Pan-cancer long mononucleotide repeat microsatellite instability testing for detection and immunotherapy selection.

Abstract

2648 Background: Microsatellite instability (MSI) and mismatch repair deficiency (dMMR) are crucial biomarkers to inform potential for hereditary risk and eligibility for immunotherapy across multiple cancers. While current standard of care testing can identify most MSI-high (MSI-H)/dMMR patients, improved testing sensitivity for MSI-H status could identify more patients who may benefit from immunotherapy. Recent studies have indicated potential for more sensitive testing using a long mononucleotide repeat (LMR)-based multiplex MSI-detection assay. Methods: Patients were identified and consented as part of an IRB-approved protocol (UW15068) based on MSI status, MMR gene variants, or high tumor mutation burden (TMB). Patient tissues, normal and tumor, were formalin-fixed and slides were prepared for pathologist annotation and immunohistochemistry (IHC) for CD8. CD8+ tumor infiltrating lymphocytes (TILs) were quantified/high powered field (HPF). Slides were scraped for DNA isolation. MSI testing was performed on the resultant DNA using the LMR MSI Analysis System (Promega, Madison, WI) and the OncoMate MSI Dx Analysis system (OM, Promega) per manufacturer protocols. LMR MSI status and score were compared with other testing methods, TMB, CD8+ TILs/HPF, immunotherapy response, progression-free survival (PFS), and overall survival (OS). Results: A cohort of 202 patients (median age: 64 (range: 24-80+)) with 20+ different cancer types (lung 27.2%, colorectal 18.8%, skin 9%, esophageal 5.9%, uterine 5.9%) and across disease stages, mutation profiles, and immunotherapy treatment regimens were consented. At least one test identified a case as dMMR/MSI-H in 31.6% of cases. In 9.4% of cases, a discrepancy between testing methods was observed. LMR detected 9 cases as MSI-H that were not detected by clinical NGS or OM. 4 cases were detected as MSI-H by LMR and OM alone, one case by OM alone, and one case by LMR and clinical NGS alone. Additionally, 2 cases were dMMR by IHC but not LMR, OM, or clinical NGS and 2 cases were dMMR by IHC and MSI-H by clinical NGS, but not LMR or OM. LMR score positively correlated with TMB (r 2 = 0.57) for MSI-H, but not MSS patients (r 2 = 0.002). LMR score weakly correlated with CD8+ TILs/HPF (r 2 = 0.16) in MSI-H, but not MSS cases (r 2 <0.001). LMR MSI-H cases have a 52.6% complete response rate compared to 6.12% of LMR MSS subjects (p<0.001). Patients identified as MSI-H by LMR testing had an increased PFS (median PFS (days) MSI-H: 711, MSS: 178; hazard ratio (HR) 0.25 [95% CI 0.13-0.50; p<0.001]) and OS (median OS (days) MSI-H: 926, MSS: 402; HR 0.43 [95% CI 0.22-0.82; p=0.015]) compared to LMR MSS patients. Conclusions: Variability can be observed between dMMR/MSI testing methods indicating that multiple methods should be considered for each patient. LMR testing demonstrates promising sensitivity and correlation with immunotherapy efficacy.