Treatment of Hepatocellular Carcinoma Using Arterial Chemoembolization with Degradable Starch Microspheres and Continuous Arterial Infusion of 5-Fluorouracil

Objective: Transcatheter arterial embolization (TAE) with gelatin sponge particles and iodized oil often yields poor results when used to treat unresectable multifocal hepatocellular carcinoma (HCC). The present study retrospectively investigated the utility of a novel combination chemotherapy regimen for treating multifocal HCC resistant to TAE. Methods: Thirteen consecutive patients with unresectable multifocal HCC and resistance to TAE were treated with combination chemotherapy consisting of arterial chemoembolization with degradable starch microspheres (DSM) (150–4500 mg on Day 1), mitomycin-C (4–8 mg on Day 1), continuous arterial infusion of 5-ﬂuorouracil (1250 mg/120 h), cisplatin (25–50 mg/120 h) and l-leucovorin (125 mg/120 h) for 10–19 weeks. Results: The response rate was 84.6%, with complete response in one patient and partial response (PR) in 10 patients. In four of 10 patients with PR, the tumor was not observable, although the tumor marker did not completely decline to the normal range. The 1-, 2- and 3-year survival rates were 100, 28.9 and 9.6% in all, and 100, 33.3 and 0% in six patients with portal vein tumor thrombosis (PVTT). The median survival was 22.1 months in all and 17.1 months in six patients with PVTT. Thrombocytopenia of Grade III or higher was observed in eight patients. Laparoscopic splenectomy was performed before therapy in four patients with platelet counts of , 70 000/mm 3 , and during therapy in ﬁve patients with severe thrombocytopenia. Conclusions: This novel chemotherapy regimen achieved favorable results and may be useful in treating patients with unresectable multifocal HCC resistant to TAE.


INTRODUCTION
Hepatocellular carcinoma (HCC) is one of the most prevalent malignancies worldwide, and recent advances in ultrasonography (US), computed tomography (CT), digital subtraction angiography (DSA) and magnetic resonance imaging (MRI) have enabled its early detection. Furthermore, several non-surgical modalities have been developed for the treatment of unresectable HCC, including percutaneous ethanol injection (PEI) (1), microwave coagulation therapy (MCT) (2) and radiofrequency ablation (RFA) (3). As a result, the prognosis for patients with HCC has improved considerably (4). However, there are a considerable number of patients with advanced multifocal HCC at first diagnosis that cannot be treated with these new modalities (4). Moreover, multicentric carcinogenesis and intrahepatic metastases make the recurrence rate for HCC high (5). Since almost all patients with HCC eventually develop multifocal tumors, development of a standard therapy for patients with advanced multifocal HCC would be beneficial.
Owing to the extensive vascularity of most hepatic tumors, transcatheter arterial embolization (TAE) with gelatin sponge particles and iodized oil is a useful alternative for the treatment of advanced multifocal HCC (6). In particular, TAE is useful for nodular type HCCs that possess capsules, and that are supplied by the hepatic artery (7,8) but is not as useful for the following tumors with portal vein flow (8): (i) non-nodular type tumors without capsules, such as massive or diffuse growth types (7,9); (ii) nodular type tumors with incomplete capsules, such as contiguous multinodular (CMN) growth type and single nodule with extranodular (SNE) growth type (7); (iii) tumors with thrombosis in hepatic vessels (9) and (iv) small intrahepatic metastases without capsules (8).
Mitomycin-C (MMC) is a dose-dependent agent with powerful and broad-spectrum anti-cancer effects (23). Use of MMC with degradable starch microspheres (DSM) can increase the anti-cancer effects while reducing the side effects. DSM, which is made from potato starch and produces transient occlusion of small arteries (24,25), has been used with anti-cancer agents for the treatment of metastatic liver cancer and HCC (25 -28). However, combination treatment using chemoembolization with DSM and continuous arterial infusion of 5-FU has not been reported. We previously experienced excellent results in three patients with metastatic liver tumors from primary colorectal cancer, who were treated with a combination of arterial chemoembolization with DSM and MMC and arterial infusion chemotherapy using 5-FU and l-LV [Isovorin, (l-LV)]. Thus, the goal of the present retrospective study was to determine the therapeutic utility of novel chemotherapy using arterial chemoembolization with DSM and MMC, and continuous arterial infusion 5-FU with its biochemical modulators for the patients with TAE-resistant HCC.  Tables 1 and 2, respectively. TAE-resistant HCC was defined by the following criteria: (i) tumors with enlargement of .20% in the sum of the longest diameter of the liver tumors on CT performed 2 months after TAE and (ii) tumors with thrombi in the hepatic vessels (9). Enrollment criteria included (i) age , 70 years; (ii) Child-Pugh classification . 9 points; (iii) absence of extrahepatic metastases; (iv) platelet count . 70 000/mm 3 ; (v) creatinine clearance . 60 ml/h; (vi) successful implantation of intra-arterial catheter and drug delivery system; (vii) absence of the other life-threatening disease, with the exception of chronic liver disease; (viii) performance status [Eastern Cooperative Oncology Group (ECOG)] level 0 -2 (29); and (ix). This retrospective study was conducted in accordance with the Declaration of Helsinki. All patients gave written informed consent.

PATIENTS
Ten patients had received TAE prior to enrollment in this study, and were diagnosed with TAE-resistant HCC by CT imaging 2 months after TAE (Criterion 1). The remaining three patients could not undergo TAE secondary to the presence of tumor thrombosis in the portal vein trunk (Criterion 2). The demographics of these patients are summarized in Table 1. Four patients with platelet counts of ,70 000/mm 3 before chemotherapy underwent laparoscopic splenectomy before entry into the study. Two patients had ascites of Grade 3 before the treatment.

CATHETER IMPLANTATION
Celiac angiography was performed according to the femoral approach under local anesthesia. After the detection of HCC and its feeding arteries, a 5-Fr heparin-coated catheter was introduced into the proper or common hepatic artery. The gastroduodenal artery and the right gastric artery were occluded with metallic coils to prevent gastroduodenal injury by the anti-cancer drugs. Aberrant hepatic arteries, if present, were occluded with a metallic coil or a mixture (1:1.5) of n-butyl cyanoacrylate and iodized oil prior to the treatment for hepatic arterial redistribution (30). After the catheter was connected to the injection port, the device was implanted in a subcutaneous pocket in the femoral site. No patients received anticoagulant therapy. The device was filled with 5000 IU of a heparin solution to prevent occlusion. Before treatment, vascularity of tumors and enhanced area by contrast medium, were confirmed using CT angiography through an implanted reservoir catheter in all patients.
First, a mixture of DSM (150 -1800 mg) and MMC (4 -8 mg) in the same volume of contrast medium was injected via the implanted port on Day 1, until hepatic arterial flow stagnated. One hour after this chemoembolization,

598
Novel chemotherapy for TAE-resistant HCC continuous intra-arterial infusion of 5-FU (1250 mg), CDDP (25 -50 mg) and l-LV (125 mg) for 120 h was performed using three disposable balloon infusers (Surefuser w ; Nipro Co., Osaka, Japan) and three 2-valves, in principle, as outpatients. In addition, this intra-arterial infusion was repeated on Day 7. A serotonin antagonist, ramosetron hydrochloride (Nasea w ; Yamanouchi Co., Tokyo, Japan), was administered intravenously or orally as antiemetic therapy. This course was repeated after the suspension of administration for 2 weeks or more. All anti-cancer therapy was discontinued when adverse effects reached level 2 of the National Cancer Institute-Common Toxicity Criteria (NCI-CTC), Version 2.0 (31) (with the exception of platelet and leukocyte counts ,50 000 and 2000/mm 3 , respectively, because of underlying cirrhosis).

ASSESSMENT OF THERAPEUTIC EFFICACY
To evaluate the response to chemotherapy, serum chemistry and tumor markers, including AFP and DCP were measured at least once every month. The normal values of AFP and DCP were defined as ,20 ng/ml and 40 mAU/ml, respectively. In addition, abdominal CT or dynamic MRI was also performed before and after the treatment at least once per month. The objective response and adverse events of the therapy were evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) (32) and NCI-CTC, Version 2.0, respectively. Cumulative survival was estimated by the Kaplan -Meier method and the log-rank test using SPSS version 14.0J software (SPSS Japan Inc).

CLINICAL EFFECTS
Therapeutic response, change in tumor marker after chemotherapy and patient's outcome are shown in Table 2. Of the 13 consecutive patients who received this therapy, complete response (CR) was observed in one patient, and partial response (PR) was observed in 12 patients. Thus, the response rate (CR or PR) was 86.7%. Two patients developed lung or lymph node metastases despite regression of liver tumors, and were regarded as having progressive disease (PD). In other words, tumor regression in the liver was observed in all patients.
In four of 10 patients with PR, CT or DSA imaging demonstrated the absence of tumor after administration of anti-cancer agents for 10 -13 weeks (mean 11.5 weeks). However, the tumor marker did not decrease within the normal range, and we regarded them as having PR.
Serum AFP values were higher than 20 ng/ml in 11 patients before chemotherapy and decreased in all 11 patients after chemotherapy, with two patients having AFP values of ,20 ng/ml. Serum DCP values were .40 ng/ml in 11 patients prior to chemotherapy and decreased in all 11 patients after chemotherapy, with seven patients having DCP values of ,40 ng/ml. The 1-, 1.5-, 2-and 3-year survival rates of all 13 patients were 100, 61.5, 28.9 and 9.6%, respectively, and the median survival was 20.4 months. In six patients with PVTT, the 1-, 1.5-, 2-and 3-year survival rates were 100, 50.0, 33.3 and 0%, and the median survival was 17.1 months. Cumulative survival was not significantly different between six patients with PVTT and seven patients without PVTT (P ¼ 0.3926). In 11 patients with therapeutic response, the 1-, 1.5-, 2-and 3-year survival rates were 100, 72.7, 34.1 and 11.4%, and the median survival was 22.1 months. Cumulative survival of 11 patients with therapeutic response significantly differed in comparison with two patients without therapeutic response (P ¼ 0.0399).

ADVERSE EVENTS
Adverse events are summarized in Table 3. Thrombocytopenia of greater than Grade III was observed in eight patients (61.5%), and two patients (15.4%) required platelet transfusions. Leukopenia of greater than Grade III was observed in three patients (23.1%), and two patients (15.4%) received granulocyte colony stimulating factor. Therapy was discontinued for more than 2 weeks in eight patients (61.5%) secondary to thrombocytopenia, and five of eight patients underwent laparoscopic splenectomy so that therapy could be reinitiated. At the time of embolization with DSM, Jpn J Clin Oncol 2008;38 (9) 599 most patients complained of upper abdominal pain or discomfort that was attributed to hepatic ischemia, but the pain was resolved following the intravenous administration of pentazocine hydrochloride, and disappeared spontaneously within 1 h. None of the patients experienced direct hepatotoxicity from the chemotherapy. Other adverse events included duodenal ulcer (three patients), and infection of the reservoir port (two patients). None of the patients experienced high fever or renal injury.

CASE PRESENTATIONS A PATIENT WITH CR (CASE 3)
A 67-year-old man was diagnosed with HCC in 1989 and underwent treatment with PEI and TAE. Beginning in May 2001, the patient underwent repeat TAE for the treatment of multifocal HCC. However, the main CMN growth type tumor in subsegment 6 showed persistent growth with many small metastases in the bilateral liver lobes ( Fig. 2a and b). Thus, these tumors were regarded as TAE-resistant HCC, and the patient underwent this chemotherapeutic regimen. Six months after the initiation of chemotherapy, the main tumor and intrahepatic metastases were no longer present, and plasma levels of tumor markers showed a marked decline ( Fig. 2c and d).
A PATIENT WITH PR, WHO UNDERWENT LAPAROSCOPIC SPLENECTOMY AGAINST SEVERE THROMBOCYTOPENIA AFTER TREATMENT (CASE 11) A 50-year-old man presented in October 2003 with complaints of abdominal fullness. He was diagnosed with advanced diffuse type HCC that occupied nearly the entire right lobe and part of the left lobe (Fig. 3a). Arterial portography suggested the presence of tumor thrombosis within the major trunk and the right branch of the portal vein; however, the left branch was observed via cavernous  A 50-year-old man had diffuse type HCC occupying more than three-quarters of the liver on CT image (a). Arterial portography suggested tumor thrombi in the major trunk and the right branch of the portal vein (arrowhead); however, the left branch was observed through cavernous transformation (b). After chemotherapy for 6 weeks, tumor had regressed markedly with small residual lesions (arrowheads) and disappearance of portal thrombi (c, d). According to the National Cancer Institute-Common Toxicity Criteria (NCI-CTC), Version 2.0.

600
Novel chemotherapy for TAE-resistant HCC transformation (Fig. 3b). The patient was treated with this chemotherapeutic regimen; however, therapy was interrupted after 3 weeks because of a platelet count ,50 000/mm 3 . Laparoscopic splenectomy was performed, and therapy was reinitiated 1 week later, as platelet counts had increased to .150 000/mm 3 . After 6 weeks of chemotherapy, tumor size had regressed markedly, portal thrombi disappeared and tumor marker levels decreased (Fig. 3c and d). The patient was considered to have PR. He died due to progression of extrahepatic metastases 13.8 months after initiation of chemotherapy, and not of liver tumor.

DISCUSSION
TAE is a widely used therapeutic modality for unresectable multifocal HCC and has an excellent anti-tumor effect (6). However, the survival benefits of TAE remain controversial (4) because TAE may promote deterioration of liver function in patients with cirrhosis (33), and because HCC is sometimes resistant to the TAE treatment. In histopathologic studies using resected liver, TAE induces a high degree of necrosis in encapsulated expanding growth type tumors but not in growth type tumors without capsules and in small daughter nodules or intrahepatic metastases (7 -9). Hashimoto et al. (7) reported that TAE induced complete necroses in 50% of single nodular type tumors, 21% of SNE growth type tumors and 9% of other types of tumors, including massive, multiple nodular or CMN growth type. TAE resistance may be mediated by blood supply from the portal vein in the tumor boundary (7,8). Sato et al. (9) reported that survival in patients with unresectable HCC and either portal thrombosis or tumor occupying .20% of the liver does not differ between TAE and conservative treatment. Furthermore, Yamashita et al. (34) reported that the tumor type was the most important prognostic factor for determining survival after TAE. Treatment of TAE-resistant HCC is a major clinical objective, and HCC has low sensitivity to several anti-cancer agents (10 -14). Several interventions were used in the present study to enhance the anti-cancer effect of 5-FU for HCC treatment. First, an implanted reservoir system was used for arterial infusion. Use of this system allows for relatively easy replenishment of administering agents into the hepatic artery (17 -22) and achieves higher intrahepatic concentrations while reducing toxicity to other organ systems (35). Because 5-FU is a time-dependent agent and exerts a stronger cytocidal effect with longer exposure to cells (36), the drug was administered in the present study by means of continuous 120-h arterial infusion. Second, low-dose CDDP and l-LV were also administrated for biochemical modulation of 5-FU (37 -39). 5-FU inhibits DNA synthesis by inactivation of thymidylate synthase (TS), which is caused by the formation of a complex of 5-fluoro-2 0 -deoxyuridine 5 0 -monophosphate (FdUMP) and methylene tetrahydrofolate (CH2FH4) (39). CDDP can increase the intracellular pool of CH2FH4 and tetrahydrofolate (39), whereas LV is a source of these increased intracellular folates (38). Both CDDP and LV can amplify complex formation of TS, FdUMP and folates. Ando et al. (17) reported that chemotherapeutic regimens using 5-FU and low dose CDDP (low dose FP) showed an objective response in 44.4% of HCC with portal tumor thrombosis of the main trunk. Furthermore, Yamasaki et al. (21) reported significantly better response and survival rates with low dose FP plus LV than with the same regimen without LV.
In the present study, chemotherapy was combined with arterial chemoembolization using DSM and MMC. DSM, a microembolization material 45 + 7 Â 10 26 m in diameter made from potato starch, induces transient occlusion of small arteries (25,26). Since DSM is degraded by serum amylase, the duration of occlusion in the hepatic arteries by DSM is limited to 80 min (40). Thus, embolization with DSM seldom injures the drug delivery route, including the hepatic artery. Furthermore, recruitment of parasitic arteries, such as the phrenic or the omental arteries, is lower with DSM embolization compared with conventional TAE with gelatin sponge particles and iodized oil. In addition, co-administration of an anti-cancer drug with DSM reduces drug dilution by the hepatic arterial flow and enhances drug retention in tumors at a higher concentration and for a longer time than a single bolus injection (25,40). Furuse et al. (41) reported that embolization with DSM alone produced a necrotic area of 50% in 35.3% of patients with HCC. In the present study, the hepatic artery was embolized with DSM until blood flow stagnated, which should have produced antitumor effects from embolization alone, and increased the concentration of MMC within the liver tumors.
Surprisingly, our chemotherapy showed a high rate of objective response (84.7%, 100% only as for liver lesion). All patients, including six cases with PVTT, survived for over 1 year. The prognosis of HCC invading the portal vein is extremely poor, and survival time is generally limited to several months after diagnosis (6). Thus, our regimen appears to have had a powerful anti-cancer effect and produced a significant improvement in survival. However, the number of patients in this study was relatively small, and further study with more number of patients is needed to confirm the efficacy and safety of this therapy.
The chemotherapy described in the present study has several limitations. First, there was a high incidence of thrombocytopenia. In most of the patients with HCC, platelet count is relatively low due to underlying cirrhosis. Therefore, severe thrombocytopenia tends to occur easily as a result of receiving chemotherapy. We experienced marked tumor progression in some patients, who needed long interruption of treatment from thrombocytopenia. Therefore, we performed laparoscopic splenectomy when the prechemotherapy platelet count was ,70 000/mm 3 , or when an interruption in chemotherapy for more than 2 weeks was required due to the thrombocytopenia that developed after treatment. In most cases, therapy could be reinitiated $1 Jpn J Clin Oncol 2008;38 (9) 601 week after splenectomy. Second, problems may occur with the use of the implanted reservoir system. For example, three patients developed duodenal ulcers due to recanalization of gastroduodenal arteries. Therefore, we advocate establishing complete obstruction of the right gastric or the duodenal arteries to prevent gastroduodenal injury. Third, the chemotherapy regimen is relatively complicated; treatment requires three arterial drug infusions in addition to chemoembolization with DSM using fluoroscopic imaging. We used three disposable balloon infusers and three two-valves for completing continuous arterial infusion of three drugs. This method could enable patients to receive this treatment only once per week as outpatients.
With the present regimen, transaminases did not increase and liver failure did not occur, and functional status (in terms of activities of daily living) of all patients remained relatively good. In fact, functional status was higher for patients treated with the present regimen compared with patients treated with TAE treatments alone. Furthermore, this regimen did not induce post-embolization syndrome, including high fever and liver failure (34,41).
In conclusion, this novel chemotherapy regimen produced good responses in patients with unresectable multifocal HCC resistant to TAE.