ABSTRACT
Most models for predicting malignant pancreatic intraductal papillary mucinous neoplasms were developed based on logistic regression (LR) analysis. Our study aimed to develop risk prediction models using machine learning (ML) and LR techniques and compare their performances. This was a multinational, multi-institutional, retrospective study. Clinical variables including age, sex, main duct diameter, cyst size, mural nodule, and tumour location were factors considered for model development (MD). After the division into a MD set and a test set (2:1), the best ML and LR models were developed by training with the MD set using a tenfold cross validation. The test area under the receiver operating curves (AUCs) of the two models were calculated using an independent test set. A total of 3,708 patients were included. The stacked ensemble algorithm in the ML model and variable combinations containing all variables in the LR model were the most chosen during 200 repetitions. After 200 repetitions, the mean AUCs of the ML and LR models were comparable (0.725 vs. 0.725). The performances of the ML and LR models were comparable. The LR model was more practical than ML counterpart, because of its convenience in clinical use and simple interpretability.
Subject(s)
Logistic Models , Machine Learning , Pancreatic Intraductal Neoplasms/pathology , Aged , Algorithms , Diagnosis, Computer-Assisted/methods , Female , Humans , Male , Middle Aged , Pancreatic Cyst/pathology , Pancreatic Intraductal Neoplasms/diagnostic imaging , Prognosis , Retrospective Studies , Risk FactorsABSTRACT
BACKGROUND: Unlike other neuroendocrine tumors of the gastrointestinal tract, management of duodenal and periampullary carcinoids remains controversial. We aimed to determine the metastatic potential and optimal choice of therapy for these neoplasms. METHODS: A retrospective review of all patients treated at the Johns Hopkins Hospital between 1996 and 2012 was conducted. Clinicopathologic factors associated with lymph nodal involvement and clinical outcomes were evaluated. RESULTS: A total of 101 patients were identified. Eighty (79.2%) tumors arose from the duodenum and 21 (20.8%) from the periampullary area. Thirty-five (34.7%) patients underwent pancreaticoduodenectomy (PD), 12 (11.9%) local resection, 38 (37.6%) endoscopic excision, and 16 (15.8%) patients harbored incidental tumors identified in the specimen after PD for another indication. Lymph node (LN) pathologic evaluation was done in 56 patients, among which 27 (48%) had positive LN. Specifically, LN positivity (LN+) for tumors <1 cm in size was 4.5% (1/22), for tumors 1-2 cm 72% (13/18), and for tumors >2 cm 81% (13/16). Tumor size was the only factor associated with LN+ (p = 0.029). CONCLUSION: Lymph nodal involvement is common for duodenal and periampullary carcinoid tumors, particularly among those >1 cm in size; therefore, resection with lymphadenectomyfor these larger tumors is recommended.
