A classification of laparoscopic central pancreatectomy determined on the basis of anatomical landmarks in 109 patients

Abstract

Background: Laparoscopic central pancreatectomy (LCP) is usually proposed for non-malignant neck-body neoplasms, but it can be proposed for head-neck lesions to avoid pancreaticoduodenectomy or for body-tail lesions to avoid distal pancreatectomy. The aim of this study was to classify CP on the basis of the proximal resection level.

Method: We retrospectively studied all consecutive LCPs performed in our institution from 2011 to 2024. LCP can be associated with vascular procedures (gastroduodenal artery or splenic vessels) and was classified into 3 types according to proximal level of pancreatic resection: head-LCP, neck-LCP, and body-LCP (results in this order). The primary objective of this study was the creation and definition of this new classification for LCP. The secondary objectives were to compare outcomes and textbook outcome (TBO) completion, defined as no clinically relevant postoperative pancreatic fistula, no clinically relevant postpancreatectomy hemorrhage, no bile leaks, no readmission, no mortality, and no severe morbidity within 90 postoperative days.

Results: In total, 109 patients underwent LCP with head-LCP, neck-LCP, and body-LCP observed in 20%, 66%, and 14%, respectively. The type was correlated with the distance of the lesion from the gastroduodenal artery (P = .0001). Head and body-LCPs were more frequently associated with vascular procedures (68% vs 17% vs 40%, P = .001) and body-LCP was associated with larger tumor size (millimeters) compared with head and neck-LCPs (17 vs 21 vs 35, P = .07). TBO did not differ significantly (41% vs 58% vs 47%, P = .31) with one patient death (mortality <1%). At median follow-up (22 months), the rate of new-onset exocrine (6%; P = .10) or endocrine (4%; P = .76) pancreatic insufficiencies was similar. On multivariate analysis, only American Society of Anesthesiologists score ≥2 (P = .03) and pancreatic texture (P = .01) were prognostic factors for TBO while LCP type was not.

Conclusion: Head and neck-LCPs were more challenging as assessed by the associated vascular procedures without impact on TBO, allowing in some selected patients parenchymal-sparing surgery. Further studies comparing CP with standard pancreatic resections are needed.