Computed tomography imaging-based predictors of the need for a step-up approach after initial endoscopic ultrasound-guided transmural drainage for pancreatic fluid collections

Background A step-up approach is recommended as a new treatment algorithm for pancreatic fluid collections (PFCs). However, determining which patients with PFCs require a step-up approach after endoscopic ultrasound-guided transmural drainage (EUS-TD) is unclear. If the need for a step-up approach could be predicted, it could be performed early for relevant patients. We aimed to identify PFC-related predictive factors for a step-up approach after EUS-TD. Methods This retrospective cohort study included consecutive patients who had undergone EUS-TD for PFCs from January 2008 to May 2020. Multivariable logistic regression analyses were performed to investigate PFC factors related to requiring a step-up approach. A step-up approach was performed for patients who did not respond clinically to EUS-TD. Results We enrolled 81 patients, of whom 25 (30.9%) required a step-up approach. In multivariate logistic regression analysis, the pre-EUS-TD number of PFC-occupied regions ≥ 3 (multivariate odds ratio [OR] 16.2, 95% confidence interval [CI] 2.68–97.6, P = 0.002), the post-EUS-TD PFC-remaining percentage ≥ 35% (multivariate OR 19.9, 95% CI 2.91–136.1, P = 0.002), and a positive sponge sign, which is a distinctive computed tomography finding in the early stage after EUS-TD (multivariate OR 6.26, 95% CI 1.33–29.3, P = 0.020), were independent predictive factors associated with requiring a step-up approach for PFCs. Conclusion Pre-EUS-TD PFC-occupied regions, post-EUS-TD PFC-remaining percentage, and a positive sponge sign were predictors of the need for a step-up approach. Patients with PFC with these findings should be offered a step-up approach whereas conservative treatment is recommended for patients without these findings. Clinical registration number UMIN 000030898.