|
Back to 2016 Annual Meeting
Tailored Approach to Reconstruction Following Pancreaticoduodenectomy: Pancreaticogastrostomy for High Risk Remnant Is Associated With Decreased Pancreatic Fistula Rate
George Kazantsev*, Austin L. Spitzer, Peter Peng, Rene Ramirez
Surgery, Kaiser Permanente, Oakland, CA
INTRODUCTION. Postoperative pancreatic fistula (POPF) remains a devastating complication following pancreaticoduodenectomy (PD). Multiple retrospective studies have demonstrated the superiority of pancreaticogastrostomy for (PG) over pancreaticojejunostomy (PJ) in reducing the rate of POPF, while several randomized trials failed to validate this finding. It has also been shown that PG may be associated with a reduced risk of POPF in a subgroup of patients with a "high risk remnant" - soft pancreas/and or small pancreatic duct. We hypothesized that selective use of PG in high risk remant patients may lead to an overall reduction in POPF rate.
METHODS. Retrospective review of patients (pts) who underwent PD with respect to type of procedure, gland type (soft vs. hard), duct size (<3mm vs >3mm), and postoperative outcomes was performed. Pancreatic remnant was classified as "high risk" if at least one risk factor (soft pancreas, small duct) was present. POPF was evaluated according to ISGPF grading system.
RESULTS. Between 2009 and 2015, a total of 98 pts underwent PD at Kaisier Oakland Medical Center for benign (22 pts) and malignant (76 pts) disease. Pancreatic ductal adenocarcinoma was the most common malignancy - 60.5%. Pylorus preservation was acieved in 66 pts and vascular resection/reconstruction was required in 15 (67% and 15.3%, respectively). PG was performed in 29 (29.5%) and PJ in 69 (69%) of cases. Type of anastomosis used was up to the individual surgeon; PG was performed only for high risk remnant while PJ was done for both types. In-hospital mortality was 2% (none related to POPF). Firm pancreas with a large duct was found in 53 cases, and soft remnant with small duct in 34. Other combinations included soft pancreas and large duct in 8, and firm pancreas with small duct in 3 pts. Clinically relevant POPF (Grade B or C) ocurred in 9 pts for the entire group (9.4%). Overall, PJ was associated with POPF in 7 pts (10.3%) and PG in 2 pts (7.1%) - NS. Subgroup analysis revealed an almost 10-fold increase in the rate of POPF when PJ was done for high risk remnant; 4 of these leaks were grade C requiring prolonged lenght of stay, additional invasive procedures or re-operations. In contrast, PG performed for high risk remnant led to onle 2 grade B leaks treated on an outpatient basis (Table 1). However, significant bleeding occured in 2 pts of PJ group (2.9%), all intraabdominal, and in 4 pts of PG group (14.3%, P=0.006 by Fisher exact test), all gastrointestinal.
CONCLUSION. Selective approach to reconstruction following PD may lead to a significant decrease in clinically relevant POPF. PJ remains the procedure of choice for low risk remnants while PG should be used for pts with high risk remnants. Refinement of PG technique of should be considered to reduce increased rate of gastrointestinal bleeding.
Table 1. Incidence of clinically relevant POPF depending on the type of pancreatic remnant and anastomosis type.
| Procedure | # | POPF(%) | P value, Fisher exact test | | PJ, low risk | 53 | 2(3.8%) | 0.006* | | PJ, high risk | 15 | 5(33.3%) | | | PG, high risk | 28 | 2(7.1%) | 0.04** |
*PJ, low risk vs. PJ, high risk; **PJ high risk vs. PG (no PG done for low risk)
Back to 2016 Annual Meeting
|
