Evaluating multiple next-generation sequencing derived tumor features to accurately predict DNA mismatch repair status
Walker,R.,Georgeson,P.,Mahmood,K.,Joo,J. E.,Makalic,E.,Clendenning,M.,Como,J.,Preston,S.,Joseland,S.,Pope,B.,Hutchinson,R.,Kasem,K.,Walsh,M.,Macrae,F.,Win,A.,Hopper,J.,Mouradov,D.,Gibbs,P.,Sieber,O.,O'Sullivan,D.,Brenner,D.,Gallinger,S.,Jenkins,M.,Rosty,C.,Winship,I.,Buchanan,D.
DOI: https://doi.org/10.1101/2022.06.20.22276419
2022-06-22
MedRxiv
Abstract:Identifying tumor DNA mismatch repair deficiency (dMMR) is important for precision medicine. We assessed tumor features, individually and in combination, in whole-exome sequenced (WES) colorectal cancers (CRCs) and in panel sequenced CRCs, endometrial cancers (ECs) and sebaceous skin tumors (SSTs) for their accuracy in detecting dMMR. CRCs (n=300) with WES, where MMR status was determined by immunohistochemistry, were assessed for microsatellite instability (MSMuTect, MANTIS, MSIseq, MSISensor), COSMIC tumor mutational signatures (TMS) and somatic mutation counts. A 10-fold cross-validation approach (100 repeats) evaluated the dMMR prediction accuracy for 1) individual features, 2) Lasso statistical model and 3) an additive feature combination approach. Panel sequenced tumors (29 CRCs, 22 ECs, 20 SSTs) were assessed for the top performing dMMR predicting features/models using these three approaches. For WES CRCs, 10 features provided >80% dMMR prediction accuracy, with MSMuTect, MSIseq, and MANTIS achieving [≥]99% accuracy. The Lasso model achieved 98.3%. The additive feature approach with [≥]3/6 of MSMuTect, MANTIS, MSIseq, MSISensor, INDEL count or TMS ID2+ID7 achieved 99.7% accuracy. For the panel sequenced tumors, the additive feature combination approach of [≥]3/6 achieved accuracies of 100%, 95.5% and 100%, for CRCs, ECs, and SSTs, respectively. The microsatellite instability calling tools performed well in WES CRCs, however, an approach combining tumor features may improve dMMR prediction in both WES and panel sequenced data across tissue types.