The Fistula Risk Score Catalog: Toward Precision Medicine for Pancreatic Fistula After Pancreatoduodenectomy

Ann Surg. 2022 Feb 1;275(2):e463-e472. doi: 10.1097/SLA.0000000000004068.


Objective: This study aims to present a full spectrum of individual patient presentations of pancreatic fistula risk, and to define the utility of mitigation strategies amongst some of the most prevalent, and vulnerable scenarios surgeons encounter.

Background: The FRS has been utilized to identify technical strategies associated with reduced CR-POPF incidence across various risk strata. However, risk-stratification using the FRS has never been investigated with greater granularity. By deriving all possible combinations of FRS elements, individualized risk assessment could be utilized for precision medicine purposes.

Methods: FRS profiles and outcomes of 5533 PDs were accrued from 17 international institutions (2003-2019). The FRS was used to derive 80 unique combinations of patient "scenarios." Risk-matched analyses were conducted using a Bonferroni adjustment to identify scenarios with increased vulnerability for CR-POPF occurrence. Subsequently, these scenarios were analyzed using multivariable regression to explore optimal mitigation approaches.

Results: The overall CR-POPF rate was 13.6%. All 80 possible scenarios were encountered, with the most frequent being scenario #1 (8.1%) - the only negligible-risk scenario (CR-POPF rate = 0.7%). The moderate-risk zone had the most scenarios (50), patients (N = 3246), CR-POPFs (65.2%), and greatest non-zero discrepancy in CR-POPF rates between scenarios (18-fold). In the risk-matched analysis, 2 scenarios (#59 and 60) displayed increased vulnerability for CR-POPF relative to the moderate-risk zone (both P < 0.001). Multivariable analysis revealed factors associated with CR-POPF in these scenarios: pancreaticogastrostomy reconstruction [odds ratio (OR) 4.67], omission of drain placement (OR 5.51), and prophylactic octreotide (OR 3.09). When comparing the utilization of best practice strategies to patients who did not have these conjointly utilized, there was a significant decrease in CR-POPF (10.7% vs 35.5%, P < 0.001; OR 0.20, 95% confidence interval 0.12-0.33).

Conclusion: Through this data, a comprehensive fistula risk catalog has been created and the most clinically-impactful scenarios have been discerned. Focusing on individual scenarios provides a practical way to approach precision medicine, allowing for more directed and efficient management of CR-POPF.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Female
  • Humans
  • Male
  • Middle Aged
  • Pancreatic Fistula / epidemiology*
  • Pancreaticoduodenectomy*
  • Postoperative Complications / epidemiology*
  • Precision Medicine*
  • Retrospective Studies
  • Risk Factors