Predictors of long-term survival for patients with pancreatic neuroendocrine neoplasm in FinPanNET: nationwide biobank study with histopathological re-evaluation

Pancreatic neuroendocrine tumours (panNETs) comprise a heterogeneous tumour group with unresolved optimal treatment. Surgery is recommended for symptomatic or >2 cm asymptomatic tumours1, due to the risk of major complications. Surveillance is considered safe for small non-functional panNETs (NF-panNETs)2, although there is a risk of metastasis. First, the critical question remains: Which tumours will develop metastases with poor survival outcomes? Second, the literature lacks nationwide panNET cohorts that include histopathological re-evaluation. A major challenge in retrospective cohorts collected over several decades is the lack of confirmed differentiation using immunohistochemical biomarkers and human variability in manual counting of mitotic rates. Grade modification rates, as high as 17–24%, have been reported after reanalysis3,4. Notably, digital image analysis for the Ki-67 proliferation index (PI) offers a high accuracy and has been endorsed previously. The aim of this study was to assemble a nationwide cohort of surgically treated panNET patients, in order to verify the diagnosis using modern immunohistochemistry and to compare clinicopathological characteristics with survival outcomes.