BACKGROUND:Interpretation of recurrence following resection of intraductal papillary mucinous neoplasms (IPMNs) is hindered by inconsistent terminology and outcome reporting. A distinction between metachronous/de novo IPMN development, progression of residual disease, and true recurrence of invasive entities is rarely considered. METHODS:A systematic review and meta-analysis were conducted following PRISMA and Cochrane guidelines (PROSPERO-ID: 1009399). The primary endpoint was postoperative recurrence of non-invasive IPMN and IPMN-derived pancreatic cancer (PC). Recurrence following resection of non-invasive IPMN was reclassified as progression of persistent cysts or de novo metachronous IPMN. Secondary endpoints comprised risk-factors for recurrence. RESULTS:Sixty-six articles with 11 464 patients were included. After a median follow-up of 26 (interquartile range (i.q.r.) 18.0-53.0) to 72 (i.q.r. 5-318) months, the recurrence rate of IPMN-derived PC was 41.9% (1646/3925 patients), with a 5-year pooled recurrence-free survival of 46.6%. Systemic recurrence was most common (1034/1646; 62.8%), followed by locoregional (430/1646; 26.1%). Secondary treatments were administered in 655/1646 patients (39.8%) presenting with recurrence and included chemotherapy (65.5%), surgery (22.1%) and radiation (5.8%). Lymph node involvement (hazard ratio 2.87, 95% confidence interval 1.51 to 5.43) and tubular subtype (hazard ratio 1.68, 1.16 to 2.43) were identified as independent predictors of recurrence-free survival. The overall recurrence rate following pancreatic resection of non-invasive IPMN was 11.2% (831/7446), after median follow-up of 28 (i.q.r. 1-153) to 114 (i.q.r. 12-204) months. Among these, 408 (49.1%) developed as de novo lesions in the remnant pancreas and 174 (20.9%) as progression of pre-existing cyst detected at the time of index surgery; the remaining 249 patients (30.0%) could not be reclassified. Non-invasive recurrence was more common (308; 37.1%) than IPMN-derived PC (118; 14.2%), whereas the type of recurrence was unspecified in 405 patients (48.7%). Secondary treatment data were available for 265 patients, of whom 131 (49.4%) underwent reoperation. CONCLUSION:The recurrence rate of IPMN-derived PC is high and warrants close surveillance policies. The majority of 'recurring' non-invasive IPMNs are de novo lesions. Standardized reporting, distinguishing true recurrence from de novo development and progression of residual disease, is essential to stratify recurrence risk and optimize surveillance protocols accurately.
原文链接
分享
引用
