Hepatocellular carcinoma invading the main portal vein: treatment with transcatheter arterial chemoembolization and portal vein stenting.

To retrospectively analyze the therapeutic results of percutaneous transhepatic portal vein stenting (PTPVS) and transcatheter arterial chemoembolization (TACE) treatment in 58 patients with hepatocellular carcinoma (HCC) invading the main portal vein (MPV). A total of 58 procedures of PTPVS were performed, immediately after which TACE was undertaken to control HCC. The clinical effects, complications, digital subtraction angiographic appearance, stent patency rates, cumulative survival rates, and predictive factors for survival were evaluated. The Kaplan-Meyer method and the log rank test were used for survival analysis. Multivariable analysis was also conducted by the Cox proportional hazard model. No patient died during stent placement or within the first 24 h. No severe procedure-related complications were observed. After stent placement, the mean +/- standard deviation portal venous pressure levels decreased from 41.43 +/- 8.56 cmH(2)O to 37.19 +/- 7.89 cmH(2)O (p < 0.01). At the time of analysis, 9 of the 58 patients survived. The 60-, 180-, 360-, and 720-day cumulative patency rates were 98.1%, 71.0%, 52.6%, and 42.1%, respectively, with a mean patency time of 552.9 +/- 88.2 days and a median patency time of 639.00 +/- 310.00 (95% confidence interval [95% CI], 31.40-1246.60) days. The 60-, 180-, 360-, and 720-day cumulative survival rates for the total study population were 74.1%, 27.1%, 17.2%, and 13.8%, respectively, with a median survival time of 113 +/- 27.29 (95% CI, 59.51-166.49) days. In the univariate analysis, the following six variables were significantly associated with the prognosis: (1) HCC type; (2) Child-Pugh grade; (3) MPV stenosis/occlusion; (4) arteriovenous shunt; (5) iodized oil deposition; and (6) number of TACE procedure. In addition, having diffuse-type HCC and Child-Pugh grade B disease were each independent factors associated with decreased survival time in the multivariate analysis. PTPVS-TACE is feasible and may be useful to control HCC invading the MPV.