Portal hypertension increases the risk of hepatic decompensation after 90Yttrium radioembolization in patients with hepatocellular carcinoma: a cohort study

BACKGROUND: Transarterial radioembolization (TARE) is increasingly used in patients with hepatocellular carcinoma (HCC). This treatment can induce or impair portal hypertension, leading to hepatic decompensation. TARE also promotes changes in liver and spleen volumes that may modify therapeutic decisions and outcomes after therapy. OBJECTIVES: We aimed to investigate the impact of TARE on the incidence of decompensation events and its predictive factors. DESIGN: In all, 63 consecutive patients treated with TARE between February 2012 and December 2018 were retrospectively included. METHODS: We assessed clinical (including Barcelona Clinic Liver Cancer stage, portal hypertension assessment, and liver decompensation), laboratory parameters, and liver and spleen volumes before and 6 and 12 weeks after treatment. A multivariate analysis was performed. RESULTS: In total, 18 out of 63 (28.6%) patients had liver decompensation (ascites, variceal bleeding, jaundice, or encephalopathy) within the first 3 months after therapy, not associated with tumor progression. Clinically significant portal hypertension (CSPH) and bilobar treatment independently predicted the development of liver decompensation after TARE. A significant volume increase in the non-treated hemi-liver was observed only in patients with unilobar treatment (median volume increase of 20.2% in patients with right lobe TARE; p = 0.007), especially in those without CSPH. Spleen volume also increased after TARE (median volume increase of 16.1%; p = 0.0001) and was associated with worsening liver function scores and decreased platelet count. CONCLUSION: Bilobar TARE and CSPH may be associated with an increased risk of liver decompensation in patients with intermediate or advanced HCC. A careful assessment considering these variables before therapy may optimize candidate selection and improve treatment planning.