Abstract
The diagnosis of high-grade gastroenteropancreatic neuroendocrine neoplasms (HG-GEP NENs) primarily relies on histopathological differentiation and mitotic count assessment. Due to interobserver variability in morphological evaluation, reliable molecular diagnostic biomarkers are needed to support accurate classification, particularly in challenging cases where the distinction between well-differentiated neuroendocrine tumours grade 3 (NET G3) and poorly differentiated neuroendocrine carcinoma (NEC) is ambiguous. The present study aimed to identify clinicopathological markers that facilitate the differential diagnosis of HG-GEP NENs. A total of 34 patients with HG-GEP NENs were included in study. Integrated bioinformatics analysis, including protein-protein interaction network construction from the GSE211485 dataset and subsequent validation using The Cancer Genome Atlas data, identified checkpoint kinase 1 (CHEK1) as a potential molecular marker for diagnosing HG-GEP NENs, which also showed prognostic significance in digestive system tumours. Despite no significant difference in overall CHEK1 DNA levels between groups, high CHEK1 expression was significantly more prevalent in the NEC group than in the NET G3 group (P=0.0113), with the small cell NEC (SCNEC) subgroup exhibiting the highest frequency (P=0.0075). Receiver operating characteristic curve analysis results revealed that high CHEK1 expression distinguished NEC from NET G3 [area under the curve (AUC)=0.8029]. When further stratified, its diagnostic performance was more pronounced for SCNEC (AUC=0.8708) than for large cell NEC (AUC=0.7102). These findings suggest that CHEK1 may serve as a potential molecular biomarker for the differential diagnosis of HG-GEP NENs. Although further large-scale clinicopathological studies are needed, CHEK1 expression demonstrates diagnostic potential and could be utilised to inform standard treatment plans.