Large-scale application of ClinGen-InSiGHT APC-specific ACMG/AMP variant classification criteria leads to substantial reduction in VUS

Abstract

Pathogenic constitutional APC variants underlie familial adenomatous polyposis, the most common hereditary gastrointestinal polyposis syndrome. To improve variant classification and resolve the interpretative challenges of variants of uncertain significance (VUSs), APC-specific variant classification criteria were developed by the ClinGen-InSiGHT Hereditary Colorectal Cancer/Polyposis Variant Curation Expert Panel (VCEP) based on the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP). A streamlined algorithm using the APC-specific criteria was developed and applied to assess all APC variants in ClinVar and the International Society for Gastrointestinal Hereditary Tumours (InSiGHT) international reference APC Leiden Open Variation Database (LOVD) variant database, which included a total of 10,228 unique APC variants. Among the ClinVar and LOVD variants with an initial classification of (likely) benign or (likely) pathogenic, 94% and 96% remained in their original categories, respectively. In contrast, 41% ClinVar and 61% LOVD VUSs were reclassified into clinically meaningful classes, the vast majority as (likely) benign. The total number of VUSs was reduced by 37%. In 24 out of 37 (65%) promising APC variants that remained VUS despite evidence for pathogenicity, a data-mining-driven work-up allowed their reclassification as (likely) pathogenic. These results demonstrated that the application of APC-specific criteria substantially reduced the number of VUSs in ClinVar and LOVD. The study also demonstrated the feasibility of a systematic approach to variant classification in large datasets, which might serve as a generalizable model for other gene- or disease-specific variant interpretation initiatives. It also allowed for the prioritization of VUSs that will benefit from in-depth evidence collection. This subset of APC variants was approved by the VCEP and made publicly available through ClinVar and LOVD for widespread clinical use.

Keywords: ACMG/AMP variant classification guidelines; APC; Adenomatous polyposis coli; ClinGen; FAP; InSiGHT; familial adenomatous polyposis.

Graphical abstract

Figure 1

APC-specific criteria in brief The ACMG/AMP guidelines defined pathogenic (P) and benign (B) criteria encompassing evidence in population, experimental, computational, and clinical domains. The criteria are weighed and coded as benign stand-alone (BA), pathogenic very strong (PVS), strong (BS/PS), moderate (PM/BM), and supporting (PP/BP), the combination of which leads to a final classification of pathogenic (P), likely pathogenic (LP), variant of uncertain significance (VUS), likely benign (LB), or benign (B). This figure is only intended as a quick reference guide to the APC-specific criteria; the most updated version can be found at https://cspec.genome.network/cspec/ui/svi/doc/GN089. ^∗Details not shown here.

Figure 2

APC-specific criteria embedded in a reclassification algorithm An algorithm demonstrating the application of all eligible APC-specific codes to APC variants in ClinVar and the InSiGHT LOVD in a stepwise approach, and the percentage of variants that reached a B/LB or P/LP classification at each step. Firstly, the highest MAF of non-structural and structural variants was calculated from gnomAD non-cancer datasets or UK Biobank non-colorectal cancer control data, and gnomAD structural variants (SVs) or database of genomic variants (DGV) gold standard, respectively. Predictive criteria were then applied based on the most severe variant consequence as predicted by Ensemble variant effect predictor (VEP). Variants with any experimental and/or clinical evidence were identified and assigned corresponding code. Finally, splice variants at the same nucleotide and missense variants at the same codon were identified, and the variants at the same position criteria were applied.

Figure 3

Reclassification workflow and suggested method of operation for ongoing VCEP activity This workflow summarized the organizational procedures undertaken in this study for variant reclassification and prioritization of variants with high clinical importance for different modes of VCEP review and approval. Relatively straightforward variants might be processed in batch and become candidates for fast-track VCEP approval (e.g., variants fulfilling BA1 or BS1 plus BP1). From the remaining 37 VUSs with some evidence for pathogenicity, five variants were reassessed as P/LP by a targeted literature review and data mining (group 1). The remaining 32 variants were reassessed based on further clinical information requested from respective ClinVar submitters and VCEP members, which ultimately lead to prioritized VCEP review.

Figure 4

Classification of all APC variants in ClinVar, InSiGHT LOVD, and the combined dataset Each plot shows the classification change between the original (left) and revised classifications (right) for the APC variants in ClinVar (top left), InSiGHT LOVD (bottom left), and the combined dataset (right). The bottom table shows the number of APC variants (%) in the combined dataset (ClinVar and InSiGHT LOVD) and their original and revised classifications.

Figure 5

Classification of all APC variants in the original database (O) and their revised classification (R) by variant type Variants are broadly categorized into seven categories: 138 gross deletions, 44 gross duplications, 1,988 frameshift/nonsense, 399 splice site, 4,313 missense, 2,579 intronic/synonymous, and 767 other variants (120 in-frame, 631 UTR, and 16 other variants, which included start-loss, stop-loss, stop-retained, Alu and SVA retrotransposon insertions, inversions, and complex variants).