ObjectiveDiagnosis and treatment of gastroenteropancreatic high-grade neuroendocrine neoplasms (GEP-HG-NENs), particularly G3 well-differentiated neuroendocrine tumours (NETs) and poorly differentiated neuroendocrine carcinomas (NECs) relies on histopathological morphology, immunohistochemistry, and molecular biological markers, which are lacking especially in cases with ambiguous histomorphology. In this study to contribute to the development of more targeted treatment strategies, we examined various immunohistochemical and molecular biological markers and their association with clinicopathological features in GEP-HG-NENs.MethodsWe included 38 patients with GEP-HG-NENs in this study, with their retrospective follow-up data. The expression of tumour protein p53 (TP53), RB transcriptional corepressor 1 (RB1), somatostatin receptor 2 (SSTR2), clusterin (CLU), and marker of proliferation Ki-67 (MKI67) was immunohistochemically analysed. KRAS proto-oncogene, GTPase (KRAS) and B-Raf proto-oncogene, serine/threonine kinase (BRAF) V600E expression was evaluated using quantitative real-time polymerase chain reaction (qRT-PCR). The relationships between immunohistochemical and molecular biological markers and clinicopathological characteristics were examined using a Cox risk regression model, receiver operating characteristic (ROC) curve, and Kaplan–Meier survival analyses.ResultsSSTR2, RB, TP53, and CLU expression differed between NET G3 and NECs, with variations among the NET G3 and small- and large-cell NEC (SCNEC and LCNEC, respectively) groups (p < 0.05). The median MKI67 proliferative index was approximately 40% and 70% in G3 NETs and NECs, respectively. The NET G3 group exhibited a median survival of 25 months, indicating a relatively better prognosis than that of the NECs group (median survival, 11 months). Both Kaplan–Meier survival analysis and the Cox risk regression model indicated a statistical correlation among treatment methods, CLU expression, and prognosis (p < 0.05). The BRAF V600E mutation rate was 32.4% in G3 NETs and SCNEC, demonstrating a significant difference between both types (p = 0.0086). Furthermore, ROC curve analysis highlighted the diagnostic significance of the positive expression of the immunohistochemical markers CLU, SSTR2, and RB in identifying NET G3.ConclusionTo guide more suitable treatment strategies, it is essential to develop and apply valuable and more targeted immunohistochemical and molecular pathological markers for a comprehensive analysis.