Delayed gastric emptying in diabetic patients undergoing pancreaticoduodenectomy: A procedure-targeted nsqip analysis

Abstract

Presenter: Alissa Greenbaum MD | Rutgers University Background: Delayed gastric emptying (DGE) is a major source of morbidity after pancreaticoduodenctomy (PD), occurring in up to 20% of patients. Patients with type I or II diabetes mellitus (DM) have a propensity for gastric dysmotility, however the exact relationship between DGE and DM is not clearly established. The aim of this study was to determine the incidence of DGE in patients with and without DM after PD. Methods: The American College of Surgeons National Quality Improvement Project (ACS-NSQIP) procedure-targeted pancreatectomy database was queried from 2014 to 2017 for patients undergoing PD and combined with the main ACS-NSQIP database. Patient demographic, clinical and perioperative variables were compared by DM status. The primary outcome was rate of DGE. A subset analysis of insulin-dependent DM (IDDM) and non-insulin-dependent DM (non-IDDM) patients was also performed. Statistical analyses were performed using chi-square, Mann Whitney-U tests, and logistic regression. Results: A total of 14,735 patients met inclusion criteria, including 10,930 non-DM patients (74.2%) and 3805 DM patients (25.8%). Median cohort age was 66 years and 53.7% were male (n=7918). DGE occurred in 17.1% of patients overall (n=2519); 17.2% in non-DM and 16.8% in DM patients (p=0.60). DM patients had a higher ASA class, increased rates of hypertension, increased proportion of pancreatic adenocarcinoma, larger duct size, harder gland texture, and required more frequent vascular resection (p<0.001). DM patients had equivalent rates compared to non-DM patients of post-operative superficial surgical site infection (7.8% in both; p=0.89), deep surgical site infection (1.4 vs 1.5%; p=0.56), urinary tract infection (3.2 vs 2.8%; p=0.23) and pneumonia (4.3 vs 3.6%; p=0.19), and lower rates of organ space infection (11.5 vs 14.0%; p<0.001) and pancreatic fistula (14.9 vs 19.0%; p<0.001). On subset analysis, IDDM had better outcomes compared to non-IDDM patients in organ space infections (8.8 vs 14.2%; p<0.001) and pancreatic fistula (11.7 vs 18.3%; p<0.001). Rates of DGE were 18.0% (n=337) in non-IDDM and 15.7% in IDDM patients (p=0.162). There were no significant differences in rates of reoperation, hospital length of stay, readmission rates or 30-day mortality. On multivariable logistic regression, male sex, advanced age, history of smoking, pancreatic fistula, and organ space infection were associated with DGE. DM status was not predictive of DGE. Conclusion: No differences in rates of DGE between DM and non-DM patients after PD were found, strengthening the hypothesis that DGE is a separate physiologic entity from diabetic gastroparesis. IDDM patients demonstrated better postsurgical outcomes compared to non-IDDM patients in terms of intra-abdominal infection and pancreatic fistula, suggesting a potential role of monitored insulin or hyperglycemia regulation in the development of these complications.