The eligibility criteria for chemotherapy were as follows: (1) 18-75 years of age; (2) preserved liver function (Pugh-Child class A and B); (3) acceptable range of thrombocytopenia due to portal hypertension without any other causes of bone marrow suppression (platelet count>50,000/mm³ and leukocyte count>2,000/mm³); (4) normal serum creatinine (Cr<1.2 mg/dL); (5) acceptable range of hepatitis activity (alanine aminotransferase, [ALT]<100 IU/L); and (6) Eastern Cooperative Oncology Group performance status (PS) 0-2. The inclusion criteria of far advanced HCC features for the protocol were as follows: (1) TNM stage III or IV by AJCC/UICC or LCSGJ classification; (2) major vascular invasion, such as main portal vein tumor thrombosis or main hepatic vein tumor thrombosis; and (3) marked vascular shunts, such as AVSs or APSs. Patients with HCC with extra-hepatic metastases at the time of initial diagnosis were also included if they had a marked APS or AVS. The exclusion criteria were as follows: (1) presence of other malignant diseases; (2) history of recent upper gastrointestinal bleeding; (3) sepsis; and (4) serious chronic disorders in other organs. We received informed consent in all patients treated with this protocol, which was approved by the Institutional Review Board of Bundang Jesaeng General Hospital.

Of the HCC patients admitted to our hospital between March 2003 and December 2009, 4 patients met the inclusion criteria of far advanced HCC. Because one patient is currently undergoing treatment, three of the four patients were assessed retrospectively.

The treatment protocol, consisted of one cycle of HAIC-EC on day 1 and PIC-IF on day 2, repeated every 4-6 weeks until efficacy was achieved as defined by the criteria of the Response Evaluation Criteria in Solid Tumors (RECIST). The dose of epirubicin (50 mg/m²/session; Il Dong Pharmaceuticals Co. Ltd., Seoul, Korea) and cisplatin (60 mg/m²/session; Dong-A Pharmaceuticals Co. Ltd., Seoul, Korea) for HAIC was adjusted based on age, liver function, and leukocyte count measured just before each treatment cycle. For the dose calculation equation for epirubicin and cisplatin, we used empirical parameters in each variable as follows: calculated dose=(dose per body surface area [BSA])×(leukocyte count/4000)×(1-[Pugh score-5]/10)×(1-[age-45]/100)6; we did not consider the intrahepatic tumor volume parameter for the adjustment of dosage because all HCCs were in the advanced stage. Fixed doses of IFN-γ (2 million units per vial, Intermax-gamma; LG Life Sciences Pharmaceuticals Co. Ltd., Daejeon, Korea) and 5-FU (500 mg per ample; Choong Wae Pharmaceuticals Co. Ltd., Seoul, Korea) were used for PIC to the far advanced HCCs (10 MU and 1500 mg per session, respectively). This cocktail ratio was decided by the maximum efficient combination, which was calculated by HCC cell line experiments.

For HAIC-EC, 1 liter of normal saline was administered intravenously before and after the transarterial angiographic procedure to prevent cisplatin-induced nephrotoxicity. Dolasetron mesylate (200 mg per tab, Anzemet, Sanofi-Aventis; HANDOK Pharmaceuticals Co. Ltd., Seoul, Korea) was used to treat nausea and vomiting following chemotherapy for 4 successive days. Megestrol acetate oral suspension (800 mg per pack, Apetrol; LG Life Sciences Pharmaceuticals Co. Ltd., Daejeon, Korea) was also used to improve appetite and quality of life.

For PIC-IF, IFN-γ (10 million units) and 5-FU (1,500 mg) were mixed well just before injection via a PEIT needle (21 gauge, 200 mm size with 3-7 holes and a closed tip) with an extension tube. Under ultrasonographic guidance, the needle tip was rapidly placed in the target area, which included HCC at the border invading the portal or hepatic veins or the inner mantle portion of the HCC. After which the syringe with the cocktail was injected very slowly over 10 minutes. The fasting state was at least 6 hours, and 50 mg of demerol and hydroxyzine (25 mg per tab, Ucerax; Yuhan Pharmaceuticals Co. Ltd., Seoul, Korea), a sedating antihistamine, was administered 30 minutes before the procedure as a premedication, and propacetamol HCl (1 g per vial, Denogan, Yungjin Pharmaceuticals Co. Ltd, Seoul, Korea), an antipyretic, was administered intravenously for 15 min to treat IFN-γ-induced flu-like reactions, such as fevers, chills and myalgias, immediately following the procedure.

The length of admission was usually from 4-7 days. Laboratory tests were usually performed 3 times during the admission (before and 1 day after treatment, and prior to discharge). Additional laboratory tests were performed from 7-10 days after discharge in the outpatient clinic to evaluate the degree of bone marrow suppression.

Treatment efficacy according to the RECIST criteria was evaluated after at least two cycles of protocol therapy, and every cycle thereafter. Treatment efficacy was based on changes in the characteristic features of the HCC as shown in clinico-radiologic studies, including an alpha-fetoprotein (AFP) level, PIVKA-II, conventional biochemical liver function tests, a triphasic liver CT scan, and a hepatic arteriogram, which were performed during each treatment cycle. The radiologic features included the changes in the number and sizes of tumors, response pattern of vascular thrombosis and occlusion of shunts, together with enhancing patterns around the tumor nodules.

Adverse reactions were assessed during and after each cycle of protocol therapy according to the criteria of the National Cancer Institute Common Toxicity Criteria (NCI-CTC; version 2.0).3

All three patients, who were AJCC-TNM stage IV, BCLC stage C, and Pugh-Child class A, had an objective response, tumor shrinkage, and disappearance of the APS or AVS. CR to our multimodality protocol, the combination treatment of HAIC-EC and PIC-IF, was obtained in two patients. The disease-free survivals of the two patients with CRs without any evidence of recurrence until the present are 28.6 months and 30.3 months, respectively. One patient had a CR in an intrahepatic HCC lesion, but progressive disease in metastatic lesions (lung and adrenal gland), and died of respiratory failure and septicemia; the survival time was 22.4 months. Each clinical course is described in detail below (Table 1).

A 60-year-old man was admitted due to variceal bleeding. Dynamic CT scan showed a massive infiltrative type of HCC with a maximum diameter ≥12 cm and the main portal vein tumor thrombosis with a massive APS in the right lobe. The baseline blood tests were as follows: white blood cell count (WBC) 17,100/mm³, hemoglobin 6.2 g/dL, platelet count 217,000/mm³, albumin 2.8 g/dL, total bilirubin 1 mg/dL, prothrombin time-INR 1.2, alkaline phosphatase (ALP) 488 IU/L, aspartate aminotransferase (AST) 170 IU/L, alanine aminotransferase (ALT) 64 IU/L, total cholesterol 109 mg/dL, creatinine 0.9 mg/dL, AFP and PIVKA-II >100,000 ng/mL and >2,000 mAU/mL respectively. HBsAg and anti-HCV were negative. After recovery from variceal bleeding by ligation therapy, when the liver function reverted to Pugh-Child A without ascites, the protocol therapy of HAIC-EC and PIC-IF was started. This patient had a CR in both serologic HCC biomarkers and radiologic evaluations after 10 cycles of the treatments, and is still alive without any evidence of recurrence for 30 months (disease-free interval [DFI], 17 months) (Fig. 1, 2).

A 67-year-old man, who had experienced exertional dyspnea and chest discomfort for 1 month, was admitted for the treatment of advanced HCC. At the time of the initial evaluation, a dynamic CT scan showed a massive infiltrating HCC with a maximum diameter of 14 cm occupying the right lobe with extensive tumor thrombosis in the right hepatic vein, inferior vena cava, and right atrium, and the main portal vein tumor thrombosis with a massive APS. Multiple metastases were observed in both lung fields. The initial laboratory test showed the following: hemoglobin 15 g/dL, WBC 5,300/mm³, platelet 94,000/mm³, albumin 3.9 g/dL, total bilirubin 1.45 mg/dL, prothrombin time-INR 1.18, ALP 1,951 IU/L, AST 118 IU/L, ALT 48 IU/L, total cholesterol 86 mg/dL, AFP and PIVKA-II 2,486 ng/mL and >2000 AU/mL, respectively. HBsAg was negative, and anti-HCV was positive, however quantitative HCV RNA (Cobas Amplicor HCV Monitor 2.0®; Roche Diagnostics, Branchburg, NJ, USA) was negative. Soon after the protocol therapy, together with an improvement in symptoms, the patient showed objective responses in all lesions, including the intrahepatic HCC lesions, vascular invasion, and pulmonary metastases (Fig. 3, 4). During the follow-up, the AFP and PIVKA-II levels were elevated without any disease progression evident on radiologic evaluations. After the oral administration of tegafur and uracil (tegafur [100 mg], uracil [224 mg per 0.5 g pack], UFT-E granules, Jeil Pharmaceuticals Co. Ltd, Seoul Korea), both serological HCC biomarkers normalized again (Fig. 3). The patient is alive without any evidence of recurrence after 28 months (DFI, 18 months).

A 66-year-old man who had anorexia and nausea for 1 month was found to have liver cirrhosis and far advanced HCC during an evaluation. The initial dynamic CT scan showed a massive infiltrating HCC with a maximum diameter of 12 cm occupying the right lobe, and complicated by a turmor thrombosis in the main hepatic vein with a large AVS. Small pulmonary metastatic lesions were also identified. The initial laboratory tests were as follows: hemoglobin 15.7 g/dL, WBC 5,000/mm³, platelet 128,000/mm³, albumin, 4.1 g/dL, total bilirubin 1.0 mg/dL, prothrombin time-INR 1.21, ALP 254 IU/L, AST 242 IU/L, ALT 60 IU/L, total cholesterol 190 mg/dL, AFP and PIVKA-II 192,320 ng/mL and >2000 AU/mL, respectively. HBsAg was positive, HBeAg and HBV DNA (Amplicor HBV Monitor®; Roche Molecular System, Pleasanton, CA, USA) were negative, and anti-HCV was negative. After the protocol therapy, together with the improvement in symptoms, the patient showed objective responses completely in the intra- hepatic HCC lesions, vascular invasion, and AVS by radiologic evaluations, and also partial improvement in the extra-hepatic lesions based on AFP and PIVKA-II tests (Fig. 5, 6). However, on serial testing, the levels of serologic HCC biomarkers rose again, and multiple metastases to both lung fields and the right adrenal gland occurred rapidly, in spite of repeated courses of chemotherapy (Fig. 5). The patient died of respiratory failure with septicemia 22 months after registration.

HAIC-EC-associated side effects were observed in all 3 patients (NCI-CTC grade 1 or 2 grade nausea developed during the admission period and grade 2 or 3 leucopenia developed 7-10 days after discharge). PIC-IF-associated side effects were observed in all 3 patients (grade 2 or 3 flu-like symptoms developed 20-30 minutes after the procedure). All side effects were transient and well-tolerated. However, there was a tendency for the degree of bone marrow suppression to be progressively aggravated by the repeated treatments, but no greater than toxicity grade 3. Neither NCI-CTC grade 4 complications nor deterioration in liver and/or renal function developed during repeat protocol therapy.