Celiac axis stenosis as an overlooked extrapancreatic predictor of fistula after pancreaticoduodenectomy.

BACKGROUND: Celiac axis stenosis (CAS) has been historically underappreciated in risk stratification for clinically relevant postoperative pancreatic fistula (CRPOPF). In this study, we aimed to investigate the role of CAS as an independent and clinically meaningful extrapancreatic predictor and assess its incremental value in established predictive models. MATERIALS AND METHODS: Overall, 1214 consecutive pancreaticoduodenectomies were conducted between January 2021 and December 2023. Patients with available preoperative abdominal contrast-enhanced computed tomography images were identified for this outcome-based retrospective cohort study. The CAS rate was assessed retrospectively using sagittal reconstruction of arterial phases and classified into no (< 30%), mild (30%-<50%), and severe (≥ 50%) stenosis. Associations between CAS and CRPOPF were analyzed using univariable and multivariable logistic regression. An extended prediction model incorporating CAS was developed based on the Alternative Fistula Risk Score (a-FRS) to evaluate the incremental predictive value of CAS. RESULTS: In total, 887 patients were eligible, of whom 228 (25.7%) had CAS. CAS significantly increased the risk of CRPOPF (p < 0.001) and remained an independent predictor (odds ratio: 3.134, 95% confidence interval: 2.153-4.564, p < 0.001), along with soft pancreatic texture, smaller main pancreatic duct, higher body mass index, and bile duct lesions. However, further stratification of stenosis into mild and severe categories did not reveal significant differences in CRPOPF rates, indicating an "all-or-none" threshold effect. Incorporating CAS into the a-FRS model yielded statistically significant but modest improvements in discrimination (p = 0.005), net benefit, and overall model fit (p < 0.001). CONCLUSION: CAS is an overlooked extrapancreatic predictor of CRPOPF. Its incorporation provided modest additional information within an established risk stratification framework. Routine reporting and preoperative intervention for CAS before pancreaticoduodenectomy may be informative and its integration into future studies may help predict and prevent CRPOPF. These results require further validation with multicenter prospective data.