Exploring the genetic basis of Lynch-like syndrome through paired germline and tumor exome sequencing.

Abstract

3592 Background: Lynch-like syndrome (LLS) is characterized by a diagnosis of mismatch repair deficient (dMMR) malignancy where somatic bi-allelic mutations in canonical MMR pathway genes ( MLH1, MSH2, MSH6, PMS2) have been identified as the main cause. Yet, a substantial proportion of cases remain unexplained by MMR somatic bi-allelic events or germline mutations. We hypothesize that LLS cases with young-onset cancers carry cryptic germline alterations in other pathways. To explore this contribution, we performed analyses of the germline and tumor mutation landscapes in LLS patients diagnosed with dMMR cancers. Methods: 18 probands with young-onset (age <50 years) dMMR colorectal or uterine cancers were selected from a familial cancer registry. The absence of deleterious germline MMR mutation and/or somatic MLH1inactivation was confirmed by standard clinical testing. We performed whole-exome sequencing (Illumina HiSeq) of germline (peripheral blood) DNA. Variant calls, quality-control, allele-frequency filtering (<1% in reference cohorts), and in silicoannotation were performed using the GATK and polyphen/SIFT tools. Pathway analysis was performed using the DAVID suite. For 16 of 18 patients, targeted exon sequencing of 408 cancer-related genes was performed on paired tumor/normal tissue samples (Ion Torrent AmpliSeq) and analyzed with VarScan 2. Results: 237,055 rare germline variants were detected in our cohort. We enriched a subset of 758 variants with putative frameshift (45.1%), stop gain or loss (25%), or splice site alteration (29.9%). Pathway analysis of genes altered by this subset revealed excess events in DNA damage repair (e.g. ERCC5, POLM, POLN, EXO5) and mRNA splicing (e.g. SCAF1, SRSF4) pathways. Preliminary analysis of somatic mutations profiles shows frequent alteration of known drivers including APC(64%) and NOTCH1(36%). Conclusions: Our exploratory analysis provides novel evidence that LLS patients may harbor an excess of deleterious germline mutations in DNA damage repair- and mRNA splicing-related genes. Future studies will identify genes which are targeted by both germline and somatic mutation with the goal of nominating putative causal genes. Defining additional mechanisms of dMMR in LLS cancers may help to refine prevention strategies for (un)affected individuals.