Parenchyma-Sparing, Limited Pancreatic Head Resection for Benign Tumors and Low-Risk Periampullary Cancer--a Systematic Review.

Background: Parenchyma-sparing local extirpation of benign tumors of the pancreatic head provides the potential benefits of preservation of functional tissue and low postoperative morbidity.

Methods: Medline/PubMed, Embase, and Cochrane library databases were surveyed for studies performing limited resection of the pancreatic head and resection of a segment of the duodenum and common bile duct or preservation of the duodenum and common bile duct (CBD). The systematic analysis included 27 cohort studies that reported on limited pancreatic head resections for benign tumors. In a subgroup analysis, 12 of the cohort studies were additionally evaluated to compare the postoperative morbidity after total head resection including duodenal segment resection (DPPHR-S) and total head resection conserving duodenum and CBD (DPPHR-T).

Results: Three hundred thirty-nine of a total of 503 patients (67.4%) underwent total head resections. One hundred forty-seven patients (29.2%) of them underwent segmental resection of the duodenum and CBD (DPPHR-S) and 192 patients (38.2%) underwent preservation of duodenum and CBD. One hundred sixty-four patients experienced partial head resection (32.6%). The final histological diagnosis revealed in 338 of 503 patients (67.2%) cystic neoplasms, 53 patients (10.3%) neuroendocrine tumors, and 20 patients (4.0%) low-risk periampullary carcinomas. Severe postoperative complications occurred in 62 of 490 patients (12.7%), pancreatic fistula B + C in 40 of 295 patients (13.6%), resurgery was experienced in 2.7%, and delayed gastric emptying in 12.3%. The 90-day mortality was 0.4%. The subgroup analysis comparing 143 DPPHR-S patients with 95 DPPHR-T patients showed that the respective rates of procedure-related biliary complications were 0.7% (1 of 143 patients) versus 8.4% (8 of 95 patients) (p ≤ 0.0032), and rates of duodenal complications were 0 versus 6.3% (6 of 95 patients) (p ≤ 0.0037). DPPHR-S was associated with a higher rate of delay of gastric emptying compared to DPPHR-T (18.9 vs. 2.1%, p ≤ 0.0001).

Conclusion: Parenchyma-sparing, limited head resection for benign tumors preserves functional pancreatic and duodenal tissue and carries in terms of fistula B + C rate, resurgery, rehospitalization, and 90-day mortality a low risk of postoperative complications. A subgroup analysis exhibited after total pancreatic head resection that preserves the duodenum and CBD an association with a significant increase in procedure-related biliary and duodenal complications compared to total head resection combined with resection of the periampullary segment of the duodenum and resection of the intrapancreatic CBD.