Dear Sir:
Mendelian stroke is a significant cause of stroke in young individuals [1,2]. However, its exact prevalence remains unclear, as most studies have focused on selective stroke populations targeting specific disorders. A recent study indicated that genetic variants associated with rare Mendelian disorders may also elevate the risk of sporadic strokes [3]. Consequently, identifying these genetic variants in the general population is crucial for understanding the broader risk of stroke.
Interestingly, the prevalence of pathogenic variants causing Mendelian stroke varies significantly across different populations, with the highest prevalence observed among East Asians in a study that analyzed the Genome Aggregation Database (gnomAD) from 101,635 individuals across seven ethnic groups [4]. In particular, pathogenic NOTCH3 variants, which cause cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) were the most frequently identified genetic variants among East Asians in the report. Recent studies from Taiwan and Jeju Island, South Korea also reported a similarly high prevalence (0.9%) of these variants among the residents [5,6]. Despite this, East Asians only represent 7.1% of the gnomAD, with merely 1,909 Koreans included [7]. This raises questions about whether Koreans have a higher frequency of these pathogenic variants linked to Mendelian stroke. In this study, we screened for genetic variants responsible for Mendelian stroke in a cohort of 15,548 Koreans without neurological disorders, including stroke.
We extracted genetic information from the whole-genome sequencing data of 15,548 individuals obtained through the Pilot Project for National Bio Big Data of Korea at the Clinical & Omics Data Archive (https://coda.nih.go.kr/stats/selectRegList.do). The mean age was 46 years and 47.6% of males. We examined 18 genes identified in an earlier report (ABCC6, APP, CCM2, CECR1, COL3A1, COL4A1, COL4A2, COLGALT1, CST3, CTSA, GLA, HTRA1, ITM2B, KRIT1, NOTCH3, PDCD10, RNF213, and TREX1) [4]. Additionally, we analyzed six more genes linked to Mendelian stroke (ACVRL1, CBS, ENG, FBN1, HBB, and NF1) (Table 1). We extracted the rare nonsynonymous variants located within exons with a minor allele frequency (MAF) of less than 0.001, based on the gnomAD v2.1.1 (https://gnomad.broadinstitute.org/) non-neuro dataset. In line with the previous report, two well-established pathogenic founder NOTCH3 variants (R544C and R1231C) and one well-established RNF213 pathogenic founder variant (R4810K) were included in the analysis regardless of MAF. Variants were classified into three groups as suggested by Grami et al. [4]: pathogenic clinical variants, variants with a Combined Annotation Dependent Depletion (CADD) score >20, and all nonsynonymous variants. For pathogenic clinical variants, we further classified them as pathogenic or likely pathogenic variants according to the ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar/). To compare variant frequencies between Koreans and gnomAD participants, Fisher’s exact test was used with Bonferroni correction for multiple testing. Detailed genomic analysis methods are provided in Supplementary Methods. This study was approved by the Institutional Review Board of Jeju National University Hospital (JEJUNUH 2022-06-006-008).
Pathogenic clinical variants were found in 17 of the 24 genes studied among Koreans, with a prevalence of 2.8% in 15,548 individuals without neurological disorders. Pathogenic clinical variants were most frequently found in the RNF213 gene (21.8 per 1,000 individuals), followed by NOTCH3 (1.8), FBN1 (1.2), and TREX1 (0.5) (Table 2). Of 39 pathogenic clinical variants found in Koreans, 18 variants were found in only Koreans, but not in gnomAD (Supplementary Table 1). Compared with gnomAD, pathogenic clinical variants of RNF213 were significantly more common, while variants in ABCC6, CBS, CECR1, HBB, and NOTCH3 were less common among Koreans. The R4810K variant in RNF213 alone was present in 339 individuals, corresponding to 21.8 per 1,000 individuals. Among the NOTCH3 pathogenic variants, we identified four pathogenic or likely pathogenic variants (R75P, R544C, R587C, and R1076C), three of which were cysteine-altering variants. Overall, the prevalence of pathogenic variants and likely pathogenic variants in Koreans was 5.4 per 1,000 individuals and 22.5 per 1,000 individuals, respectively (Supplementary Table 2). For CADD-predicted variants, the most frequently involved gene was also RNF213, followed by NOTCH3, FBN1, COL4A2, and NF1. Among nonsynonymous variants, RNF213 remained the most frequent gene, showing similar patterns to pathogenic or CADD-predicted variants. Limiting the analysis to the 18 genes suggested in the earlier report kept the pathogenic carrier frequency at 2.5%, as six additional genes contributed little.
Previous studies have reported a high prevalence of pathogenic NOTCH3 variants among East Asians, particularly the R544C variant [8]. This variant was found in 0.9% of the Taiwanese general population and among the residents of Jeju Island, South Korea [5,6]. However, in our analysis, the overall frequency of pathogenic NOTCH3 variants was only 0.18%, consistent with our previous report indicating a frequency range of 0.12%-0.44% [5]. Therefore, the prevalent R544C variant does not appear to be a general feature among Koreans, and the effect of genetic drift associated with the founder effect should be considered for Jeju Island and Taiwan.
Despite the lower-than-expected frequency of NOTCH3 pathogenic variants, the overall frequency of pathogenic clinical variants associated with Mendelian stroke in Koreans was similar to the frequency reported in East Asians by Grami et al. [4] This was primarily due to the significantly higher frequency of the RNF213 variant among Koreans in this study compared to the report by Grami et al. [4] The RNF213 variant is well-known as a susceptibility gene for moyamoya disease, and our findings align with the previously reported prevalence (2.25%-2.65%) of this variant among 1,516 healthy Koreans [9].
Pathogenic variants in the ABCC6 gene are known to cause pseudoxanthoma elasticum, a connective tissue disorder characterized by claudication, gastrointestinal hemorrhage, ischemic stroke, or intracranial aneurysms [10]. ABCC6 variants were reported in 5.1% of early-onset Korean ischemic stroke patients in a recent report and they were the third most common Mendelian stroke variants reported in East Asian general populations [2,4]. However, the frequency was only 0.3 per 1,000 individuals in our analysis, partly due to the reclassification of the most common c.*38G>A variant as benign in this study.
In summary, the prevalence of pathogenic clinical variants associated with Mendelian stroke was 2.8% among 15,548 Korean individuals without neurological disorders. The overall prevalence rate of genetic variants associated with stroke seems to be high among the Korean general population, underscoring the importance for clinicians to be well-informed about genetic disorders that contribute to increased stroke risk.