# Relevance of the International Study Group of Pancreatic Surgery and the Dutch Pancreatic Cancer Group Classifications of Pancreas-Specific Risk Factors in Predicting Clinically Relevant Postoperative Pancreatic Fistula in the Whipple Procedure

## TL;DR

This study compares two classification systems for predicting postoperative pancreatic fistula after Whipple surgery, finding both effective but one simpler to use.

## Contribution

The study evaluates and compares the predictive accuracy of ISGPS and DPCG classification systems for postoperative pancreatic fistula in a single-center cohort.

## Key findings

- Main pancreatic duct diameter ≤3 mm was the strongest independent predictor of clinically relevant postoperative pancreatic fistula.
- Both ISGPS and DPCG classifications showed similar predictive accuracy (AUC 0.707 vs 0.710) for postoperative complications.
- Type D (ISGPS) and two-risk-factor group (DPCG) had the highest complication rates at 40.5%.

## Abstract

Background

The International Study Group of Pancreatic Surgery (ISGPS)introduced a four-tier classification system, including pancreatic texture and pancreatic duct diameter, to aid the risk stratification of clinically relevant postoperative pancreatic fistula. The Dutch Pancreatic Cancer Group (DPCG) validated the ISGPS risk classification and proposed a three-tier classification system. This study was conducted to compare the clinically relevant postoperative pancreatic fistula rate among two classification systems.

Methods

This study was conducted by a retrospective review of the prospectively maintained data of 165 patients who underwent pancreaticoduodenectomy, also known as the Whipple Procedure, between 2015 and 2024 in a single unit of the Department of Surgical Gastroenterology at Tribhuvan University Teaching Hospital, Kathmandu, Nepal. The preoperative, intraoperative, and postoperative variables were analyzed to assess the relevance of the two classifications to predict clinically relevant postoperative pancreatic fistula.

Results

Ampullary carcinoma was the most common indication of pancreaticoduodenectomy (47.3%, n=78). Fifty patients (30.3%) had a main pancreatic duct diameter ≤3 mm, and 62.4% (n=103) had soft pancreatic texture. Twenty-eight patients (17.0%) developed clinically relevant postoperative pancreatic fistula, 44 (26.7%) had major complications (Clavien Dindo ≥3), and in-hospital mortality was seen in 13 (7.9%). Main pancreatic duct diameter ≤3 mm (36.0% vs 8.7%, P: <0.001), blood loss ≥500 ml (21.4% vs 7.5%, P: 0.027), and non-pancreatic pathology (21.5% vs 4.5%, P: 0.010) were significantly associated with clinically relevant postoperative pancreatic fistula but main pancreatic duct diameter ≤3 mm (OR: 7.313, P: 0.007, 95%CI: 1.462-12.124) was the only independent predictor. The rate of clinically relevant postoperative pancreatic fistula was significantly different in the subclasses in both the ISGPS and the DPCG classifications, being highest in Type D (40.5%, n=17). Both the classification systems showed similar predictivity for clinically relevant postoperative pancreatic fistula, with similar area under the curve, 0.707 for the ISGPS classification and 0.710 for the DPCG classification.

Conclusion

This study showed that the Type D (as per the ISGPS classification) or the two-risk-factor group (as per the DPCG classification) has the highest rate of postoperative complications after pancreaticoduodenectomy. On further analysis of the classification of pancreas-specific risk factors, including pancreatic texture and main pancreatic duct diameter, according to the ISGPS and DPCG classification systems, predictive accuracy was similar for clinically relevant postoperative pancreatic fistula; however, the DPCG classification with the simpler three-tier system is easier to apply in practice.