Systematic Review and Meta-Analysis of Enucleation Versus Standardized Resection for Small Pancreatic Lesions

The appropriate surgical strategy in patients with small pancreatic lesions of low malignant potential, such as pancreatic neuroendocrine tumors, remains unknown. Increasing reports suggest limited pancreatic surgery may be a safe option for parenchymal preservation. PubMed and MEDLINE were searched in the English literature for studies from January 2000 to February 2015 examining enucleation for pancreatic lesions that were single-arm and comparative studies (versus resection). Single-arm enucleation studies were systematically reviewed. Comparative studies were included for meta-analysis. Endpoints include safety, complications, mortality, survival, and parenchymal-related outcomes. Thirteen studies comprising of 1101 patients undergoing enucleation were included. Seven studies were comparative studies of enucleation and standardized pancreatic resection. Enucleation was a shorter procedure (pooled mean differences (MD) = 109, 95 % confidence interval (CI) 105-114; Z = 46.37; P < 0.001) associated with less blood loss (pooled MD = 314, 95 % CI 297-330; Z = 37.47; P < 0.001). Both enucleation and resection had similar mortality and complication rates, but the rate of pancreatic fistula (all grades) (pooled odds ratio (OR) = 1.99; 95 % CI 1.2-3.4; Z = 2.57; P = 0.01] and rate of pancreatic fistula (grade B/C) (pooled OR = 1.58; 95 % CI 1.0-2.5; Z = 2.06; P = 0.04) was higher in the enucleation group. Enucleation resulted in lower rates of endocrine (pooled OR = 0.22; 95 % CI 0.1-0.5; Z = 3.21; P = 0.001) and exocrine (pooled OR = 0.07; 95 % CI 0.02-0.2; Z = 5.08; P < 0.001) insufficiency. The median 5-year survival was 95 % (range 93-98) and 84 % (range 79-90). Enucleation appears to be a safe procedure and achieves parenchymal preservation for small pancreatic lesions of low malignant potential. Its oncologic efficacy compared with standardized pancreatic resection with respect to long-term survival and recurrences have not been reported adequately and hence may not be concluded as being comparable.