Personalized medicine for pancreatic fistula after pancreatoduodenectomy: Targeting improvement through risk profiling

Abstract

Background: Despite having many technical options, most surgeons employ a “one size fits all” management strategy for preventing pancreatic fistula (CR-POPF) following pancreatoduodenectomy (PD). Yet, the risk of CR-POPF development is different in every patient. The 4-element/10-point Fistula Risk Score (FRS) quantitates that risk, which can be segregated into four discrete zones for comparison (Negligible, Low, Moderate, High). More specifically, 80 unique biologic risk scenarios can be derived by combining the four FRS risk components (gland texture, pathology, duct size, and blood loss). We propose that risk mitigation should be targeted to best practice for the individualized presentation of patient risk. Methods: FRS profiles and outcomes of 5,323 PDs were accrued from a 17-institution, multinational collaborative (2003–2016). Associations between CR-POPF and varying mitigation strategies were analyzed for both the four defined FRS risk zones and each of the 80 distinct scenarios. Commonly encountered, or highly vulnerable, scenarios were individually investigated using multivariate analysis to derive directed mitigation approaches. Results: The CR-POPF rate in the overall cohort was 12.9% and varied by FRS zone (Negligible: 0.7%; Low: 5.2%; Moderate: 14.3%; High: 31.6%). The median and mode FRS number was 3. The overall distribution of cases across FRS zones was: Negligible-8.1%; Low-22.6%; Moderate-58.8%; High- 10.5.%. Among the overall series, each of the 80 possible distinct FRS scenarios were encountered at least twice. Neither CR-POPF rates, nor severity, necessarily correlate with scenario frequency. For instance, the most common combination was the only Negligible risk (FRS 0) scenario (#1: firm gland, PDAC/pancreatitis pathology, duct≥5mm, and ≤400 mL blood loss), with an 8.1% (n = 433) occurrence and just a 0.7% rate, where no specific mitigation strategies are useful. On the other hand, scenario #67 (FRS 7), demonstrating the highest CR-POPF rate (50%), was relatively infrequently encountered (n = 16). The majority (446/688: 64.8%) of all CR-POPFs actually occurred in the Moderate risk zone – where, notably, the most common scenario (#60) was more vulnerable to fistula occurrence than other scenarios in that zone in a risk-matched analysis (21.1% vs. 14.3%, p < 0.001). Fourteen Moderate risk scenarios have higher CR-POPF rates than four high risk scenarios. Overall, 36.3% of CR-POPF encountered come from just five scenarios. Optimal mitigation approaches have been identified for these most prolific situations. For example, for Scenario #60 (soft gland, high risk pathology, 2 mm duct, and ≤400 mL blood loss), a markedly reduced fistula occurrence (OR 0.22, 95% CI 0.12–0.44) was achieved using a “best-practice bundle” consisting of drain use without applications of stents or octreotide. Table 1 shows the wide heterogeneity of scenarios associated with each important CR-POPF outcome metric. Conclusion: Through this data a comprehensive catalogue has been created that links specific and unique CR-POPF risk scenarios (based on the biologic interplay of recognized risk factors) to their outcomes. The most commonly encountered case is that of Negligible risk. While High risk scenarios are rare, with high stakes, most fistulas are actually derived from Moderate risk (FRS 3-6) situations. An emphasis on improving management of these particular scenarios should yield the most value in decreasing the problem of CR-POPF. The uniqueness associated with every scenario requires optimal mitigation strategies be utilized at different times. Individualized approaches have been identified for the most common, as well as the most vulnerable, situations that surgeons face during PD.