Clinical characteristics and survival prediction of surgical patients with invasive pancreatic cystic neoplasm: a large retrospective study over two decades

Abstract

Purposes Invasive pancreatic cystic neoplasms (iPCNs) are an uncommon and biologically heterogeneous group of malignant neoplasms. We aimed to investigate the clinicopathological characteristics of iPCN patients and to develop nomograms for individual survival prediction after radical surgery. Methods Data of patients diagnosed with iPCN and pancreatic ductal adenocarcinoma (PDAC) between 2000 and 2018 from the SEER database were retrieved. The differences in clinical outcomes were evaluated using the Kaplan–Meier analysis. Nomograms were proposed based on the Cox regression model and internally validated by C-index, area under the curve (AUC) value, and calibration plot. Results A total of 7777 iPCN patients and 154,336 PDAC patients were enrolled. Most neoplasms were advanced, with 63.1% at stage IV. The 3-year overall survival (OS) and cancer-specific survival (CSS) rates in surgical patients were as follows: 45.7% and 50.1% for invasive intraductal papillary mucinous neoplasm (IPMN), 54.8% and 59.3% for invasive mucinous cystic neoplasm (MCN), 97.8% and 98.2% for invasive solid pseudopapillary neoplasm (SPN), 88.9% and 88.9% for invasive serous cystic neoplasm (SCN), and 27.3% and 30.5% for PDAC. Subgroup analyses showed no clinical benefit from chemotherapy or radiotherapy in lymph node-negative iPCN patients who underwent surgery. The following variables associated with OS and CSS were identified: age, race, chemotherapy, radiotherapy, histologic type, pathological grade, regional nodes examined, and T, N, and M stage. The nomograms had good discrimination and calibration by internal validation, with an AUC value of 0.800 for 3-year OS and 0.814 for 3-year CSS. Conclusion Our study showed that the prognosis of iPCN patients was significantly better than PDAC patients. The proposed nomograms demonstrated substantially better discrimination and calibration.