ClinVar Genomic variation as it relates to human health

This variant is considered likely benign or benign based on one or more of the following criteria: it is a conservative change, it occurs at a poorly conserved position in the protein, it is predicted to be benign by multiple in silico algorithms, and/or has population frequency not consistent with disease.

PMS2: PP3, BP2, BS1

This variant was observed in the ICSL laboratory as part of a predisposition screen in an ostensibly healthy population. It had not been previously curated by ICSL or reported in the Human Gene Mutation Database (HGMD: prior to June 1st, 2018), and was therefore a candidate for classification through an automated scoring system. Utilizing variant allele frequency, disease prevalence and penetrance estimates, and inheritance mode, an automated score was calculated to assess if this variant is too frequent to cause the disease. Based on the score, this variant could not be ruled out of causing disease and therefore its association with disease required further investigation. A literature search was performed for the gene, cDNA change, and amino acid change (if applicable). No publications were found based on this search. This variant was therefore classified as a variant of unknown significance for this disease.

Variant summary: PMS2 c.86G>C (p.Gly29Ala) results in a non-conservative amino acid change located in the DNA mismatch repair protein family, N-terminal of the encoded protein sequence. Five of five in-silico tools predict a damaging effect of the variant on protein function. The variant allele was found at a frequency of 0.00061 in 245776 control chromosomes in the gnomAD database, including 1 homozygotes. The observed variant frequency is approximately 9 fold of the estimated maximal expected allele frequency for a pathogenic variant in PMS2 causing Hereditary Nonpolyposis Colorectal Cancer phenotype (7.1e-05), strongly suggesting that the variant is benign. c.86G>C has been reported in the literature in individuals affected with Hereditary Nonpolyposis Colorectal Cancer. These reports do not provide unequivocal conclusions about association of the variant with Hereditary Nonpolyposis Colorectal Cancer. Co-occurrence with a pathogenic variant has been reported (PMS2 exon 10 deletion, Goodenberger_2015), providing supporting evidence for a benign role. To our knowledge, no experimental evidence demonstrating an impact on protein function has been reported. Fourteen clinical diagnostic laboratories have submitted clinical-significance assessments for this variant to ClinVar after 2014 without evidence for independent evaluation. Multiple laboratories reported the variant with conflicting assessments (benign/likely benign n=9, VUS n=5). Based on the evidence outlined above, the variant was classified as benign.

The PMS2 c.86G>C;p.Gly29Ala variant has been published in the literature in reportedly healthy individuals (Bodian 2014, Johnston 2012) as well as individuals with colorectal cancer or breast cancer (Goodenberger 2016, Nikitin 2020, Tung 2016, Yurgelun 2017). Several of these individuals had other known pathogenic or risk factor variants in cancer susceptibility genes. This variant is also reported in ClinVar (Variation ID: 41721). This variant is found in the general population with an allele frequency of 0.057% (150/264,494 alleles, including 1 homozygote) in the non-cancer cohort of Genome Aggregation Database with a higher frequency observed in the Ashkenazi Jewish population (0.63%; 62/9,774 alleles). Although this frequency is higher than expected for a pathogenic variant, this variant was significantly associated with breast cancer in a recent case-control study (Nikitin 2020). The glycine at codon 29 is highly conserved, and computational analyses predict that this variant is deleterious (REVEL: 0.825). Due to conflicting information, the clinical significance of the p.Gly29Ala variant is uncertain at this time.

This alteration is classified as likely benign based on a combination of the following: seen in unaffected individuals, population frequency, intact protein function, lack of segregation with disease, co-occurrence, RNA analysis, in silico models, amino acid conservation, lack of disease association in case-control studies, and/or the mechanism of disease or impacted region is inconsistent with a known cause of pathogenicity.

Converted during submission to Uncertain significance.

The PMS2 p.Gly29Ala variant was identified 2 exome sequencing studies looking at cancer susceptibility genes in subjects of preterm birth or atherosclerotic phenotype, being observed in 5 of 2506 proband chromosomes (frequency 0.002) (Johnston 2012, Bodian 2014). The variant was also identified in dbSNP (ID: rs146176004) “With other allele”, ClinVar (with conflicting interpretations of pathogenicity, submitters: likely benign by GeneDx, Invitae, Ambry Genetics; uncertain significance by Illumina and Biesecker Lab/Human Development Section-NIH; and classification not provided by ITMI), Clinvitae (4x), and Insight Hereditary Tumors Database (3x); but was not identified in the Genesight-COGR, Cosmic, Insight Colon Cancer Gene Variant Database, Zhejiang Colon Cancer Database, Mismatch Repair Genes Variant Database. The variant was identified in control databases in 160 (1 homozygous) of 271376 chromosomes at a frequency of 0.0006 increasing the likelihood this could be a low frequency benign variant (Genome Aggregation Consortium Feb 27, 2017), identified in the following populations: African in 1 of 22874 chromosomes (frequency: 0.00004), Other in 9 of 6380 chromosomes (frequency: 0.001), Latino in 3 of 34310 chromosomes (frequency: 0.00009),European Non-Finnish in 63 (1 homozyogus) of 122742 chromosomes (frequency: 0.0005),European Finnish in 13 of 25778 chromosomes (frequency: 0.0005),Ashkenazi Jewish in 62 of 9996 chromosomes (frequency: 0.006), and South Asian in 9 of 30552 chromosomes (frequency: 0.0003). The p.Gly29 residue is conserved across mammals and other organisms, and computational analyses (PolyPhen-2, SIFT, AlignGVGD, BLOSUM, MutationTaster) suggest that the variant may impact the protein; however, this information is not predictive enough to assume pathogenicity. The variant occurs outside of the splicing consensus sequence and 2 of 5 in silico or computational prediction software programs (SpliceSiteFinder, MaxEntScan, NNSPLICE, GeneSplicer, HumanSpliceFinder) predict a greater than 10% difference in splicing; this is not very predictive of pathogenicity. In summary, based on the above information the clinical significance of this variant cannot be determined with certainty at this time. This variant is classified as a variant of uncertain significance.

This variant is classified as likely benign based on ACMG/AMP sequence variant interpretation guidelines (Richards et al. 2015 PMID: 25741868, with internal and published modifications).