The Model for End-Stage Liver Disease (MELD)-Na score was calculated as described by Kim et al.[12] Details of the inclusion criteria for SBRT and treatment procedure have been described previously.[13] The summary of treatment procedure was as follows. TACE was underwent before SBRT. click here If respiratory motion was greater than 5 mm, patients held their breath in the end-expiratory phase using a spirometer or Abches (APEX Medical, Tokyo, Japan). A fiducial marker was not used for targeting the tumor. An arterial phase of dynamic computed tomography (CT) scan was used for radiation treatment

planning. Gross tumor volume (GTV) was defined by iodized oil and early enhancement. A clinical target volume (CTV) margin of 3 mm was usually added to GTV, and a planning target volume (PTV) margin of 5–8 mm was added to CTV. Eight non-coplanar ports were selected, and beams were delivered using 6–10-MV photons. The prescribed dose was calculated at the isocenter and was delivered on consecutive days. The prescribed dose was 50 Gy in five fractions until September 2004. Thereafter, 48 Gy

in four fractions was usually used, and 60 Gy in eight fractions was used when the PTV included the portal vein, inferior vena cava or heart. The patient receiving 52.5 Gy in seven fractions was planned to receive 60 Gy in eight fractions, but the last fraction was discontinued because of a BMS-777607 femoral neck fracture due to a fall. Portal vein thrombosis, bile duct stenosis, blood bilirubin increase, ascites, gastrointestinal disorders and ulcers were graded according to the Common Terminology Criteria for Adverse Events version 4.0. Portal vein thrombosis was non-tumoral as confirmed by dynamic CT scan or dynamic magnetic

resonance imaging. We retrospectively delineated the portal vein and bile duct on the planning dynamic CT scan. The portal vein was delineated from the main trunk to the first branch. The common bile duct, cystic duct find more and the first branch of the hepatic duct were delineated as the bile duct. The dose received by 2% of the volume (D2) of the portal vein and bile duct was calculated. The median follow-up duration was 17 months (range, 6–39). Median D2 of the portal vein was 12.6 Gy (range, 0.4–58.7). Portal vein thrombosis was observed in three patients (4.8%), all of whom developed grade 3. Common points of these patients were Child–Pugh class B and D2 of the portal vein of 40 Gy or higher (Fig. 1). Prescribed doses varied for D2 of the portal vein; thus, the biological equivalent dose (BED) with α/β ratio of 3 Gy (BED3) was calculated as an indicator. The BED3 values of D2 of the portal vein for patients 1, 2 and 3 were 217.4, 202.0 and 202.3 Gy, respectively. A77-year-old man suffered from non-B, non-C liver cirrhosis and was in Child–Pugh class B. His MELD-Na score was 11. He had received previous percutaneous ethanol injection and TACE for HCC.