Parenchyma-sparing, local pancreatic head resection for premalignant and low-malignant neoplasms - A systematic review and meta-analysis.

Background: Parenchyma-sparing, local pancreatic head resection, but not pancreaticoduodenectomy (PD) preserves tissue and maintains the pancreatic metabolic functions.

Methods: PubMed/Medline, Embase, and Cochrane library collections were systematically searched. Twenty-six cohort studies with 523 cumulative patients, who underwent duodenum-sparing pancreatic head resection (DPPHR), were retrieved. The meta-analysis was based on 14 controlled studies.

Results: In total, 338 patients suffered cystic neoplasms and 59 PNETs, IPMN-174, MCN-43 and SPN-23 patients. Eighty-one patients (15.5%) histo-pathologically displayed a low-malignant tumor, of which 27 were carcinoma in-situ. Tumor recurrence was observed after a mean follow-up of 47.1 months in 11 patients. In-hospital and late mortality after DPPHR was 0.6% and 1.7%, respectively. The meta-analysis was based on 318 DPPHR compared to 404 PD patients. DPPHR was performed for premalignant neoplasm and PNET in 164 and 46 patients, and PD in 181 and 46 patients, respectively. Events of recurrence displayed no statistically significant difference between the DPPHR and PD groups.

Conclusion: DPPHR is associated with oncologically complete tumor resection for patients suffering premalignant IPMN, MCN, or SPN and for low-risk cancer.