FH (odds ratio 14.78, p < 0.001) (Lye et al., 2013).
Two independent studies from the same team in Malaysia used SNPs for genetic analysis (Alex et al., 2012;Lye et al., 2013). Although they found SNPs with different allele frequencies between Malaysians, European Whites, Han Chinese, Yoruba, and Gujarati Indians, the functional significance of these SNPs requires further investigation.
Singapore FH genetic studies in Singapore are summarized in Table 7.
In Singapore, 96 clinically suspected FH patients were genetically analyzed via NGS of 26 lipid-related genes. The cohort included Chinese (78.1%), Malays (13.5%), Indians (5.2%), and other ethnicities. A total of 50 patients (52.1%) had LDLR mutations and 4 (4.2%) had APOB mutations. No PCSK9 mutations were detected in this cohort. There were 15 novel LDLR mutations. The study concluded that the mutation distribution was similar to other Asian countries, but the spectrum was different locally (Pek et al., 2018).    Philippines FH genetic studies in Philippines are summarized in Table 8.
In the Philippines, 60 suspected FH patients were analyzed for LDLR mutations. LDLR mutations were detected in 12 patients (20%), and six novel mutations were identified. The study found that LDLR mutations were significantly associated with LDL-C levels, FH scores, and a family history of dyslipidemia (Punzalan et al., 2005).
In addition to LDLR mutations, further studies are still needed to identify relevant mutations in APOB, PCSK9, and other FHrelated genes in this population.

Sri Lanka
FH genetic studies in Sri Lanka are summarized in Table 9.
In Sri Lanka, a total of 27 clinically diagnosed FH patients were analyzed for LDLR mutations using Sanger sequencing. Among these patients, pathogenic missense mutations were detected in only five (18.5%). Four patients were heterozygous for LDLR mutations, including LDLR c.682 G > C (p.E228Q) in 2 patients, LDLR c.1720C > A (p.R574S) in 1, and LDLR c.1855 T > A (p.F619L) in 1. One patient was compound heterozygous for LDLR c.2289 G > T (p.E763D) and LDLR c.1670 C > G (p.T557S). The study found that LDLR mutations were markedly low in this population (Paththinige et al., 2018). This is the first report on LDLR mutations in this population. However, the detection rate of genetic mutations was low. Furthermore, the study did not screen for other FH mutations, including those in APOB, PCSK9, and LDLRAP1.

India
FH genetic studies in India are summarized in Table 10.
In one Indian study, a total of 100 potential FH cases were genetically analyzed using Sanger sequencing, MLPA, and NGS. Genetic mutations were identified in 47 cases (47.0%). Based on modified Dutch Lipid Clinic Network criteria, mutations were detected in 91.4% of definite FH cases, in 40% of probable FH cases, and in 18.8% of possible FH cases. A total of 38 pathogenic variants were identified, including 33 LDLR mutations, 3 APOB mutations, and 2 PCSK9 mutations. There were ten novel pathogenic variants. Interestingly, a likely founder mutation in intron 10 (c.1587-1 G > A) of the LDLR gene was observed in 6 North Indian families. However, the conventional pathogenic variants in previously reported LDLR, APOB, and PCSK9 mutations were not detected. This study found different genetic variants between the Indian population and Western populations (Setia et al., 2020).
In addition, there were other genetic studies as abstracts from India. However, we did not report them in this review article.

GENETIC DIFFERENCES OF FH BETWEEN DIFFERENT ETHNICITIES
Among the genetic studies in Asia, the detection rates of genetic mutations varied. The differences may be due to either the diagnostic methods used or the ethnicities sampled. With the advances in genetic diagnostic methods, detection rates will be further improved. Most of the genetic studies in Asia were confined to one ethnicity. Only a few studies enrolled FH patients with different ethnicities. Data from Singapore revealed a similar distribution of mutations as compared to other Asian countries (Pek et al., 2018). Data from Malaysia revealed significant differences in the allele frequency of markers for LDLR, APOB, and PCSK9 genes compared to publicly available data from other populations (Alex et al., 2012). Until now, there is limited literature discussing the genetic differences in FH between Asian and non-Asian populations. Further studies are still needed to compare the genetic differences in FH between different ethnicities.

CONCLUSION
Although there has been a continuous improvement in FH detection rates, there are still important gaps to fill in Asian countries in comparison to other countries and the United Kingdom. FH awareness, knowledge, and perception remain low in Asian physicians, particularly in less economically developed countries. More efforts are required to close this gap, such as extensive education and training programs. Although there have been some FH genetic studies in the Asian population, each utilizing different techniques, further studies are required to understand the different genetic backgrounds of FH in Asia.

AUTHOR CONTRIBUTIONS
C-CH and M-JC contributed to conception and design of the study, manuscript revision, read, and approved the submitted version. C-CH wrote the first draft of the manuscript.