Background: Partial splenic embolization (PSE) has been proposed to treat the consequences of hypersplenism in the context of portal hypertension, especially thrombocytopenia. However, a high morbidity/mortality rate has made this technique unpopular. We conducted a multicenter retrospective nationwide French study to reevaluate efficacy and tolerance.

Methods: All consecutive patients who underwent PSE for hypersplenism and portal hypertension in 7 tertiary liver centers between 1998 and 2023 were included.

Results: The study population consisted of 91 procedures in 90 patients, with a median age of 55.5 years [range 18-83]. The main cause of portal hypertension was cirrhosis (84.6%). The main indications for PSE were (1) an indication of medical treatment or radiological/surgical procedure in the context a severe thrombocytopenia (59.3%), (2) a chronic hemorrhagic disorder associated with a severe thrombocytopenia (18.7%), and (3) a chronic pain associated with a major splenomegaly (9.9%). PSE was associated with a transjugular intrahepatic portosystemic shunt in 20 cases. Median follow-up after PSE was 41.9 months [0.5–270.5]. Platelet count increased from a median of 48.0 G/L [IQR 37.0; 60.0] to 100.0 G/L [75.0; 148]. Forty-eight patients (52.7%) had complications after PSE; 25 cases were considered severe (including 7 deaths). A Child-Pugh B-C score (p<0.02) was significantly associated with all complications, a history of portal vein thrombosis (p<0.01), and the absence of prophylactic antibiotherapy (p<0.05) with severe complications.

Conclusion: Our results strongly confirm that PSE is very effective, for a long time, although a quarter of the patients experienced severe complications. Improved patient selection (exclusion of patients with portal vein thrombosis and decompensated cirrhosis) and systematic prophylactic antibiotherapy could reduce morbidity and early mortality in the future.