From January 2004 to September 2006, 1,359 patients who were newly diagnosed with HCC were referred to the angiography unit of our institution for chemoembolization. For 138 out of 1,359 patients, MPV tumor thrombosis was evident at initial presentation. Until August 2008, seven patients (5%) have been followed-up without tumor recurrence. The seven patients were all male with an age range of 32-72 years (mean age, 53 years). Six patients were hepatitis B carriers, while one was a hepatitis C carrier. In all cases, a diagnosis of HCC was made based on laboratory testing (e.g., elevated α-fetoprotein levels and expression of viral markers) with typical CT and angiographic findings. All patients had received repeated chemoembolization, and one patient received additional radiation therapy for portal vein thrombosis. The criteria of 'complete remission' are as follows: 1) no enhancement on a portion in the liver, suggesting the presence of a viable tumor as seen on a CT image, 2) no hypervascular staining as detected on angiography, 3) a serum α-fetoprotein level less than 20 ng/mL, 4) no metastasis detected on a CT image. The institutional review board approved this study and patient informed consent was waived due to the retrospective nature of the study.

Every patient had undergone a CT examination every two or three months. The CT examinations were performed using various multi-detector row CT scanners. The scanners included a four-detector row CT scanner (MX 8000, Philips Medical Systems, Cleveland, OH), an eight-detector row CT scanner (Light Speed Ultra, GE Medical Systems, Milwaukee, WI), a 16-detector row CT scanner (Sensation 16, Siemens, Erlangen, Germany), and a 64-detector row CT scanner (Brilliance 64, Philips Medical Systems). All patients who received chemoembolization at our institution underwent CT scanning before each chemoembolization session.

The respective scanning parameters used for the 4-, 8-, 16- and 64-multidetector row CT scanners included the following: (detector configurations) 4 × 2.5, 8 × 1.25, 16 × 0.75, and 64 × 0.625 mm; (slice thickness) 3.2, 2.5, 3, and 3 mm; (reconstruction intervals) 3, 2.5, 3, and 2 mm; (table speed) 12.5, 13.5, 24 and 46 mm/rotation; (rotation time) 0.5-0.75 seconds.

Each CT examination included four phases that consisted of unenhanced images, hepatic arterial, portal venous, and equilibrium phases. After first acquiring unenhanced liver images in the craniocaudal direction, contrast medium (Iopromide, Ultravist 370; Schering, Berlin, Germany) was administered, which followed by a 30-mL sterile saline flush using a power injector (Multi-level CT; Medrad, Pittsburgh, PA). Contrast medium and saline solution were injected at 3 mL/sec through an 18-gauge plastic intravenous catheter placed in an antecubital vein. Contrast medium volume (delivered at 2 mL/kilogram body weight) varied from 100-130 mL. Hepatic arterial phase scan delays were 11-17 seconds after the descending aorta enhancement reached 100 Hounsfield units (HU), as measured using a bolus-tracking technique. Portal venous phase interscan delays were 20-30 seconds. The equilibrium phase commenced 180 seconds after completing the administration of contrast medium.

All angiographic examinations were performed by an experienced interventional radiologist (with 15 years of experience). Conventional chemoembolizations were performed as selectively as possible through the segmental or subsegmental arteries using a microcatheter (Microferret; Cook, Bloomington, IN or Progreat; Terumo, Tokyo, Japan). Initially, an iodized oil (Lipiodol; Andre Guerbet, Aulnay-sous-Bois, France) and doxorubicin hydrochloride (Adriamycin RDF; Ildong Pharmaceutical, Seoul, Korea) emulsion was administered into the feeders. The iodized oil volume ranged from 3 mL to 10 mL, and the amount of doxorubicin ranged from 20 mg to 50 mg. Gelatin sponge particles (Gelfoam; Upjohn, Kalamazoo, MI) that were mixed with mitomycin (Mitomycin-10; Kyowa Hakko Kogyo, Tokyo, Japan) and contrast material were administered into the feeders until the blood flow had nearly completely stopped. Every effort was made to seek and catheterize small arterial feeders that supplied tumor thrombi. Additional superselective cisplatin (Cisplan; Donga, Seoul, Korea) infusion was started after obtaining near complete stasis of vascular flow of the target arteries. Cisplatin was prepared as a solution at a concentration of 0.5 mg/mL, and the drug was infused as selectively as possible at a rate of 5-10 mL/min. The total amount of cisplatin ranged from 70 mg to 100 mg according to the patient body weight.

All CT scans were reviewed, and four CT examinations in each patient were analyzed quantitatively, including the initial CT scan, the CT scan after the first chemoembolization, the CT scan after the last chemoembolization session, and the last CT scan during follow-up. CT scans were retrospectively reviewed by two radiologists and focused on the following parameters: the extent of portal vein thrombosis, the diameter of the affected portal vein (MPV, right portal vein [RPV] and left portal vein [LPV]), enhancement of portal vein thrombosis, and the existence of a collateral vein as depicted on CT scans. Enhancement of portal vein thrombosis was assessed by the use of a region of interest (ROI) measurement, which was an average value of three ROI values of 20 mm² in the tumor thrombus. In the case of iodized oil uptake for portal vein thrombosis, an ROI measurement was not possible.