Outcomes of pancreato-jejunostomies in a low volume hepato-pancreato biliary surgery unit in Sri Lanka

Abstract

Background: Leaks from the pancreatojejunostomy (PJ) after a pancreaticoduodenectomy (PD) occur in 20%-40% patients. Clinically significant grade B & C leaks occurring in 12.3%-16.5% result in morbidity and mortality. This study looks at the outcomes of patients who had a PJ in a low volume HPB surgical unit. Methods: Retrospective analysis of a HPB patient database and records from 2011-2021 was done. Patient demography, clinico-pathological details, pancreatic leak rates (ISGPF classification) and survival was recorded. PJs were done using a duct to mucosa technique with 3-0 and 5-0 polypropylene with a stent in-situ. Results: Of the 59 patients, mean age was 55.1 years with a M:F ratio of 1.1:1. 93.2% (n = 55) of the PJs were done as part of a PD, 3.4% (n = 2) after central pancreatectomy and 3.4% (n = 2) after longitudinal PJs. Histologically, 78% (n = 46) were malignant and 81.4% had a R0 resection margin (n = 35/43) while 18.6% (n = 8) were R1 resections. Grade B and C leaks occurred in 6.8% (n = 4). One grade B and both grade C fistulae were after PD. The other grade B fistula followed a central pancreatectomy. One grade C leak led to death on postoperative day 18 despite reopening. Among those that developed Grade B or C fistulae, the pancreas was soft in 50% (2/4) and firm in 25% (1/4), whereas 75% (3/4) had a PD diameter ≤ 3 mm. The median postoperative ICU and hospital stay were 2 (IQR = 3, n = 24) and 11 days (IQR = 8, n = 31), respectively. Conclusions: Clinically significant PJ leaks were comparatively low in this group. However, grade B & C fistulae resulted in significant morbidity and in one case mortality. Even in low volume centres, adhering to standard practice can result in good outcomes.