Abstract
BACKGROUND: Paracentesis temporarily relieves malignant ascites but causes hypoalbuminemia. Cell-free and concentrated ascites reinfusion therapy (CART) reinfuses autologous proteins to prevent hypoalbuminemia and has been increasingly used in Japan. However, CART has not been widely adopted outside of Japan, and its benefit remains unclear. We evaluated the clinical outcomes and healthcare costs of CART compared with paracentesis in metastatic cancer. METHODS: This retrospective cohort study included hospitalized patients with metastatic solid cancer receiving CART or paracentesis across Japan (April 2016-March 2023). Baseline characteristics were balanced using overlap propensity-score weighting. Primary outcomes were in-hospital mortality, functional disability, and 30-day unplanned readmission. Secondary outcomes were length of stay (LOS), albumin administration or re-drainage rates, and costs. Mortality risk was assessed using a modified Poisson regression. The composite primary outcomes were assessed using a win-ratio approach. RESULTS: Among 1159 patients (CART: 457, paracentesis: 702) from 51 hospitals, the CART group had lower mortality than the paracentesis group (28.6% vs. 36.7%; risk ratio: 0.78, 95% confidence intervals [95% CI]: 0.64-0.94). The win-ratio analysis also favored the CART group over the paracentesis group (win ratio: 1.34, 95% CI: 1.09-1.64). Additionally, CART was associated with lower mortality and better composite outcomes than paracentesis, particularly among males, patients with serum albumin ≤ 2.5 g/dL, and those with non-gastrointestinal cancer. Despite higher procedural costs, CART was associated with shorter median LOS (14.1 vs. 19.0 days), lower albumin administration (11.6% vs. 17.3%) and re-drainage (32.7% vs. 52.7%) rates, and lower total median costs (4490.9 [interquartile range: 2042.3-7054.5] vs. 5084.1 [interquartile range: 3054.7-8659.7] USD) than paracentesis. CONCLUSIONS: CART was associated with improved clinical outcomes and healthcare costs over paracentesis among hospitalized patients with metastatic cancer, particularly in males, patients with serum albumin ≤ 2.5 g/dL, and those with non-gastrointestinal cancer. These findings may support clinical decision-making and resource allocation.