Clinical characteristics and association of percutaneous catheter drainage in children with acute lymphoblastic leukemia complicated by asparaginase-associated pancreatitis and ascites

Background Asparaginase-associated pancreatitis (AAP) is a serious complication of acute lymphoblastic leukemia (ALL) chemotherapy and is frequently accompanied by ascites; however, evidence on percutaneous catheter drainage (PCD) in children with mild AAP and ascites is lacking. This study evaluated the association between PCD, performed using a retrospectively defined ascites-depth threshold of > 30 mm, and clinical outcomes in children with mild AAP. Methods We retrospectively reviewed 47 children with ALL who developed AAP under the CCLG-ALL-2020 protocol at a single center between January 2022 and December 2024. Severity was classified according to the INSPPIRE criteria and the NASPGHAN Pancreas Committee classification. Among the 28 patients with mild AAP and ascites depth > 30 mm, 18 underwent PCD and 10 received conservative treatment; because treatment was not randomized, patients were compared according to the treatment actually received (Mann-Whitney U and Fisher’s exact tests). Results Of the 47 patients (median age 6.5 years; 59.6% male), 40 (85.1%) had mild, 3 (6.4%) moderately severe, and 4 (8.5%) severe AAP, and 35 (74.5%) developed ascites. In the mild AAP subgroup, PCD was associated with shorter fasting duration (median 7.0 vs. 10.5 days, P = 0.031) and total hospital stay (26.0 vs. 31.0 days, P = 0.043), with similar pancreatic enzyme recovery times ( P > 0.05 for all). One patient (5.6%) in the PCD group developed self-limited puncture-site oozing; in the conservative group, two patients (20.0%) required rescue paracentesis and one (10.0%) developed a pancreatic pseudocyst. All 4 patients with severe AAP died despite PCD. Conclusion In children with mild AAP and ascites, PCD guided by a retrospectively defined, data-derived 30-mm threshold was associated with shorter fasting duration and hospital stay in this small, non-randomized cohort. The threshold is prone to selection bias and lacks validation; these hypothesis-generating findings require confirmation in larger prospective studies before PCD can be recommended in mild AAP.