PDF
Citation
Print
INTRODUCTION
Amyloidosis refers to a group of disorders characterized by the extracellular deposition of misfolded proteins (termed pathogenic amyloids) that cause organ-specific diseases [1]. The two primary systemic forms include light chain amyloidosis, resulting from tissue deposition of misfolded monoclonal immunoglobulin light chains, and transthyretin amyloidosis (ATTR), caused by misfolded transthyretin deposition [1, 2]. ATTR is further classified into wild-type transthyretin amyloidosis (ATTRwt) and hereditary transthyretin amyloidosis (ATTRv), based on the presence of variants in the transthyretin (TTR) gene [3, 4].
Congenital pathogenic variants (PVs) of TTR are associated with a specific genetic etiology of ATTRv, which is inherited in an autosomal-dominant manner [5]. ATTRv is a rare disease, primarily characterized by progressive cardiomyopathy and neuropathy, exhibiting varied clinical presentations depending on the specific variant and individual’s medical background [6, 7]. Owing to heterogeneous symptoms and similarities with other common diseases, including hypertrophic cardiomyopathy [8], ATTRv diagnosis often requires invasive procedures, such as cardiac biopsy, to confirm amyloid deposits [9, 10]. Consequently, diagnosis may be delayed until symptoms progress to severe stages, leading ATTRv to become potentially life-threatening without timely detection and intervention [11]. TTR analysis may function as a valuable screening tool to diagnose ATTRv and guide patient management.
Since 2007, the Transthyretin Amyloidosis Outcomes Survey (THAOS) has accumulated genetic and clinical data on patients with ATTRwt and ATTRv, including information from participating institutions in Korea [11]. Data from this registry highlight genetic and phenotypic heterogeneity and geographic differences; the V30M variant is predominantly identified in Europe, Asia, and South America, whereas ATTRwt is most frequently found in North America [11]. Substantial variation occurs even within geographic regions; for example, E89Q is the most prevalent variant in Italy and Bulgaria [12]. Clinical heterogeneity is also observed among individuals harboring the same variant; TTR V30M with early-onset ATTRv typically manifests with prominent neurological features, whereas late-onset cases more frequently exhibit cardiac involvement [13]. In Korea, D38A is the most prevalent variant among patients with ATTRv, representing a distinct genetic characteristic from other Asian countries [14–16]. Despite this known prevalence, comprehensive data on the distribution and clinical features of TTR variants in Korean patients with ATTRv remain limited. Here, we describe the characteristics of TTR variants in Korean patients with ATTRv.
MATERIALS AND METHODS
Study participants
We included 801 participants who underwent TTR analysis at Samsung Medical Center (Seoul, Korea) between April 2012 and October 2024. The participants were categorized into “in-house” and “externally referred” groups.
The in-house group was further subdivided into probands and those undergoing cascade testing (in-house relatives). Specifically, in-house probands included participants who underwent TTR analysis at our hospital owing to clinical manifestations associated with ATTRv. In-house relatives comprised participants related to probands harboring pathogenic (P) or likely pathogenic (LP) TTR variants, regardless of ATTRv-associated symptoms.
Externally referred participants were those whose blood samples were submitted to our laboratory from other institutions for TTR analysis. Based on clinical information provided in test request forms, we reclassified them as referred probands when at least one clinical manifestation suggestive of ATTRv was documented. Alternatively, the participants were reclassified as other referrals when the sample was drawn for a cascade testing (referred relatives) or when the indication was not specified. Participants were considered Korean, based on nationality or ethnicity data in hospital records, including their name, resident registration, medical records, and enrollment in the Korean National Health Insurance Service. Non-Korean participants were excluded from the study.
We comprehensively reviewed the medical records of in-house participants with TTR variants and collected demographic information, family history, and clinical data regarding organ involvement [heart, autonomic nervous system (ANS), peripheral nervous system (PNS), soft tissue, gastrointestinal (GI) tract, and kidneys]. Organ involvement was comprehensively evaluated based on criteria proposed by Gertz et al. [17] and Desport et al. [18], using tests available at our hospital in conjunction with clinical history-taking (Supplemental Data Table S1). This study was approved by the Institutional Review Board (IRB) of Samsung Medical Center (approval number 2024-12-033). Given the retrospective nature of the study, the IRB waived the requirement for informed consent.
TTR analysis
Genomic DNA was extracted from peripheral blood leukocytes using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). PCR amplification was performed using in-house primers (Supplemental Data Table S2), and direct sequencing was conducted with the ABI Prism 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) and the BigDye Terminator Cycle Sequencing Ready Kit (Applied Biosystems). Sequences thus obtained were compared with the TTR reference sequence (GRCh37/hg19, NM_000371.4). The pathogenicity of the identified variants was determined according to the 2015 American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines [19]. For in silico analyses, we employed the Rare Exome Variant Ensemble Learner (REVEL) [20] for identifying missense variants, SpliceAI [21] for splice variants, and Sorting Intolerant from Tolerant Indel (SIFT Indel [22]) and Protein Variation Effect Analyzer (PROVEAN) [23] for duplication variants. Variants classified as P, LP, or variants of uncertain significance (VUSs) were reported clinically, whereas benign or likely benign variants were excluded. Pedigrees were constructed to document the presence of reported variants in probands and to evaluate co-segregation with ATTRv-related clinical features.
Regarding TTR variant nomenclature, we followed the recommendations of the International Society of Amyloidosis Nomenclature Committee [24], which are based on mature protein levels. Nucleotide and amino acid changes according to the genetic reference sequence (NM_000371.4), including the signal peptide coding region, are also provided for variant descriptions.
Statistical analysis
To compare the allele frequencies of identified variants in patients with ATTRv with those in the general population, we utilized two population databases: Genome Aggregation Database (gnomAD v2.1.1) [25] and the Korean Reference Genome Database (KRGDB) [26]. Odds ratios were calculated by comparing the frequency of each variant in our patients with ATTRv with those in gnomAD (Korean subpopulation) and KRGDB. P was determined using Fisher’s exact test, and the confidence interval (CI) was calculated for each odds ratio. All statistical analyses were performed using R software, version 4.4.1 (R Foundation for Statistical Computing, Vienna, Austria).
RESULTS
Participant characteristics
The composition of participants who underwent TTR analysis is summarized in Fig. 1. In the overall cohort (N=801), 448 participants were male (55.9%), and the median age was 57 yrs (range, 9–93 yrs). The in-house group (N=313) contained 174 males (55.6%) with a median age of 52 yrs (9–89 yrs), comprising 165 in-house probands (102 males [61.8%]; median age, 62 yrs [16–87 yrs]) and 148 in-house relatives (72 males [48.6%]; median age, 42 yrs [9–89 yrs]). The externally referred group consisted of 488 participants, 274 of whom were male (56.1%), with a median age of 60 yrs (range, 13–93 yrs). Within this group, 96 were classified as referred probands [55 males (57.3%); median age, 71 yrs (16–93 yrs)] and 12 were classified as referred relatives (four males [33.3%]; median age, 42 yrs [27–68 yrs]). The remaining 380 externally referred participants had no documented indications for TTR analysis.
P/LP TTR variants were identified in 162 of the 801 participants (20.2%), including 36/165 in-house probands (21.8%), 75/148 in-house relatives (50.7%), 22/96 referred probands (22.9%), 10/12 referred relatives (83.3%), and 19/380 externally referred participants without a documented indication (5.0%). VUSs were identified in four additional participants: two in-house probands, one in-house relative, and one referred proband.
Genetic landscape of TTR variants in Korean participants
The distribution of P/LP TTR variants was dominated by D38A (87/162 patients; 53.7%), followed by M13dup (19; 11.7%); D38V (12; 7.4%); E89K (9; 5.6%); K35N (8; 4.9%); G47E and V122I (3 each; 1.9%); L12M, D18E, V30M, F44S, L55P, T59K, and T75I (2 each; 1.2%); and T40A, T40I, A45V, G47R, E54K, T60A, and I73V (1 each; 0.6%). The VUSs identified included H88N (2) and D99N (2). All identified variants were heterozygous.
In-house and referred probands
D38A was the most frequent variant among all probands with P/LP TTR variants (30/58; 51.7%), followed by M13dup (8/58; 13.8%) (Supplemental Data Fig. S1). Consistently, D38A remained the leading variant among the in-house probands (18/36; 50.0%), followed by M13dup and E89K (3 each; 8.3%) and T75I (2; 5.6%) (Fig. 2). Ten additional variants—L12M, D18E, K35N, D38V, T40A, T40I, F44S, G47R, T59K, and V122I—were each identified once (2.8%), as were the VUSs H88N and D99N. Among the referred probands, D38A was predominant (12/22; 54.5%), followed by M13dup (5; 22.7%) (Supplemental Data Fig. S1). Single occurrences (1/22; 4.5%) were observed for V30M, K35N, D38V, L55P, and I73V, and one VUS (D99N) was identified.
In-house and referred relatives
D38A was the most common variant among in-house relatives with P/LP TTR variants (38/75, 50.7%), followed by M13dup and D38V (10 each; 13.3%) (Supplemental Data Table S3). Additional variants included E89K (6; 8.0%), G47E (3; 4.0%), V122I (2; 2.7%), and single cases of L12M, D18E, V30M, K35N, F44S, and T59K (1.3% each). One asymptomatic in-house relative carried a VUS (H88N). Among the referred relatives, we detected K35N in five participants (50.0%), D38A in four (40.0%), and T60A in one (10.0%).
Clinical characteristics of in-house participants with P/LP TTR variants
Cardiac amyloid involvement was the most frequent manifestation in the in-house probands with ATTRv, observed in 97.2% (35/36) of those in that group. Involvement in other organs was also observed, as follows: PNS, 94.4% (34/36); ANS, 88.9% (32/36); soft tissue, 55.6% (20/36); GI tract, 41.7% (15/36); and kidneys, 5.6% (2/36) (Table 1). Among the 75 in-house relatives with P/LP TTR variants, 48 underwent at least one amyloidosis-related organ involvement workup. Of them, 34 (70.8%) showed signs of organ involvement, with the following distribution: ANS, 61.5% (24/39); heart, 52.1% (25/48); PNS, 48.7% (19/39); soft tissue, 27.3% (12/44); GI tract, 11.4% (5/44); and kidneys, 2.2% (1/46) (Supplemental Data Table S3).
Pathogenic M13dup variant in three Korean families
The M13dup variant was observed in three in-house probands (Table 2) and was recently reported by Jeon et al. [27]. Briefly, the first proband (SMC-P056) had a 20-yr history of cardiomegaly and developed indigestion, postural hypotension, and numbness 3 yrs before diagnosis. The second proband (SMC-P074) presented with acute heart failure, atrial fibrillation, and left ventricular hypertrophy. The last proband (SMC-P109) showed recurrent presyncope with left ventricular wall thickening and a restrictive filling pattern. All three probands exhibited transthyretin amyloid cardiomyopathy (ATTR-CMP) on cardiac imaging and biopsy. Neurological examinations revealed evidence of autonomic and peripheral nervous dysfunction.
TTR analysis was conducted for 12 family members of the probands with the M13dup variant, which identified 10 additional participants harboring M13dup variants (Fig. 3). Among them, eight underwent evaluations for organ involvement associated with amyloidosis, five of whom (62.5%) exhibited signs of organ involvement (Table 2). The ANS was most frequently affected (80.0%; 4/5), followed by soft tissue (40.0%; 2/5) and the heart and PNS (20.0%; 1/5 each).
Using the M13dup data from this study along with allele frequency data from public databases [gnomAD (2 of 3,818) and KRGDB (2 of 1,722)], the calculated odds ratios of M13dup for ATTRv were 37.5 (95% CI: 10.7–147.4, P<0.001) and 16.9 (95% CI: 4.8–66.6, P<0.001), respectively. In silico analysis predicted the M13dup variant to be damaging (SIFT Indel score: 0.894) and deleterious (PROVEAN score: −8.153). Considering the high odds ratio for ATTRv among M13dup carriers, co-segregation data within families, low allele frequency in public databases, and nature of the variant as an in-frame duplication, this variant was ultimately classified as a PV.
DISCUSSION
This comprehensive analysis of TTR variants in Korean patients with ATTRv reveals a unique genetic landscape dominated by the D38A and M13dup, differing markedly from variant distributions in other populations. Our findings represent the first systematic characterization of ATTRv genetics in Korea, including the reclassification of M13dup as a pathogenic variant, and offers insights into organ involvement patterns that have important implications for clinical practice.
Our findings helped identify D38A as the most frequent TTR variant among Korean patients with ATTRv, being present in 50.0% of the in-house probands. M13dup and E89K were the next most common, each with a frequency of 8.3%. When combined with data from the referred probands (D38A, 54.5%; M13dup, 22.7%), the D38A (51.7%) and M13dup (13.8%) variants emerged as the most common variants in Korean patients with ATTRv. These findings partially align with a recent nationwide ATTR-CMP study by Kim et al., who identified E89K (9.8%) and M13dup (8.2%) as the most frequent variants in Koreans, following D38A (55.7%) [16]. Among east Asian countries, G83R and V30M/A were identified as the most common variants in China, whereas V30M was the predominant variant in Japan [28, 29]. Despite the geographical proximity of these countries to Korea and their shared genetic characteristics, the TTR genetic variant distribution in Korea differs markedly from those in neighboring populations.
Currently, over 140 TTR variants have been identified in patients with ATTRv [30]. According to data in the THAOS registry, the most common variants globally are V30M and V122I [11]. However, V30M was not observed in the in-house probands in this study and was detected at very low frequencies in other studies on Korean patients; for instance, Kim et al. [16] detected 1.6%, whereas Choi et al. [15] did not detect any instances of V30M. Conversely, we observed the D38A variant, accounting for only 0.6% of cases in the THAOS registry [11], in approximately half of all Korean patients with ATTRv, highlighting the unique genetic characteristics of the Korean population. Considering that most TTR amino acid sequences are highly conserved [31], this distinct variant distribution may reflect a founder effect. Such an effect has been previously demonstrated for the E89Q variant in a Bulgarian population, where over 75% of affected individuals share a common haplotype and ancestry, whereas the predominance of the V30M variant in other European countries is suggestive of distinct regional origins and endemic clustering [32].
Notably, M13dup was classified as a newly identified disease-causing variant, according to the 2015 ACMG/AMP guidelines. Although M13dup has been sporadically reported in Korean patients with ATTR [27, 33, 34], its pathogenicity has not been defined owing to a lack of functional and clinical evidence. The results of this study, including cascade testing and comprehensive evaluation of M13dup carriers, provide evidence of co-segregation with ATTRv and a higher odds ratio than that of the general population, based on data in the gnomAD and KRGDB repositories. In light of these findings, we reclassified M13dup as a rare duplication type of PV. In addition to M13dup, Klimtchuk et al. [35] reported data on the E51_S52dup variant suggesting that although most pathogenic TTR variants are missense, in-frame insertions or duplications should also be considered as potential causative variants of ATTRv. However, sufficient evidence is currently lacking for both VUSs identified in this study (H88N and D99N) to establish their pathogenicity. Nonetheless, considering that these variants involve highly conserved amino acids in TTR, they may be reclassified as P/LP variants in the future upon identification in additional patients.
The clinical presentation of ATTRv is highly variable, ranging from multiple neuropathies to cardiac disorders, which contributes to diagnostic delays [36]. Some TTR variants have been studied for their association with specific affected organs, and efforts are being made to perform organ involvement workups based on TTR analysis results to screen for potential complications and facilitate early intervention in affected carriers [6]. For example, the V122I variant has been associated with involvement of the heart, PNS, and soft tissue, whereas the V30M and T60A variants can affect the heart, ANS, and PNS [37]. All in-house probands with D38A exhibited cardiac manifestations, and 88.9% of the probands (16/18) showed signs of both ANS and PNS involvement, consistent with findings reported by Choi et al. [15], where all eight patients with D38A exhibited heart and ANS involvement. Soft tissue and GI tract involvement were observed in 50.0% (9/18) and 38.9% (7/18) of probands with the D38A variant, respectively. All probands with M13dup demonstrated involvement of the heart, ANS, and PNS, along with carpal tunnel syndrome. As more data are accumulated, the phenotypic characteristics associated with specific variants will become clearer.
Among the 48 carriers with P/LP TTR variants identified via cascade testing who underwent organ involvement workup, 34 (70.8%) showed signs of organ involvement. Although cardiac involvement was predominant among the in-house probands, ANS involvement was more frequent among the in-house relatives. The most common variant in Koreans, D38A, was also associated with ANS involvement in 66.7% (14/21), PNS involvement in 61.9% (13/21), and heart involvement in 57.7% (15/26) of in-house relatives. Similarly, V30M, the most common variant worldwide, has been reported to predominantly cause polyneuropathy in early-onset cases, whereas late-onset cases often involve cardiac manifestations [38]. M13dup carriers in our cascade testing also showed a tendency for ANS involvement (66.7%; 4/6) to precede cardiac manifestations (12.5%; 1/8). Our findings suggest that D38A and M13dup carriers tend to present neurological manifestations earlier than cardiac symptoms. Regular monitoring of neurological manifestations is essential for carriers with P/LP TTR variants, as cardiac manifestations may develop later.
This study has some limitations. First, owing to the retrospective study design, we relied solely on medical chart reviews obtained at the time of diagnosis, without conducting a longitudinal investigation that included the clinical course and survival outcomes. Second, in cascade testing, evaluations of organ involvement were often performed selectively based on clinical findings rather than via standardized evaluations, which may have introduced bias into the data regarding organ involvement. Standardized assessments are required for P/LP TTR variant carriers to ensure accurate evaluation and disease management. Third, for externally referred participants, the status of the probands was determined based solely on the information provided in the test request form. This approach could have included participants referred for cascade testing rather than true probands, and the detailed clinical features of these referred probands could not be thoroughly reviewed; therefore, the frequencies observed in the referred probands should only be used as a validation set to confirm the findings from the in-house probands. Lastly, although the M13dup variant was classified as a PV according to the 2015 ACMG/AMP guidelines, this study did not involve functional studies to directly evaluate the impact of this variant at the mRNA- or protein-expression levels.
In summary, we delineated the genetic profile of ATTRv in Koreans using the largest single-institution dataset with systematic organ assessment to date. The predominance of D38A, together with the frequent detection of M13dup, highlights a variant spectrum that differs from those in other populations and provides a nationally representative reference for clinical practice and future research in Korea. Characterizing this distinct variant spectrum not only provides a valuable reference for clinicians but also supports the development of population-specific screening strategies, guides early diagnostic approaches, and may inform national policies for ATTRv management, ultimately improving patient outcomes in Korea.
XML Download