Predictive factors for platelet increase after partial splenic embolization in liver cirrhosis patients

Background and Aim: Partial splenic embolization (PSE) is often performed for improving thrombocytopenia in cirrhotic patients. We investigated the largely unclear predictive factors for platelet increase at both 1 month and 1 year after PSE. Methods: Aimed at increasing the platelet count, PSE was performed in 42 cirrhotic patients with thrombocytopenia (platelets < 80 x 10(4)/mL) caused by hypersplenism. The clinical data were analyzed to clarify the predictive factors for platelet increase at 1 month (n = 42) and 1 year (n = 38) after PSE. Results: The mean splenic infarction ratio was 76.7% +/- 11.2%. The platelet count increased to 259% +/- 112% and 228% +/- 75% of the pretreatment values at 1 month and at 1 year after PSE, respectively. Stepwise multiple linear regression analysis showed that the infarcted splenic volume had a positive independent association with the increase in platelet count at both 1 month (P = 0.00004) and 1 year (P = 0.005) after PSE (increase in platelet count (x10(4)/mL): at 1 month = 0.752 + 0.018 x infarcted splenic volume (mL), R-2 = 0.344; at 1 year = 2.19 + 0.01 x infarcted splenic volume (mL), R-2 = 0.203). Receiver operating characteristic analysis yielded a cut-off value of 388 mL of infarcted splenic volume for achieving an increase of 5.0-8.0 x 10(4)/mL in platelet count at 1 year. Conclusions: PSE can reduce the platelet pool and induce an increase in platelet count. This increase is greatly dependent on the infarcted splenic volume.