Quantitative approaches to variant classification increase the yield and precision of genetic testing in Mendelian diseases: the case of hypertrophic cardiomyopathy

Abstract

Background: International guidelines for variant interpretation in Mendelian disease set stringent criteria to report a variant as (likely) pathogenic, prioritising control of false-positive rate over test sensitivity and diagnostic yield. Genetic testing is also more likely informative in individuals with well-characterised variants from extensively studied European-ancestry populations. Inherited cardiomyopathies are relatively common Mendelian diseases that allow empirical calibration and assessment of this framework.

Methods: We compared rare variants in large hypertrophic cardiomyopathy (HCM) cohorts (up to 6179 cases) to reference populations to identify variant classes with high prior likelihoods of pathogenicity, as defined by etiological fraction (EF). We analysed the distribution of variants using a bespoke unsupervised clustering algorithm to identify gene regions in which variants are significantly clustered in cases.

Results: Analysis of variant distribution identified regions in which variants are significantly enriched in cases and variant location was a better discriminator of pathogenicity than generic computational functional prediction algorithms. Non-truncating variant classes with an EF ≥ 0.95 were identified in five established HCM genes. Applying this approach leads to an estimated 14-20% increase in cases with actionable HCM variants, i.e. variants classified as pathogenic/likely pathogenic that might be used for predictive testing in probands' relatives.

Conclusions: When found in a patient confirmed to have disease, novel variants in some genes and regions are empirically shown to have a sufficiently high probability of pathogenicity to support a "likely pathogenic" classification, even without additional segregation or functional data. This could increase the yield of high confidence actionable variants, consistent with the framework and recommendations of current guidelines. The techniques outlined offer a consistent and unbiased approach to variant interpretation for Mendelian disease genetic testing. We propose adaptations to ACMG/AMP guidelines to incorporate such evidence in a quantitative and transparent manner.

Keywords: ACMG/AMP guidelines; Hypertrophic cardiomyopathy; Mendelian genetics; Variant interpretation.

Fig. 1

The use of etiological fractions to evaluate variant classification methods. Illustration of how EFs can be used to evaluate methods for distinguishing pathogenic from benign variants (for a hypothetical gene). The overall EF of 0.85 [1] is based on a case frequency of 9.5% and a reference frequency of 1.5%. The aim of variant classification methods is to fully distinguish between pathogenic variants (producing an EF of 1.0 with frequency equal to case excess [2]) and benign variants (producing an EF of 0 with frequency equal to population reference, here ExAC [3]). We propose that an EF of 0.95 would be required to indicate a likely pathogenic variant

Fig. 2

Distribution of rare variants in HCM and ExAC cohorts for 6 genes with HCM clustering. Clustering analyses identify regions enriched for disease-associated variation, and therefore within which variants have a high likelihood of pathogenicity. For six HCM genes, the location of rare missense and single amino acid inframe indel variants found in cases (all variants regardless of clinical classification) and controls are shown alongside a cartoon of the cDNA structure. Darker grey indicates higher variant density (overlapping variants not plotted separately). Regions in which variants cluster significantly in cases are shown in red, and regions with clustering in population controls (ExAC) are shown in yellow. The HCM clusters detected were: MYH7 (residues 167–931), MYBPC3 (485–502, 1248–1266), TNNI3 (141–209), TNNT2 (79–179), MYL3 (143–180) and CSRP3 (44–71). For MYH7, existing functional annotations (as described in the “Discussion” section) are superimposed: In green, key residues of the converter kinetic domain and myosin mesa surface area enriched in disease-associated variants (Homburger et al. [37]); in blue, sites of inter- and intramolecular interaction between pairs of myosin heads (Alamo et al. [38]); and in grey, regions previously identified as constrained (intolerant of variation as evidenced by depletion of protein-altering variation in population controls), with the darker shades indicating higher constraint (Samocha et al. [36]). The coordinates describe amino-acid position within the canonical protein sequence

Fig. 3

Proposed adaptation of ACMG/AMP guidelines for rule PM1, relating to the relative frequencies of non-truncating variants in case cohorts and population controls

Fig. 4

Effect of EF-based approach to variant classification in HCM cohorts. a Proportion of cases from the OMGL/LMM HCM cohorts with variants in 8 sarcomeric genes (only rare variants, ExAC filtering frequency < 4 × 10⁻⁵, are shown, excluding non-essential splice site variants). Coloured shading represents the clinical classification of the original diagnostic laboratory (OMGL and LMM), and, for variants originally classified as VUS, the proportion that could be reclassified as Likely Pathogenic based on occurrence within a gene or region with EF ≥ 0.95. Eighty-nine variants in 123 cases for MYH7, 12 variants in 27 cases for MYBPC3, 18 variants in 34 cases for TNNI3, 15 variants in 18 cases for TNNT2 and 22 variants in 33 cases for TPM1 would be upgraded based on this analysis. b Proportion of cases in a prospective HCM cohort classified as actionable based on application of fixed and automatable ACMG/AMP rules, alongside the addition of manual curation of published evidence and the proposed EF-calibrated PM1 rules. Thirty-one extra cases (4.5%) are upgraded with EF-based rules compared to just 4 (0.6%) with manual curation. c Comparison of indexed LV mass in cases with pathogenic variants, VUS in high EF (≥ 0.95) regions, and VUS in low EF regions (< 0.95) in MYH7/MYBPC3 as well as genotype-negative cases, from the prospective HCM cohort. The clinical phenotype of individuals with VUS at locations anticipated to be pathogenic is indistinguishable from known pathogenic/likely pathogenic variants, while individuals with VUS in other regions have a clinical phenotype more similar to individuals without a sarcomere variant. d Kaplan-Meier survival curve for the overall composite endpoint (including mortality, ventricular arrhythmia and heart failure composites) of the SHaRe cardiomyopathy registry stratified by genotype (HCM cases with pathogenic variants, VUS in high EF region (≥ 0.95), VUS in lower EF regions (< 0.95), and genotype-negative cases)

Fig. 5

Distribution of rare variants in CPVT and ExAC cohorts for RYR2. All rare RYR2 non-truncating (missense and single amino acid inframe indel variants) variants in 1355 CPVT cases (well-phenotyped and referral) and ExAC are shown alongside a cartoon of the cDNA structure. Darker grey indicates higher variant density (overlapping variants not plotted separately). Three regions enriched for disease-associated variation were identified (shown in red)—residues 2138–2538, 3935–4196 and 4721–4959. Exons used in previously defined hotspot regions (original 41 exons and refined 21 exons) are highlighted as shown