Validation microsatellite path score in a population-based cohort of patients with colorectal cancer

Abstract

Purpose: Bethesda guidelines are used to recognize patients at risk for Lynch syndrome. However, obtaining personal and familial tumor data can sometimes be difficult. The Microsatellite Path Score (MsPath), a pathological score, based on age, tumor location, and pathologic features, has been developed to effectively predict colorectal cancer with DNA mismatch repair (MMR) deficiencies. However, the MsPath model's performance in an unselected, population-based colorectal cancer (CRC) population is unknown.

Patients and methods: We analyzed all patients with CRC regardless of age, personal or family history, and tumor characteristics from the EPICOLON study, an independent, prospective, multicenter, population-based cohort (N = 1,222). All patients underwent tumor microsatellite instability (MSI) analysis and immunostaining for MLH1/MSH2, and those with MMR underwent tumor BRAF mutation analysis and MLH1/MSH2 germline testing. All the pathologic features were centralized and evaluated blinded to the MMR status.

Results: MsPath score for prediction of having MSI high, with the recommended MsPath cutoff score ≥1.0, had a sensitivity, specificity, and positive predictive value (PPV) of 92.8% (95% CI, 86.9 to 98.3), 64.1% (95% CI, 61.1 to 66.8), and 15.8% (95% CI, 12.2 to 18.6), respectively. MsPath score had a sensitivity, specificity, and PPV of 81.8% (95% CI, 59.0 to 99.8), 60.6% (95% CI, 57.8 to 63.4), and 1.9% (95% CI, 0.7 to 3.1), respectively, for the identification of MLH1/MSH2 gene carriers. Application of the MsPath score, resulted in two (18%) of 11 mutation carriers being missed, both pathogenic germline MSH2 mutations.

Conclusion: In the general nonselected population, the MsPath score accurately predicted the probability of bearing a MSI high CRC, but it was insufficiently accurate to use for the selection of patients warranting MLH1/MSH2 mutation testing in the setting of Lynch syndrome.