The aim of the current study was to assess the short and long-term outcome of total pancreatectomy (TP) for IPMN based on the largest series to date.
Literature data are scarce regarding TP for IPMN, though increasingly performed in this setting.
Data of 888 IPMN patients operated between 2004 and 2013 were collected in a multicentric retrospective AFC database. Ninety-three patients (10.5%) who had TP entered this study. Patient demographics, indications, intraoperative data, 3-month morbi-mortality (Clavien), and long-term outcome were analyzed.
Most patients had mixed type IPMN (59%) and underwent upfront (56%) or intraoperatively-decided (33%) TP. Morbidity and mortality rates were 47.3% and 4.3%, respectively, with no lethal hypoglycemia; morbidity was higher for intraoperatively-decided TP. Misdiagnoses were frequent regarding main pancreatic duct involvement (12%), invasiveness (33%), or mural nodules (50%), resulting in 12 TPs (13%) performed for asymptomatic IPMN showing only low/moderate dysplasia (LMD). On histopathological examination, there were 54 (58%) invasive IPMN (mostly pT3/T4 (76%), N+ (60%), R0 (75%)), with a significantly worse 5-year survival (21.2%) compared to noninvasive group (85.7%; P < 0.0001). In the former, 24 (58.5%) developed recurrence showing mostly distant metastasis, within 2 years in 92%.
This large series of TP for IPMN reported acceptable morbi-mortality rates with no long-term death from diabetes-related complication. Morphologic assessment was imperfectly reliable with 13% of TP done for LMD only. More than half of patients were operated at an invasive carcinoma stage with poor outcome. Conversely, long-term survival was excellent after TP for noninvasive IPMN.