http://orcid.org/0000-0002-3822-745X
Figures
Abstract
Introduction
Hypertrophic cardiomyopathy (HCM) is the most prevalent inherited heart disease. Next-generation sequencing (NGS) is the preferred genetic test, but the diagnostic value of screening for minor and candidate genes, and the role of copy number variants (CNVs) deserves further evaluation.
Methods
Three hundred and eighty-seven consecutive unrelated patients with HCM were screened for genetic variants in the 5 most frequent genes (MYBPC3, MYH7, TNNT2, TNNI3 and TPM1) using Sanger sequencing (N = 84) or NGS (N = 303). In the NGS cohort we analyzed 20 additional minor or candidate genes, and applied a proprietary bioinformatics algorithm for detecting CNVs. Additionally, the rate and classification of TTN variants in HCM were compared with 427 patients without structural heart disease.
Results
The percentage of patients with pathogenic/likely pathogenic (P/LP) variants in the main genes was 33.3%, without significant differences between the Sanger sequencing and NGS cohorts. The screening for 20 additional genes revealed LP variants in ACTC1, MYL2, MYL3, TNNC1, GLA and PRKAG2 in 12 patients. This approach resulted in more inconclusive tests (36.0% vs. 9.6%, p<0.001), mostly due to variants of unknown significance (VUS) in TTN. The detection rate of rare variants in TTN was not significantly different to that found in the group of patients without structural heart disease. In the NGS cohort, 4 patients (1.3%) had pathogenic CNVs: 2 deletions in MYBPC3 and 2 deletions involving the complete coding region of PLN.
Conclusions
A small percentage of HCM cases without point mutations in the 5 main genes are explained by P/LP variants in minor or candidate genes and CNVs. Screening for variants in TTN in HCM patients drastically increases the number of inconclusive tests, and shows a rate of VUS that is similar to patients without structural heart disease, suggesting that this gene should not be analyzed for clinical purposes in HCM.
Citation: Mademont-Soler I, Mates J, Yotti R, Espinosa MA, Pérez-Serra A, Fernandez-Avila AI, et al. (2017) Additional value of screening for minor genes and copy number variants in hypertrophic cardiomyopathy. PLoS ONE 12(8): e0181465. https://doi.org/10.1371/journal.pone.0181465
Editor: Chunhua Song, Pennsylvania State University, UNITED STATES
Received: March 22, 2017; Accepted: June 30, 2017; Published: August 3, 2017
Copyright: © 2017 Mademont-Soler et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Individual-level variant data is queryable from Cardiovascular Genetics Center under terms determined by Hospital General Universitario Gregorio Marañón and Institut d'Investigació Biomèdica de Girona for scientific collaboration with not-for profit entities. The data sharing would take place after acceptance of the conditions stipulated through a material transfer agreement. This was imposed by an Institutional Review Board. For more details, please contact Dr. Ramon Brugada (ramon@brugada.com) or Dr. Raquel Yotti (raquel.yotti@salud.madrid.org).
Funding: This work was supported by: ió “Obra social La Caixa”: Ramon Brugada; Instituto de Salud Carlos III (Fondo Investigación Sanitaria -FIS- (PI14/01773): Ramon Brugada; Sociedad Española de Cardiología (Proyecto Investigación Básica Cardiología 2015 de los Socios Estratégicos SEC): Oscar Campuzano; Instituto de Salud Carlos III (Fondo Investigación Sanitaria -FIS- (PI15/02222; BA16/00032): Raquel Yotti. The funder provided support in the form of salaries for authors CF-C, PA and SC, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.
Competing interests: I have read the journal's policy and the authors of this manuscript have the following competing interests: Dr. Ramon Brugada is consultant of Ferrer-inCode; and Dr. Ferrer-Costa, Dr. Patricia Álvarez and Dr. Sergio Castillo are employed by Gendiag.exe SL. This commercial affiliation does not alter our adherence to PLOS ONE policies on sharing data and materials. The other authors declare no conflicts of interest to disclose.
Introduction
Hypertrophic cardiomyopathy (HCM) is characterized by left ventricular hypertrophy with histologic features of cellular hypertrophy, myofibrillar disarray, and interstitial fibrosis. With a prevalence of 0.2% in the adult population, HCM is the most common inherited cardiac disease and a major cause of sudden cardiac death (SCD) in young people [1]. The disease has marked phenotypic variability, and clinical manifestations range from asymptomatic clinical course to severe heart failure and SCD. The identification of a disease-causing variant in a patient is crucial for diagnosis confirmation in borderline cases, early management of at-risk family members, genetic counseling and avoidance of unnecessary follow-up of non-carriers. The latter, besides doubtless clinical benefit, enables significant health-care costs saving [2–4]. For all these reasons, current guidelines recommend genetic testing in patients fulfilling diagnostic criteria for HCM, but the advantages of screening for genes without a definitive evidence of disease association versus more conservative approaches remain to be determined [5].
Overall, in patients fulfilling HCM diagnostic criteria, genetic testing leads to the identification of disease-causing genetic variants in 32–78.9% of cases, depending on the clinical characteristics of the patients, the number of genes studied, and the criteria used for variant classification [4, 6–19]. Most HCM cases are caused by mutations in genes that encode sarcomere proteins [19–21]. Among them, about 85% of pathogenic variants are found in MYBPC3 and MYH7, 10% in cardiac troponin T (TNNT2) and troponin I (TNNI3), up to 2% in TPM1, and less than 3% in other sarcomere genes (MYL2, MYL3, ACTC1 and TNNC1). For this reason, initial studies using Sanger sequencing in HCM recommended to focus on the 5 principal sarcomere genes [20]. Recent improvements in DNA-sequencing technologies offer the opportunity to screen for a larger number of genes in a time and cost-effective manner. However, this approach also results in an increase of the number of rare genetic variants of unknown significance (VUS), which may entail a clinical challenge.
The analysis of a predefined panel of HCM-related genes using Next-Generation Sequencing (NGS) technologies has emerged as the preferred genetic testing methodology for clinical purposes in HCM. This approach allows the additional screening for genes that have been previously proposed to cause a relatively small number of HCM cases (minor genes) and genes with a controversial role in the disease (candidate genes) [12, 19, 22]. Moreover, panels can easily include genes related to metabolic disorders that account for rare cases of unexplained left ventricular hypertrophy in adults (<5%) but whose identification is of great clinical relevance [5]. Finally, NGS enables the detection of alterations in the number of copies of large genomic regions, known as Copy Number Variants (CNVs). Recently, two large NGS series involving the screening for these variants in HCM-associated genes have shown that 0.56–0.8% of HCM cases may be explained by these large imbalances [23, 24].
The aim of the present study was to determine the prevalence and spectrum of clinically relevant genetic variants in a Spanish cohort of HCM patients and analyze the additional clinical value provided by the screening for minor and candidate HCM genes and CNVs using NGS. The value and clinical challenges derived from the screening for variants in TTN were specifically addressed and compared with an independent cohort of patients without structural heart disease.
Materials and methods
Study population
The study cohort includes 387 consecutive unrelated Spanish patients with clinical diagnosis of HCM according to current clinical criteria [5], referred for genetic testing between 2012 and 2016. The study was approved by the ethical committee of Hospital Universitari Dr. Josep Trueta de Girona (Spain) and conformed to the ethical guidelines of the Declaration of Helsinki 2008. Informed written consent was obtained from all subjects. Patients were recruited at the Inherited Heart Diseases Units from Hospital General Universitario Gregorio Marañón, Hospital Universitari Dr. Josep Trueta, Hospital Clínic de Barcelona and Hospital Universitario Puerta del Hierro. The mean age at the time of the genetic study was 48±20 years, 255 patients (65.9%) were men and 132 (34.1%) women. Family members of carriers of rare non-synonymous variants, indels and/or CNVs were invited to undergo genetic analysis. During the period of the study 180 relatives were referred for genetic testing and were studied by Sanger sequencing or MLPA.
The detection rate and classification of rare variants in TTN in the HCM cohort was compared with the results obtained in an independent group of 427 unrelated patients without echocardiographic evidence of structural heart disease (30 healthy subjects, 191 patients with Brugada syndrome, 138 with long QT syndrome, 20 with catecholaminergic polymorphic ventricular tachycardia, 8 with short QT syndrome, 9 with atrioventricular block, 7 with idiopathic ventricular tachycardia/ ventricular fibrillation, 5 with atrial fibrillation and 19 with other arrhythmias).
Genetic analysis
Total genomic DNA was isolated from blood or saliva samples using Chemagen MSM I (PerkinElmer, Germany). All patients were screened for the 5 more frequent sarcomere genes (MYBPC3, MYH7, TNNI3, TNNT2 and TPM1) (isoforms analyzed are listed in S1 Table). The first 84 patients underwent Sanger sequencing of these 5 genes and the remaining 303 patients were studied using expanded NGS panels.
Sanger sequencing.
The coding regions and exon-intron boundaries (±10bp) of the 5 main sarcomere genes were amplified by PCR and, after purification, the PCR products were directly sequenced in both directions using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, TX, USA). Sequencing products were run on 3130XL Genetic Analyzer (Applied Biosystems) and analyzed by means of SeqScape Software v2.5 (Life Technologies, CA, USA).
NGS.
Three hundred and three patients were analyzed with custom NGS panels (55 or 78 genes) including the coding regions and exon-intron boundaries (±10bp) of the most prevalent genes associated with inherited cardiac diseases (the 55-gene panel includes the UTR sequences of some genes). Coordinates of sequence data were based on UCSC human genome version hg19 (GRCh37). Both NGS panels were developed by Gendiag.exe S.L. and commercialized by Ferrer InCode as SudD inCode®.
Using either panel, we focused on the analysis of 25 genes previously associated or candidate for HCM. Reference sequence transcripts listed in S1 Table were analyzed. Only eight sarcomere genes definitively associated with HCM (MYBPC3, MYH7, TNNI3, TNNT2, TPM1, ACTC1, MYL2 and MYL3) and genes robustly associated with metabolic diseases that can mimic HCM (GLA, LAMP2 and PRKAG2) were considered validated genes [22]. The additional 14 genes (CSRP3, PLN, ACTN2, MYOZ2, MYH6, TNNC1, CAV3, JPH2, LDB3, RYR2, TCAP, VCL, PDLIM3 and TTN) were classified as candidate genes. Following the criteria described by Walsh et al. [22], these candidate genes have different levels of evidence for their association with HCM, with the exception of PDLIM3 and VCL (no supporting evidence), and RYR2 and TTN (evidence not analyzed). We analyzed separately the additional value and clinical challenges derived from the screening for variants in TTN, due to the limited evidence for its association with HCM and the high background variation of this gene.
Sample libraries were prepared following the SureSelect XT Target Enrichment System for Illumina Paired-End Sequencing Library protocol (Agilent Technologies, CA, USA). Indexed libraries were sequenced in ten-sample pools on a MiSeq platform (Illumina, CA, USA), with 2x75 bp reads length.
An algorithm developed by Gendiag.exe SL was used to process the FASTQ files to obtain clean BAM files for the subsequent analysis of both SNVs and indels. In brief, the processed raw reads obtained after sequencing were trimmed and mapped with BWA-MEM [25]. The output from mapping steps was joined and sorted, and only the uniquely and properly mapped read pairs were selected. Then the variant call was performed with SAMtools v.1.2 [26], together with an ad hoc developed script. Both the custom NGS gene panels and the bioinformatics algorithm used for the detection of SNVs and indels had been previously validated in our center, obtaining a sensitivity of 100% and a specificity of 99.5% (unpublished data). The identified SNVs and indels were annotated with dbSNP [27], Exome Sequencing Project (ESP) [28], 1000 Genomes Project [29], Exome Aggregation Consortium (ExAC) [30], Human Gene Mutation Database (HGMD) [31] and ClinVar [32]. Sanger sequencing was performed to sequence regions with coverage lower than 30X, as well as to validate the uncommon non-synonymous variants identified. Genetic variants were reported following the recommendations of the Human Genome Variation Society.
We used a bioinformatics algorithm developed in our laboratory to detect CNVs using NGS data. The approach focuses on capturing significant differences between the expected and observed normalized coverage for a given sample in every exon of the genes included in the NGS panels. Raw coverage is first normalized by the amount of DNA yielded for each sample in the run. Then the insert size and the low probe affinity bias for targeted regions with a too high or too low GC content (>75% and <45%, respectively) are corrected. Finally, the ratio between each sample and a built-in baseline is evaluated. If the ratio falls outside a signal-to-noise window and is greater or lower than the duplication or deletion cut-offs (0.45 and -0.8, respectively), the gain or loss is inferred. Each potential CNV was visually reviewed to discard possible false positives due to artefacts caused by samples with enrichment inconsistencies generated during the library preparation protocol. Sensitivity and specificity of the method were assessed in an independent cohort including 108 patients with different cardiovascular diseases (16 of them with known CNVs and the remaining without this type of rearrangements), and they were 100% and 90.7%, respectively (unpublished data).
Each CNV identified by NGS was validated by an alternative method (Multiplex Ligation-dependent Probe Amplification -MLPA- or quantitative PCR -qPCR-). MLPA analyses were performed using commercially available SALSA MLPA probemixes and following manufacturer’s instructions (MRC-Holland, Amsterdam, The Netherlands). After the multiplex PCR reaction, electrophoresis was performed using ABI3130XL Genetic Analyzer with LIZ500 size standard (both from Applied Biosystems), and results were analyzed using Coffalyser.Net (MRC-Holland). qPCR analyses were performed with the QuantStudio 7 Flex System using Power Up Sybr Green master mix (both from Life Technologies), following manufacturer’s recommendations. Results were analyzed with QuantStudio Real-Time PCR Software v1.2 (Life Technologies).
For precise characterization of the CNVs, the breakpoints were assessed using NGS split-read data when the breakpoint regions were covered by the NGS panels, and then they were confirmed by Sanger sequencing. When no split-read data were available, Sanger sequencing was performed in an attempt to characterize the breakpoints using primers located in the non-altered regions of the gene of interest.
Summary of NGS data results.
In the present study, including all MiSeq runs, the average call rate achieved at 30x with the custom enrichment gene designs of 55 and 78 genes was 99.7% and 99.8%, respectively. The median percentage of reads overlapping our target regions was 48% (range 39% to 51%) for the first panel and 66% (range 52% to 69%) for the second one. The median coverage per sample was 870 (721 to 1069) and 679 (479 to 867), respectively. The 25 and 75 percentiles were 571 and 1099 for the 55 gene design, and 509 and 843 for the 78 gene design.
Criteria for interpretation of SNVs, indels and CNVs
Rare variants (SNVs and indels) were defined as variants with a minor allele frequency (MAF) <0.002 [19, 33] in the databases dbSNP [27], ESP [28], 1000 Genomes Project [29] and Exome Aggregation Consortium (ExAC) [30]. We chose this conservative and inclusive cut-off to ensure the selection of all potentially relevant variants for the subsequent process of individual variant classification (see below). Additionally, to analyze the impact of the MAF filter applied on the final number and classification of genetic variants, we compared this approach with a more restrictive hard filtering recently proposed by Walsh et al., based on the frequency of the most common pathogenic HCM variant (MAF <0.0001 in ExAC) [22]. Both MAF criteria were also used to compare the detection rate of rare variants in TTN in HCM patients and individuals without a structural heart disease.
We used the updated American College of Medical Genetics and Genomics (ACMG) 2015 guidelines for variant interpretation to classify variants in 5 categories: pathogenic (P), likely pathogenic (LP), VUS, likely benign (LB) or benign (B) [34]. For the assessment of the clinical significance of previously reported variants we first searched for information in public variant databases, population cohorts and scientific literature. Then, available clinical, experimental and computational data were integrated with potential additional information obtained from the study of the particular family to reach a final clinical conclusion. The strength of the association with the disease at the gene-level was classified as strong, moderate, weak, only supported in functional data or no evidence [22].
Novel variants were defined as variants not previously reported in patients (published literature, HGMD [31] or ClinVar [32]) and absent from controls in ESP, 1000 Genomes Project [29], ExAC and Genome Aggregate Database [30]. Novel variants that did not meet strict ACMG criteria of pathogenicity (VUS) but exhibited at least one supportive criteria were denominated novel candidate variants. For this purpose we considered the following criteria: 1) location in a mutational hot spot and/or critical and well-established functional domain, 2) protein length changes as a result of in-frame deletions/insertions in a nonrepeat region, 3) missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before, 4) cosegregation with disease in ≥2 affected family members, or 5) multiple lines of computational evidence support a deleterious effect (probably or possibly damaging/deleterious/disease causing by three in silico prediction tools: PolyPhen-2 [35], Provean [36] and Mutation Taster [37]).
Finally, if the segregation study of a large family enabled the classification of a variant as LB (no segregation), such labeling was extrapolated to other index cases with the same genetic variant. In the description of TTN variants, we included their location in the main protein domains and the percentage spliced in (PSI) of the affected exon (estimation of the percentage of TTN transcripts that incorporate the mutation) [38], but we did not use this information to modify the variant classification.
Confirmed CNVs were first compared with published literature and databases HGMD [31], ClinVar [32], DECIPHER [39], Database of Genomic Variants [40] and ClinGen [41]. If the CNV (identical after precise characterization) had been previously robustly classified as P/LP or B/LB, the classification was extrapolated to our case. Novel CNVs were classified as pathogenic variants if: 1) it was a deletion in/of a gene where loss of function is a known mechanism of patient’s disease, 2) it was an intragenic in tandem duplication (not involving the last exon of the gene) in a gene where loss of function is a known mechanism of disease, or 3) it was a whole gene duplication in a gene for which triplosensitivity is known to cause patient’s disease.
Statistical analysis
Categorical variables were compared using the chi-square test and two-sided p values <0.05 were considered significant. Specifically, we compared the percentage of patients with rare variants observed after the screening for 25 genes with the percentage obtained when only the 5 main sarcomere genes were analyzed in the same cohort. We also compared the rate of P/LP variants found using these two different approaches. The same comparison was performed excluding TTN (set of 24 genes). Contingence tables were built to identify the number of patients with rare variants in TTN that also carried rare or P/LP variants in sarcomere genes. The role of the MAF hard filter on the final number and classification of genetic variants was analyzed comparing the proportions obtained using two different MAFs (<0.002 vs. <0.0001). We also used both MAF cut-offs to compare the detection rate of rare variants in TTN in patients with HCM and patients without structural heart disease. The statistical analysis was performed using R version 3.3.2.
Results
Genetic variants in main sarcomere genes
Overall, including both the Sanger sequencing and the NGS cohorts (n = 387 patients), we found 187 rare variants in the 5 principal sarcomere genes (MYBPC3, MYH7, TNNI3, TNNT2 and TPM1) in 269 patients (69.5%) (S1 and S2 Tables). After applying the ACMG criteria of causality, 135 variants were classified as P/LP (72.2%), 41 (21.9%) were considered VUS, and only 11 (5.9%) were LB. No significant differences in the percentage of rare, P/LP or novel variants in the five main sarcomere genes were observed between the Sanger and the NGS cohorts (split data are shown in Table 1). The percentage of patients with at least one P/LP variant (positive test) was 33.3%. We found 48 novel variants in these sarcomere genes. Among them, 35 were classified as P/LP (Table 2) and 1 as candidate novel variant.
The distribution of the 187 rare variants found in the 5 main sarcomere genes (pooled data from Sanger sequencing and NGS cohort, 387 patients) and their clinical classification is shown in Fig 1. We found 114 rare variants in MYBPC3, 48 in MYH7, 11 in TNNT2, 4 in TNNI3 and 10 in TPM1. As expected, most P/LP variants were found in MYBPC3 (64.2%) and MYH7 (27%). In accordance with a well-known loss of function mechanism for MYBPC3, most P/LP variants in this gene (60.2%) were radical variants, whereas missense variants were the most frequent variants among the other genes. Cascade genetic screening allowed us to establish the penetrance of 40 P/LP variants: 17 with complete penetrance and 23 with incomplete penetrance. One variant was de novo.
We identified several sarcomere variants previously associated with other cardiomyopathies (S2 Table). Among P/LP variants, MYH7_p.R249G has been previous described in association with left ventricular non-compaction cardiomyopathy (LVNC) [42], and MYH7_p.T1019N in association with both dilated cardiomyopathy (DCM) and HCM [43, 44]. Additionally, cascade screening for the pathogenic variant TNNT2_p.E163del found in one HCM index case showed that 1 of the 3 genotype-positive relatives had HCM but the other 2 had a LVNC phenotype.
Rare genetic variants identified in the NGS cohort
The screening for 25 genes in the NGS cohort showed 401 rare variants, but 3 variants were not confirmed by Sanger sequencing (false discovery rate = 0.75%). Therefore, 398 confirmed variants were reported in 231 patients. In comparison with the screening for only the 5 main sarcomere genes, the percentage of patients with non-benign variants was significantly higher (72.6% vs. 42.2%, p<0.001). Among them, 117 variants in 111 patients (36.6%) were classified as P/LP (Table 1). The proportion of patients with a positive test was not significantly different to the proportion found when screening for only the 5 main sarcomere genes (Table 1). The distribution of the 398 rare variants in the 25 genes in the NGS cohort and their clinical classification is shown in Fig 2. All genes except CSRP3, LDB3, MYOZ2 and PLN showed at least one rare variant. The classification of the novel variants identified in this cohort is shown in Fig 3.
In 99 patients who did not have rare variants in the 5 main sarcomere genes the screening for minor and candidate genes identified 250 rare variants. Among them, 12 patients carried P/LP variants. Most of these P/LP variants were found in validated minor sarcomere genes (ACTC1, MYL2 and MYL3) and genes related to metabolic diseases (GLA and PRKAG2) (Figs 2 and 4). The lack of definitive association with HCM implied that most rare variants found in candidate genes did not meet enough standardized ACMG criteria to be considered P/LP. However, at the individual variant level, we identified 2 unrelated HCM patients with the variant TNNC1_p.A8V, which has been previously reported in at least 7 additional unrelated HCM patients, is absent in control population (ExAC), and has functional studies supportive of a damaging effect. The findings of our cohort may help to increase the supportive evidence for enrichment of this infrequent variant in HCM cases.
The expanded genetic study identified 72 novel variants out of the 5 more frequent genes. On the other hand, the number of VUS drastically increased with the screening of 25 genes (23.6% vs. 61.1%, p<0.001). Most of the additional VUS were TTN variants (97.9% of them missense). Accordingly, the number of inconclusive tests also increased (9.6% vs. 36.0%, p<0.001) (Table 1).
Effect of MAF cut-off on the number and classification of rare variants.
Using a ExAC MAF <0.0001 we identified 308 variants in 205 patients (67.7%) from the NGS cohort. This cut-off filtered out 28 of the 39 LB variants and 59 of the 241 VUS with MAF<0.002. However, the rate of inconclusive studies did not significantly change (31.7% vs. 36.0%, p = 0.26). Three LP variants in 3 different patients were missed (MYBPC3_p.V771M in two patients and TNNT2_p.R278C in another one). Although there is enough supporting evidence to consider both variants as LP, published data suggest incomplete penetrance and a mild effect in isolation. Complete analysis with MAF <0.0001 is provided in S3 Table. The effect of the different MAF cut-offs on the number and distribution of variants according to the gene-level association with the disease is presented in Fig 4.
Identification of multiple variants: Compound and double heterozygotes.
Overall, considering only the 8 validated sarcomere genes and excluding LB variants, 14 patients carried two variants. In 6 cases we found 2 P/LP variants, and in 8 cases one P/LP variant in combination with one VUS. Six patients had two variants in MYBPC3 (1 was a compound heterozygote, while in the remaining cases we cannot discard the possibility of both variants being located in the same allele), 1 patient had two variants in MYH7, and 7 patients were double-heterozygous for validated sarcomere genes. In the NGS cohort, 105 patients had more than one non-benign variant. After excluding TTN from analysis, 38 patients had multiple non-benign variants.
Detection rate of TTN variants in HCM and comparison with patients without structural heart disease
TTN variant classification and filtering.
We found 163 rare variants in TTN in 117 patients, 49 novel and 114 previously described in clinical and/or population databases. After a careful revision of published data and family studies, 17 variants were classified as LB (15 previously reported and 2 novel variants). In particular, segregation studies allowed us to classify the variant as LB in 8 families. The remaining 146 variants in 108 patients were classified as VUS (28 patients had 2 additional VUS, and 5 had 3 VUS). Fifty-eight patients also carried variants in the main sarcomere genes, classified as P/LP in 40 cases. In an attempt to select the VUS in TTN with a higher potential clinical significance we focused on 20 novel candidate variants found in 18 patients in which three in silico tools consistently predicted a deleterious effect (Table 3). Fourteen of these variants were found in 13 patients who also carried variants in sarcomere genes (9 patients had a P/LP variant and 4 a VUS). Only 6 of these filtered TTN variants in 5 patients were found in cases without variants in the main sarcomere genes.
Comparison with patients without structural heart disease.
The number of HCM patients with non-synonymous rare variants in TTN (38.6% with MAF <0.002 and 29% with MAF <0.0001) was not significantly different to the group of patients without evidence of structural heart disease (39.3% with MAF <0.002, and 26% with MAF <0.0001) (p >0.3 for both case-control comparisons) (Table 4). The detection rate of VUS and novel variants was also similar (Table 4). A detailed description of the TTN variants analyzed in this cohort is provided in S4 Table.
With respect to the total number of TTN variants, filtering with MAF <0.002, the percentage of LB variants was higher in the group of patients without structural heart disease, and the relative percentage of novel variants was lower. However, considering only variants with MAF <0.0001, the percentage of novel variants was not significantly different. The location of the variants in the protein and the percentage of variants located in constitutive exons (PSI = 100) were similar in both groups when using MAF <0.0001 (Table 4). Applying the same step-wise algorithm in the non-structural cohort we found 17 novel VUS consistently predicted as deleterious, 3 of them non-sense variants located in the A-band, I-band and M-band, respectively (S4 Table). The detection rate of patients with this type of selected novel variants was not significantly different between HCM patients and patients without structural heart disease.
Identification, characterization and classification of CNVs
Screening for CNVs in our NGS cohort revealed that 4 out of the 303 patients had a validated CNV in one of the 25 genes analyzed (1.3%). Twelve additional signals were detected but they were not validated by MLPA or qPCR (false discovery rate = 75%). Among confirmed CNVs, 2 patients had deletions involving MYBPC3 gene, and 2 patients had a deletion of the entire coding region of the PLN gene. According to our criteria for interpretation of CNVs, all of them were considered pathogenic variants.
CNVs in MYBPC3.
One case (P168) had a deletion of the entire exon 27 (Fig 5. Panel A), and the other one (P259) had a deletion spanning from exon 4 to exon 12 (Fig 5. Panel B). Both deletions were confirmed by MLPA. No split-read data were available for the establishment of the rearrangement breakpoints of these cases, but both CNVs could be precisely characterized by Sanger sequencing: c.2737+148_2905+40del727insG for P168 and c.406+69_1091-1154del5654 for P259. None of the deletions has been previously described.
NGS results, schematic representation of the breakpoints and precise characterization by Sanger sequencing of (A) the deletion of exon 27 of MYBPC3 (P168, brown sample in the graph), (B) the deletion spanning from exon 4 to exon 12 of MYBPC3 (P259, turquoise sample in the graph), and (C) the well-characterized PLN deletion (blue sample in the graph). (D) NGS results are shown for the non-characterized PLN deletion (orange sample in the graph).
The case P168 also carried two previously reported variants in MYBPC3 (p.V771M –classified as LP–and p.A522T –classified as LB–), and a VUS in TTN. The case P259 also carried two missense variants in TTN: p.E2055K (classified as LB) and p.R25906C (classified as VUS). Cascade genetic testing in this family showed that the deletion from exon 4 to exon 12 cosegregated with the disease in one affected relative; this patient also carried the variant TTN_p.R25906C but not the variant TTN_p.E2055K (S2 Table).
CNVs in PLN.
In both cases, the deletions involved the entire coding region of the PLN gene and were confirmed by qPCR using the pair of primers 5’CTCAACAAGCACGTCAAAAGC3’ and 5’GCATCACGATGATACAGATCAGC3’. None of the patients had any uncommon SNP or indel. The first patient had a deletion of 7936bp, involving a portion of the first intron and the second exon of the gene (Fig 5. Panel C). The deletion was identified with the 55-gene panel, which included the UTR regions for this gene. The breakpoints could be identified using NGS split-read data, and were confirmed by Sanger sequencing. The precise description of the rearrangement was c.1-7587_159+190del7936. Such rearrangement has not been previously described. The second deletion of PLN was detected with the 78-gene panel, which does not include the UTR regions (Fig 5. Panel D). The rearrangement could not be further characterized using neither split-read data nor Sanger sequencing, but most probably is different from the one diagnosed in the other patient, as widening the coding region coordinates 500bp upstream and downstream the deletion signal was still present, and Sanger sequencing with the same pair of primers did not amplify any fragment. Overlapping rearrangements have not been previously described in HGMD, but there are several overlapping deletions described in DGV [40], DECIPHER [39], ClinGen [41] and ClinVar [32]. The variant found in DGV is a deletion of 58Kb involving the genes CEP85L and PLN, and the variants found in the other databases are larger and have been found in patients with congenital abnormalities.
Discussion
In this study, we report the results of the genetic screening of 387 unrelated Spanish patients clinically diagnosed with HCM. Using NGS, we focus on the additional diagnostic value of screening for minor and candidate genes for HCM, and expose a comprehensive study of CNVs. Our data show that screening for these genes and CNVs in HCM patients identifies the genetic cause of the disease in a small number of cases, but this approach does not increase the global detection rate. The screening for variants in TTN in HCM patients shows a high number of VUS and increases the rate of inconclusive test. In an independent cohort without structural heart disease, we have found a number and classification of rare variants in TTN similar to that found in HCM patients; adding evidence against the role of this gene in HCM.
Detection rate and clinical classification of rare variants
In the present study, the analysis of the most frequent sarcomere genes (MYBPC3, MYH7, TNNI3, TNNT2 and TPM1) in patients with HCM identified a potentially relevant variant in 42.2% of the patients. Applying recommended criteria for clinical classification, P/LP variants were identified in only one third of the patients. The screening for additional sarcomere genes and other known and putative HCM genes using a 25-gene NGS panel showed potentially relevant variants in about two thirds of the patients, but identified an additional P/LP mutation in a small number of patients, without significantly increasing the percentage of patients with a positive test. Although some initial studies using Sanger sequencing reported a detection rate of pathogenic variants of 63% [7, 20], most studies have reported rates below 50% [6, 8, 9, 11, 13–18]. The detection rate in recent NGS studies including different additional genes range from 32% to 78.9% [4, 12, 17, 19, 44–52]. This huge variation is most likely due to selection bias in these studies, differences in the clinical characteristics of the patients included and, importantly, the different criteria applied for the classification of genetic variants. As extensively recognized, NGS offers a high reliability [44, 47], so the sequencing process itself does not justify the differences in the reported yields.
Interestingly, the percentage of patients with P/LP variants found in our study is in agreement with the largest published NGS series in patients with HCM, which includes more than 2900 unrelated HCM patients (with genetic screening involving from 10 to 51 known or putative HCM genes) [4]. The criteria for the assessment of genetic variants used in their study was similar to the ACMG classification followed in the present study. It has been recognized that this strict classification may result in a larger proportion of variants being considered VUS [34], but the main purpose of this tool is to guide clinical decisions, which must be always based on strong supporting evidence. Nevertheless, we recognize that these restrictive clinical classifications do not fulfill the requirements of a research study to identify new disease-causing variants, especially when genes with a non-definitive association with the disease are included. For this reason, an additional effort to weight the evidence of pathogenicity during the assessment of novel candidate variants identified by NGS should be attempted.
Genetic spectrum of the disease.
In agreement with most series [4, 7–9, 14, 15, 17], our study shows that MYBPC3 is the gene with a higher proportion of P/LP variants, followed by MYH7. While MYH7 variants are almost exclusively missense (in our cohort all of them), MYBPC3 is characterized by a significant incidence of radical variants [3, 4]. Overall, considering only the 8 validated sarcomere genes, the 3.6% of our patients carried two non-benign variants, which is consistent with published data [17, 20]. The screening for additional candidate genes using NGS increased the proportion of carriers of multiple non-benign variants to 34.7%, mostly due to the existence of rare missense TTN variants. It has been demonstrated that the presence of multiple pathogenic variants in the 8 validated sarcomere genes may confer a more severe form of disease with a higher incidence of adverse outcomes including heart failure and sudden death [53]. However, the clinical significance of the presence of multiple variants in the remaining genes is unknown and this information should not currently be used for prognostic purposes.
Variants of unknown significance and role of TTN
While the screening for 25 genes provides a definitive diagnostic in particular cases without P/LP variants in the main sarcomere genes, the proportion of cases with VUS increases exponentially. These VUS represent nowadays a major clinical challenge, as proper genetic diagnosis and genetic counseling cannot be provided. In the assessment of VUS, the study of large pedigrees for segregation analyses and in vitro assays may provide useful information, but these studies are not possible or feasible in most cases.
Current available data are not enough to support pathogenicity of novel variants in candidate genes, but the absence in controls and the existence of consistent computational data supporting a deleterious effect should be taken into consideration to undertake segregation and/or functional studies. It is important to underlie that variants in candidate genes should not be used for clinical purposes, such as genetic counseling or cascade screening testing. However, reporting and carefully addressing them are necessary steps for the improvement of genetic diagnosis in HCM. Additionally, scientific literature and databases should be periodically searched for new information to reclassify these variants.
In the present study, the drastic increase of VUS in the NGS cohort is mainly due to the analysis of TTN, which is the gene with the largest coding sequence in the human genome. Whereas the pathogenic role of truncating variants in TTN has been demonstrated for DCM [54], the frequency of these radical variants in patients with HCM is similar to that found in control populations [54], and the pathogenic role of TTN missense variants is unknown [55]. In the present study we show that the number and classification of missense variants in TTN is not significantly different in patients with HCM and patients without structural heart disease. Even the rate of occurrence of novel variants that are consistently predicted as deleterious by in silico tools is not significantly different between both cohorts. Moreover, whereas a potential modifier role of selected missense TTN variants cannot be definitively ruled out, the finding of a high proportion of rare TTN variants in patients with variants in sarcomere genes increases the concerns about their actual pathogenic role. Altogether, these data may advice against the inclusion of this gene in clinical HCM panels.
Role of CNVs in HCM
To date, the role of CNVs in patients with HCM is a relatively unexplored field. The first CNV in a patient with HCM was reported in 1992 and consisted of a 2.4Kb deletion in MYH7, which was identified by Southern blotting, analyzing restriction fragment length polymorphisms [56]. A second patient with HCM and two CNVs in MYBPC3 was reported in 2009 [57]. Since then, few series studying CNVs in HCM patients have been published, and most studies have evaluated only 1 or 2 genes [58–63]. The search for single-exon deletions by long-range PCR in MYH7 in a cohort of 150 patients did not identify CNVs [58]. Three studies that performed MLPA of MYBPC3 (and in some cases TNNT2) reported a detection rate for CNVs of 0% (0/108) [59], 1% (1/100) [60] and 1.4% (1/72) [61]. Interestingly, the CNV identified in the last two studies was an identical MYBPC3 large deletion involving several exons (starting in the intron 27 and ending 485 bp after the MYBPC3 stop codon). Three other cases with CNVs in MYBPC3 have been reported, but detailed information of each case is not available [62, 63].
The first comprehensive study that searched for CNVs in multiple genes in a large group of HCM patients was published in 2015 by Lopes et al. [23]. They analyzed 19 HCM-related or candidate genes by NGS in 505 patients and detected 4 CNVs (0.8%): 1 deletion in MYBPC3, 1 deletion in PDLIM3, 1 duplication of the entire TNNT2 gene, and 1 duplication in LMNA. Deletions were considered LP variants, while duplications were considered VUS. Recently, Ceyhan-Birsoy et al. screened 708 HCM patients for CNVs using a NGS panel including 18 HCM-related (or putative) genes or 46 genes covering the full spectrum of cardiomyopathies, and detected CNVs in 4 of them (0.56%): a duplication in MYOZ2; a deletion in MYBPC3; a whole gene duplication of NEXN; and a whole gene duplication of GLA, LAMP2, EMD and TAZ (patient with trisomy X) [24]. Only the deletion of MYBPC3 was classified as pathogenic. In the present study, to further elucidate the role of CNVs in HCM we screened 303 HCM patients for CNVs in 25 genes associated with or candidate for HCM. Among them, we detected 4 CNVs (1.3% of our patients). Two CNVs were novel deletions in MYBPC3, one of them involving exon 27 and the other one ranging from exon 4 to exon 12. Both CNVs were classified as pathogenic variants, as radical variants in MYBPC3 are a well-known cause of HCM [4]. Interestingly, the first patient also harbored one LP variant and one LB variant in MYBPC3 (p.V771M and p.A522T, respectively), and a VUS in TTN. As no family members were available, we were not able to determinate if the MYBPC3 variants were located in the same allele. To the best of our knowledge, this kind of complex genotype in MYBPC3 has not been reported before. A previous study in 113 patients designed to search for CNVs in MYBPC3 in HCM patients carrying one pathogenic point variant did not identify any large rearrangement [61]. Our study demonstrates that even in patients with a LP variant in a main sarcomere gene, the screening for CNVs may add valuable information.
The other two rearrangements identified in our study were deletions of the entire coding region of PLN gene. Such PLN deletions were classified as pathogenic variants, because the patients only have a single functional copy of the gene, and reduction of the expression of PLN (due to nonsense and promoter pathogenic variants) has been previously associated with the development of HCM [64, 65]. These patients did not harbor any other P/LP variant that could explain the HCM phenotype.
The CNV prevalence in our cohort (1.3%) is not significantly different to that reported by Lopes et al. [23] (0.8%, p = 0.4630) and Ceyhan-Birsoy et al. (0.56%, p = 0.2144) [24]. These data suggest that large rearrangements explain a small number of cases that do not carry SNVs and indels, and in selected cases can be part of complex genotypes in combination with variants in sarcomere genes.
Limitations
Only the protein-coding and flanking intronic regions of known or putative HCM genes were analyzed, and some HCM cases may be explained by pathogenic variants in non-coding regions or other genes. In fact, during the enrollment period of this study new genes that were not included in the cardiomyopathy panels used, such as FLNC [66] or FHL1 [67], have been associated with HCM. Additionally, the inclusion in the HCM panels of genes related to other inherited diseases that involve left ventricular hypertrophy (i.e. Pompe disease, amyloidosis, mitochondrial cardiomyopathies or rasopathies) could also increase the diagnosis of some unexplained cases. The two steps process used for discovery and validation of CNVs showed a high false discovery rate, but this finding might be at least partially related to the inclusion of signals with low quality score in the validation step, in an attempt to minimize the rate of false negatives in a clinical scenario. Even using this low threshold, the number of CNV signals identified is small and their validation does not significantly impact the total cost of genetic testing in the whole sample.
Conclusion
Only a small percentage of HCM cases without point mutations in the 5 principal sarcomere genes are explained by pathogenic variants in minor or candidate genes for HCM or CNVs, but their identification is of major clinical relevance and can be easily performed by widely available NGS techniques. Screening for TTN in HCM patients drastically increases the number of inconclusive tests, and provides a rate of rare variants similar to that found in patients without structural heart disease, suggesting that this gene should not be analyzed for clinical purposes in HCM patients.
Supporting information
S1 Table. Isoforms analysed of the 25 known or candidate HCM genes included in the custom NGS panels.
https://doi.org/10.1371/journal.pone.0181465.s001
(XLSX)
S2 Table. Genetic variants with ExAC MAF <0.002 identifed in 387 consecutive unrelated Spanish patients with hypertrophic cardiomyopathy.
https://doi.org/10.1371/journal.pone.0181465.s002
(XLSX)
S3 Table. Rare variants (MAF <0.0001) in the 5 most frequent sarcomere genes, 25 genes associated with or candidates for HCM and 24 genes (same panel excluding TTN).
https://doi.org/10.1371/journal.pone.0181465.s003
(XLSX)
S4 Table. Nonsynonymous variants in TTN with ExAC MAF <0.002 identifed in 427 consecutive unrelated Spanish patients without structural heart disease.
https://doi.org/10.1371/journal.pone.0181465.s004
(XLSX)
Acknowledgments
The authors acknowledge Instituto de Salud Carlos III (ISCIII). The CIBERCV is an initiative of the ISCIII, Spanish Ministry of Economy and Competitiveness.
References
- 1. Maron BJ, Gardin JM, Flack JM, Gidding SS, Kurosaki TT, Bild DE. Prevalence of hypertrophic cardiomyopathy in a general population of young adults. Echocardiographic analysis of 4111 subjects in the CARDIA Study. Coronary Artery Risk Development in (Young) Adults. Circulation. 1995;92(4):785–9. pmid:7641357
- 2. Cobo-Marcos M, Cuenca S, Gamez Martinez JM, Bornstein B, Ripoll Vera T, Garcia-Pavia P. Usefulness of genetic testing for hypertrophic cardiomyopathy in real-world practice. Rev Esp Cardiol (Engl Ed). 2013;66(9):746–7.
- 3. Das KJ, Ingles J, Bagnall RD, Semsarian C. Determining pathogenicity of genetic variants in hypertrophic cardiomyopathy: importance of periodic reassessment. Genet Med. 2014;16(4):286–93. pmid:24113344
- 4. Alfares AA, Kelly MA, McDermott G, Funke BH, Lebo MS, Baxter SB, et al. Results of clinical genetic testing of 2,912 probands with hypertrophic cardiomyopathy: expanded panels offer limited additional sensitivity. Genet Med. 2015;17(11):880–8. pmid:25611685
- 5. Authors/Task Force m, Elliott PM, Anastasakis A, Borger MA, Borggrefe M, Cecchi F, et al. 2014 ESC Guidelines on diagnosis and management of hypertrophic cardiomyopathy: the Task Force for the Diagnosis and Management of Hypertrophic Cardiomyopathy of the European Society of Cardiology (ESC). Eur Heart J. 2014;35(39):2733–79. pmid:25173338
- 6. Van Driest SL, Ommen SR, Tajik AJ, Gersh BJ, Ackerman MJ. Sarcomeric genotyping in hypertrophic cardiomyopathy. Mayo Clin Proc. 2005;80(4):463–9. pmid:15819282
- 7. Olivotto I, Girolami F, Ackerman MJ, Nistri S, Bos JM, Zachara E, et al. Myofilament protein gene mutation screening and outcome of patients with hypertrophic cardiomyopathy. Mayo Clin Proc. 2008;83(6):630–8. pmid:18533079
- 8. Millat G, Bouvagnet P, Chevalier P, Dauphin C, Jouk PS, Da Costa A, et al. Prevalence and spectrum of mutations in a cohort of 192 unrelated patients with hypertrophic cardiomyopathy. Eur J Med Genet. 2010;53(5):261–7. pmid:20624503
- 9. Waldmuller S, Erdmann J, Binner P, Gelbrich G, Pankuweit S, Geier C, et al. Novel correlations between the genotype and the phenotype of hypertrophic and dilated cardiomyopathy: results from the German Competence Network Heart Failure. Eur J Heart Fail. 2011;13(11):1185–92. pmid:21750094
- 10. Garcia-Pavia P, Vazquez ME, Segovia J, Salas C, Avellana P, Gomez-Bueno M, et al. Genetic basis of end-stage hypertrophic cardiomyopathy. Eur J Heart Fail. 2011;13(11):1193–201. pmid:21896538
- 11. Gruner C, Ivanov J, Care M, Williams L, Moravsky G, Yang H, et al. Toronto hypertrophic cardiomyopathy genotype score for prediction of a positive genotype in hypertrophic cardiomyopathy. Circ Cardiovasc Genet. 2013;6(1):19–26. pmid:23239831
- 12. Lopes LR, Zekavati A, Syrris P, Hubank M, Giambartolomei C, Dalageorgou C, et al. Genetic complexity in hypertrophic cardiomyopathy revealed by high-throughput sequencing. J Med Genet. 2013;50(4):228–39. pmid:23396983
- 13. Marsiglia JD, Credidio FL, de Oliveira TG, Reis RF, Antunes Mde O, de Araujo AQ, et al. Screening of MYH7, MYBPC3, and TNNT2 genes in Brazilian patients with hypertrophic cardiomyopathy. Am Heart J. 2013;166(4):775–82. pmid:24093860
- 14. Kassem H, Azer RS, Saber-Ayad M, Moharem-Elgamal S, Magdy G, Elguindy A, et al. Early results of sarcomeric gene screening from the Egyptian National BA-HCM Program. J Cardiovasc Transl Res. 2013;6(1):65–80. pmid:23233322
- 15. Reguero JR, Gomez J, Martin M, Florez JP, Moris C, Iglesias S, et al. The G263X MYBPC3 mutation is a common and low-penetrant mutation for hypertrophic cardiomyopathy in the region of Asturias (Northern Spain). Int J Cardiol. 2013;168(4):4555–6. pmid:23870641
- 16. Kapplinger JD, Landstrom AP, Bos JM, Salisbury BA, Callis TE, Ackerman MJ. Distinguishing hypertrophic cardiomyopathy-associated mutations from background genetic noise. J Cardiovasc Transl Res. 2014;7(3):347–61. pmid:24510615
- 17. Li Z, Huang J, Zhao J, Chen C, Wang H, Ding H, et al. Rapid molecular genetic diagnosis of hypertrophic cardiomyopathy by semiconductor sequencing. J Transl Med. 2014;12:173. pmid:24938736
- 18. Berge KE, Leren TP. Genetics of hypertrophic cardiomyopathy in Norway. Clin Genet. 2014;86(4):355–60. pmid:24111713
- 19. Lopes LR, Syrris P, Guttmann OP, O'Mahony C, Tang HC, Dalageorgou C, et al. Novel genotype-phenotype associations demonstrated by high-throughput sequencing in patients with hypertrophic cardiomyopathy. Heart. 2015;101(4):294–301. pmid:25351510
- 20. Richard P, Charron P, Carrier L, Ledeuil C, Cheav T, Pichereau C, et al. Hypertrophic cardiomyopathy: distribution of disease genes, spectrum of mutations, and implications for a molecular diagnosis strategy. Circulation. 2003;107(17):2227–32. pmid:12707239
- 21. Veselka J, Anavekar NS, Charron P. Hypertrophic obstructive cardiomyopathy. Lancet. 2016. Epub 2016/12/04.
- 22. Walsh R, Buchan R, Wilk A, John S, Felkin LE, Thomson KL, et al. Defining the genetic architecture of hypertrophic cardiomyopathy: re-evaluating the role of non-sarcomeric genes. Eur Heart J. 2017. Epub 2017/01/14.
- 23. Lopes LR, Murphy C, Syrris P, Dalageorgou C, McKenna WJ, Elliott PM, et al. Use of high-throughput targeted exome-sequencing to screen for copy number variation in hypertrophic cardiomyopathy. Eur J Med Genet. 2015;58(11):611–6. pmid:26455666
- 24. Ceyhan-Birsoy O, Pugh TJ, Bowser MJ, Hynes E, Frisella AL, Mahanta LM, et al. Next generation sequencing-based copy number analysis reveals low prevalence of deletions and duplications in 46 genes associated with genetic cardiomyopathies. Mol Genet Genomic Med. 2016;4(2):143–51. pmid:27066507
- 25. Marco-Sola S, Sammeth M, Guigo R, Ribeca P. The GEM mapper: fast, accurate and versatile alignment by filtration. Nat Methods. 2012;9(12):1185–8. pmid:23103880
- 26. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25(16):2078–9. pmid:19505943
- 27. Sherry ST, Ward MH, Kholodov M, Baker J, Phan L, Smigielski EM, et al. dbSNP: the NCBI database of genetic variation. Nucleic Acids Res. 2001;29(1):308–11. pmid:11125122
- 28.
Exome Variant Server, NHLBI GO Exome Sequencing Project (ESP) [database on the Internet]. 2016. Available from: http://evs.gs.washington.edu/EVS/.
- 29. Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, Korbel JO, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68–74. pmid:26432245
- 30. Lek M, Karczewski KJ, Minikel EV, Samocha KE, Banks E, Fennell T, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536(7616):285–91. pmid:27535533
- 31.
The Human Gene Mutation Database [database on the Internet]. Institute of Medical Genetics in Cardiff. 2016. Available from: http://www.hgmd.cf.ac.uk/ac/index.php.
- 32. Landrum MJ, Lee JM, Riley GR, Jang W, Rubinstein WS, Church DM, et al. ClinVar: public archive of relationships among sequence variation and human phenotype. Nucleic Acids Res. 2014;42(Database issue):D980–5. pmid:24234437
- 33. Duzkale H, Shen J, McLaughlin H, Alfares A, Kelly MA, Pugh TJ, et al. A systematic approach to assessing the clinical significance of genetic variants. Clin Genet. 2013;84(5):453–63. pmid:24033266
- 34. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, 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. pmid:25741868
- 35. Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, Bork P, et al. A method and server for predicting damaging missense mutations. Nat Methods. 2010;7(4):248–9. pmid:20354512
- 36. Choi Y, Sims GE, Murphy S, Miller JR, Chan AP. Predicting the functional effect of amino acid substitutions and indels. PLoS One. 2012;7(10):e46688. pmid:23056405
- 37. Schwarz JM, Cooper DN, Schuelke M, Seelow D. MutationTaster2: mutation prediction for the deep-sequencing age. Nat Methods. 2014;11(4):361–2. pmid:24681721
- 38. Roberts AM, Ware JS, Herman DS, Schafer S, Baksi J, Bick AG, et al. Integrated allelic, transcriptional, and phenomic dissection of the cardiac effects of titin truncations in health and disease. Sci Transl Med. 2015;7(270):270ra6. pmid:25589632
- 39. Firth HV, Richards SM, Bevan AP, Clayton S, Corpas M, Rajan D, et al. DECIPHER: Database of Chromosomal Imbalance and Phenotype in Humans Using Ensembl Resources. Am J Hum Genet. 2009;84(4):524–33. pmid:19344873
- 40.
Database of Genomic Variants [database on the Internet]. 2016. Available from: http://dgv.tcag.ca/dgv/app/home.
- 41.
ClinGen: Clinical Genome Resource [database on the Internet]. 2016. Available from: http://clinicalgenome.org/.
- 42. Tian T, Wang J, Wang H, Sun K, Wang Y, Jia L, et al. A low prevalence of sarcomeric gene variants in a Chinese cohort with left ventricular non-compaction. Heart Vessels. 2015;30(2):258–64. pmid:24691700
- 43. Villard E, Duboscq-Bidot L, Charron P, Benaiche A, Conraads V, Sylvius N, et al. Mutation screening in dilated cardiomyopathy: prominent role of the beta myosin heavy chain gene. Eur Heart J. 2005;26(8):794–803. pmid:15769782
- 44. Gomez J, Reguero JR, Moris C, Martin M, Alvarez V, Alonso B, et al. Mutation analysis of the main hypertrophic cardiomyopathy genes using multiplex amplification and semiconductor next-generation sequencing. Circ J. 2014;78(12):2963–71. pmid:25342278
- 45. Meder B, Haas J, Keller A, Heid C, Just S, Borries A, et al. Targeted next-generation sequencing for the molecular genetic diagnostics of cardiomyopathies. Circ Cardiovasc Genet. 2011;4(2):110–22. pmid:21252143
- 46. Mook OR, Haagmans MA, Soucy JF, van de Meerakker JB, Baas F, Jakobs ME, et al. Targeted sequence capture and GS-FLX Titanium sequencing of 23 hypertrophic and dilated cardiomyopathy genes: implementation into diagnostics. J Med Genet. 2013;50(9):614–26. pmid:23785128
- 47. D'Argenio V, Frisso G, Precone V, Boccia A, Fienga A, Pacileo G, et al. DNA sequence capture and next-generation sequencing for the molecular diagnosis of genetic cardiomyopathies. J Mol Diagn. 2014;16(1):32–44. pmid:24183960
- 48. Glotov AS, Kazakov SV, Zhukova EA, Alexandrov AV, Glotov OS, Pakin VS, et al. Targeted next-generation sequencing (NGS) of nine candidate genes with custom AmpliSeq in patients and a cardiomyopathy risk group. Clin Chim Acta. 2015;446:132–40. pmid:25892673
- 49. Liu X, Jiang T, Piao C, Li X, Guo J, Zheng S, et al. Screening Mutations of MYBPC3 in 114 Unrelated Patients with Hypertrophic Cardiomyopathy by Targeted Capture and Next-generation Sequencing. Sci Rep. 2015;5:11411. pmid:26090888
- 50. Bottillo I, D'Angelantonio D, Caputo V, Paiardini A, Lipari M, De Bernardo C, et al. Molecular analysis of sarcomeric and non-sarcomeric genes in patients with hypertrophic cardiomyopathy. Gene. 2016;577(2):227–35. pmid:26656175
- 51. Cecconi M, Parodi MI, Formisano F, Spirito P, Autore C, Musumeci MB, et al. Targeted next-generation sequencing helps to decipher the genetic and phenotypic heterogeneity of hypertrophic cardiomyopathy. Int J Mol Med. 2016;38(4):1111–24. pmid:27600940
- 52. Rubattu S, Bozzao C, Pennacchini E, Pagannone E, Musumeci BM, Piane M, et al. A Next-Generation Sequencing Approach to Identify Gene Mutations in Early- and Late-Onset Hypertrophic Cardiomyopathy Patients of an Italian Cohort. Int J Mol Sci. 2016;17(8).
- 53. Maron BJ, Maron MS, Semsarian C. Double or compound sarcomere mutations in hypertrophic cardiomyopathy: a potential link to sudden death in the absence of conventional risk factors. Heart rhythm: the official journal of the Heart Rhythm Society. 2012;9(1):57–63.
- 54. Herman DS, Lam L, Taylor MR, Wang L, Teekakirikul P, Christodoulou D, et al. Truncations of titin causing dilated cardiomyopathy. N Engl J Med. 2012;366(7):619–28. pmid:22335739
- 55. Gigli M, Begay RL, Morea G, Graw SL, Sinagra G, Taylor MR, et al. A Review of the Giant Protein Titin in Clinical Molecular Diagnostics of Cardiomyopathies. Front Cardiovasc Med. 2016;3:21. pmid:27493940
- 56. Marian AJ, Yu QT, Mares A Jr., Hill R, Roberts R, Perryman MB. Detection of a new mutation in the beta-myosin heavy chain gene in an individual with hypertrophic cardiomyopathy. J Clin Invest. 1992;90(6):2156–65. pmid:1361491
- 57. Herman DS, Hovingh GK, Iartchouk O, Rehm HL, Kucherlapati R, Seidman JG, et al. Filter-based hybridization capture of subgenomes enables resequencing and copy-number detection. Nat Methods. 2009;6(7):507–10. pmid:19543287
- 58. Coto E, Reguero JR, Palacin M, Gomez J, Alonso B, Iglesias S, et al. Resequencing the whole MYH7 gene (including the intronic, promoter, and 3' UTR sequences) in hypertrophic cardiomyopathy. J Mol Diagn. 2012;14(5):518–24. pmid:22765922
- 59. Bagnall RD, Yeates L, Semsarian C. The role of large gene deletions and duplications in MYBPC3 and TNNT2 in patients with hypertrophic cardiomyopathy. Int J Cardiol. 2010;145(1):150–3. pmid:19666196
- 60. Chanavat V, Seronde MF, Bouvagnet P, Chevalier P, Rousson R, Millat G. Molecular characterization of a large MYBPC3 rearrangement in a cohort of 100 unrelated patients with hypertrophic cardiomyopathy. Eur J Med Genet. 2012;55(3):163–6. pmid:22314326
- 61. Pezzoli L, Sana ME, Ferrazzi P, Iascone M. A new mutational mechanism for hypertrophic cardiomyopathy. Gene. 2012;507(2):165–9. pmid:22820391
- 62. Jouven X, Hagege A, Charron P, Carrier L, Dubourg O, Langlard JM, et al. Relation between QT duration and maximal wall thickness in familial hypertrophic cardiomyopathy. Heart. 2002;88(2):153–7. pmid:12117842
- 63. Nannenberg EA, Michels M, Christiaans I, Majoor-Krakauer D, Hoedemaekers YM, van Tintelen JP, et al. Mortality risk of untreated myosin-binding protein C-related hypertrophic cardiomyopathy: insight into the natural history. J Am Coll Cardiol. 2011;58(23):2406–14. pmid:22115648
- 64. Haghighi K, Kolokathis F, Gramolini AO, Waggoner JR, Pater L, Lynch RA, et al. A mutation in the human phospholamban gene, deleting arginine 14, results in lethal, hereditary cardiomyopathy. Proc Natl Acad Sci U S A. 2006;103(5):1388–93. pmid:16432188
- 65. Medin M, Hermida-Prieto M, Monserrat L, Laredo R, Rodriguez-Rey JC, Fernandez X, et al. Mutational screening of phospholamban gene in hypertrophic and idiopathic dilated cardiomyopathy and functional study of the PLN -42 C>G mutation. Eur J Heart Fail. 2007;9(1):37–43. pmid:16829191
- 66. Valdes-Mas R, Gutierrez-Fernandez A, Gomez J, Coto E, Astudillo A, Puente DA, et al. Mutations in filamin C cause a new form of familial hypertrophic cardiomyopathy. Nat Commun. 2014;5:5326. pmid:25351925
- 67. Friedrich FW, Wilding BR, Reischmann S, Crocini C, Lang P, Charron P, et al. Evidence for FHL1 as a novel disease gene for isolated hypertrophic cardiomyopathy. Human molecular genetics. 2012;21(14):3237–54. pmid:22523091