Actual malignancy risk of either operated or non-operated presumed mucinous cystic neoplasms of the pancreas under surveillance

Background Despite the low malignant potential of pancreatic mucinous cystic neoplasms (MCNs), surgery is still performed. The aim of this pragmatic study was to assess the outcome of surgery and surveillance for patients presenting with a presumed MCN at the first evaluation. Methods Data for patients with a presumed MCN observed from 2000 to 2016 at the Verona Pancreas Institute and San Raffaele Hospital were extracted from prospective databases. The endpoints were risk of malignancy at pathology and rate of misdiagnosis for the surgical series, expressed as an odds ratio (OR) with 95 per cent confidence interval, and disease-specific survival (DSS) for the surveillance cohort investigated by the Kaplan–Meier method. Results A total of 424 patients were identified. In the surgical series (229 patients), the rate of misdiagnosis was 19.2 per cent. The rate of malignant MCNs was 10.9 per cent (25 patients). The overall rate of malignancy, including misdiagnoses, was 11.3 per cent (26 patients). Predictors of malignancy were mural nodules (OR 27.75, 95 per cent c.i. 4.44–173.61; P < 0.001), size at least 50 mm (OR 13.39, 2.01 to 89.47; P = 0.007), and carbohydrate antigen 19.9 level (OR 3.98, 1.19 to 13.30; P = 0.025). In the absence of mural nodules and enhancing walls, none of the resected presumed MCNs smaller than 50 mm were malignant. Only patients with high-risk stigmata undergoing surgery experienced a significantly reduced 5-year DSS compared with all other patients (88 versus 100 per cent; P = 0.031). Conclusion Presumed MCNs with mural nodules, enhancing walls or cysts of 50 mm or larger should be considered for upfront surgical resection owing to the high risk of malignancy. In the absence of these features, the incidence of malignancy is negligible, favouring surveillance in selected patients given the low risk of malignancy and the high rate of misdiagnosis.