RESUMEN
BACKGROUND: Reliability of mismatch repair proteins and microsatellite instability assessment is essential in order to define treatment strategy and identify candidates to immune checkpoint inhibitors in locally advanced gastroesophageal carcinoma. We evaluated the concordance of deficient mismatch repair (dMMR) and microsatellite instability-high (MSI-H) status between endoscopic biopsies and surgical specimens. METHODS: Consecutive patients with resectable gastric or gastroesophageal junction adenocarcinoma classified as MSI-H/dMMR by polymerase chain reaction (PCR) or immunohistochemistry (IHC) and operated at three referral Institutions were included. The primary endpoint was the rate of concordance between biopsy and surgical samples. If needed, central revision by IHC/PCR was performed by specialized pathologists from coordinating Institutions. RESULTS: Thirteen (19.7%) out of 66 patients showed discordant MSI-H/dMMR results in the original pathology reports. In most cases (11, 16.7%) this was due to the diagnosis of proficient mismatch repair status on biopsies. Among the ten cases available for central review, four were due to sample issues, four were reclassified as dMMR, one case showed dMMR status but was classified as microsatellite stable by PCR, one was linked to misdiagnosis of endoscopic biopsy by the local pathologist. Heterogeneity of mismatch repair proteins staining was observed in two cases. CONCLUSIONS: Available methods can lead to conflicting results in MSI-H/dMMR evaluation between endoscopic biopsies and surgical samples of gastroesophageal adenocarcinoma. Strategies aiming to improve the reliability of assessment should be primarily focused on the optimization of tissue collection and management during endoscopy and adequate training of dedicated gastrointestinal pathologists within the multidisciplinary team.