Minimally invasive distal pancreatectomy for pancreatic cystic neoplams and localized solid tumors: variation in utilization and potential as a quality measure

Abstract

Background: Minimally invasive distal pancreatectomy (MIDP) has been demonstrated to be a technically feasible and oncologically sound alternative to open distal pancreatectomy (ODP) for pancreatic cystic neoplasms and localized solid tumors, including pancreatic adenocarcinoma (PDAC) and pancreatic neuroendocrine tumors (PNET). However, it is unknown which patients are being selected for MIDP or the extent to which hospitals are adopting this approach. The objectives of this study were to (1) identify patient-level factors associated with utilization of MIDP, (2) describe hospital-level variation in utilization of MIDP, and (3) compare outcomes between ODP and MIDP for cystic lesions and early stage solid tumors. Methods: The American College of Surgeons National Surgical Quality Improvement Program Pancreas Targeted Data File was used to identify patients from 2014-2017 who underwent distal pancreatectomy for pancreatic cysts (serous cystadenomas, mucinous cystadenomas, and intraductal papillary mucinous neoplasms), Stage I PDAC, or Stage I PNET. Surgical approach was categorized as OPD or MIDP (laparoscopic or robotic). Outcomes of interest included unplanned conversion of MIDP to OPD, postoperative length of stay, and a composite outcome of death or serious morbidity (DSM). Associations between patient factors, surgical approach, and postoperative complications were assessed by risk-adjusted hierarchical multivariable logistic regression. Hospital-level utilization of MIDP was calculated to assess institutional variability in surgical approach. Results: Analysis identified 3,276 patients undergoing distal pancreatectomy at 141 hospitals. Overall, 55.1% of patients had pancreatic cysts, 12.0% had Stage I PDAC, and 32.9% had Stage I PNET. ODP was performed in 37.3% of cases (35.6% of cyst cases, 61.6% of PDAC cases, and 31.4% of PNET cases), while 62.7% were MIDP (of which 76.1% were laparoscopic and 23.9% were robotic). The rate of unplanned conversion to open during MIDP was 11.8% (12.1% laparoscopic, 10.8% robotic, p = 0.700). Patients were more likely to undergo MIDP if they were <55 years old (66.6% vs 56.4% if ≥75; aOR 1.34, 95%CI [1.08-1.67]) or had a BMI≥30 (67.1% vs 57.8%, aOR 1.34, 95%CI [1.11-1.62]). Patients were less likely to undergo MIDP if they had PDAC pathology (38.4% vs 66.0% for PNET; aOR 0.33, 95%CI [0.25-0.44]). There was significant variation in hospital-level MIDP utilization (range: 0% to 100% of cases, Figure). Only 36.9% of hospitals attempted MIPD on at least 75% of patients with localized neoplasms. Similar to previous studies, length of stay was significantly shorter following MIDP (5.4 vs 7.2 days, p<0.001), and patients were significantly less likely to experience DSM following MIDP (16.2% vs 20.1%; aOR 0.74, 95%CI [0.62-0.89]). There was no difference in DSM between cystic neoplasms (DSM 17.2%), PDAC (DSM 18.0%), and PNET (DSM 18.0%; p = 0.519). Conclusion: Despite being a safe alternative to ODP with fewer complications and shorter length of stay, MIDP is still underutilized even in cases of cystic neoplasms and localized solid pancreatic tumors. While some patient-level factors are associated with use of MIDP, hospital variation in adoption of MIPD appears to be the principal driver of utilization and should be the primary focus of strategies to increase its uptake. In addition to short-term and oncologic outcome measures, hospital-level MIPD rates in appropriate patients could be utilized as a measure of hospital quality to both improve patient outcomes and encourage adoption of minimally invasive techniques.