Case Report: Apparent progressive complete mismatch repair protein loss during platinum-based chemotherapy in metastatic rectal neuroendocrine tumour grade 2

Background Mismatch repair (MMR) deficiency is uncommon in gastroenteropancreatic neuroendocrine tumours (GEP-NETs), yet it represents a critical biomarker for immune checkpoint inhibitor (ICI) eligibility. Treatment-associated reduction or loss of MMR protein expression, often without corresponding MSI-H, has been reported in colorectal adenocarcinoma, but comparable changes have not been described in neuroendocrine tumours. Persistent discordance between immunohistochemistry (IHC)-based MMR protein assessment and next-generation sequencing (NGS)-based microsatellite instability (MSI) testing is an emerging and clinically challenging phenomenon. Case presentation We report a 55-year-old woman initially diagnosed with rectal NET Grade 1 on endoscopic biopsy; the tumour was upgraded to NET Grade 2 (Ki-67 approximately 20%) on the surgical resectionk specimen. Hepatic metastases were confirmed four months after resection. The first liver biopsy (January 16, 2026) demonstrated partial MMR protein loss by IHC (MLH1-, MSH2+, MSH6-, PMS2-), while concurrent NGS showed microsatellite stability (MSS) and a tumour mutational burden (TMB) of <1 mut/Mb. The patient received two cycles of surufatinib 250 mg orally once daily combined with intravenous carboplatin (500 mg, day 1) and etoposide (160 mg, days 1-3). Restaging CT demonstrated disease progression. A repeat liver biopsy (March 17, 2026) revealed complete loss of all four MMR proteins by IHC (MLH1-, MSH2-, MSH6-, PMS2-), while NGS again confirmed MSS and TMB <1 mut/Mb. Conclusions To the best of our knowledge, this is the first reported case of apparent progressive complete MMR protein loss in the context of sequential platinum-based chemotherapy in metastatic rectal NET Grade 2, with persistent IHC-NGS discordance throughout. These findings raise an important question regarding how IHC-defined dMMR should be interpreted in the absence of NGS-confirmed MSI-H, and highlight the complexity of immunotherapy eligibility decisions in rare tumour types. Prospective data are needed to guide clinical practice.