Transcatheter Arterial Embolization with Only Particles for the Treatment of Unresectable Hepatocellular Carcinoma

Transcription

1 Transcatheter Arterial Embolization with Only Particles for the Treatment of Unresectable Hepatocellular Carcinoma Mary A. Maluccio, MD, MPH, Anne M. Covey, MD, Leah Ben Porat, PhD, Joanna Schubert, MD, Lynn A. Brody, MD, Constantinos T. Sofocleous, MD, George I. Getrajdman, MD, William Jarnagin, MD, Ronald DeMatteo, MD, Leslie H. Blumgart, MD, Yuman Fong, MD, and Karen T. Brown, MD PURPOSE: To determine the survival of patients with hepatocellular carcinoma (HCC) treated with a standardized method of transcatheter arterial embolization (TAE) with small embolic particles intended to impart terminal vessel blockade, and to evaluate prognostic factors that impact overall survival. MATERIALS AND METHODS: A total of 322 patients with HCC who underwent 766 embolizations from January 1997 to December 2004 were retrospectively reviewed. Selective embolization of vessels feeding individual tumors was performed with small (50 m) polyvinyl alcohol or spherical embolic particles ( m) intended to cause terminal vessel blockade. Repeat embolization was performed in cases of evidence of persistent viable tumor or development of new lesions. Patient, tumor, and treatment characteristics were prospectively recorded and tested for prognostic significance by univariate and multivariate analysis. RESULTS: The median survival time was 21 months, with 1-, 2-, and 3-year overall survival rates of 66%, 46%, and 33% respectively. In patients without extrahepatic disease or portal vein involvement by tumor, the overall 1-, 2-, and 3-year survival rates increased to 84%, 66%, and 51%, respectively. Okuda stage, extrahepatic disease, diffuse disease >5 ( tumors), and tumor size were independent predictors of survival on multivariate analysis. There were 90 complications (11.9%) in 75 patients, including eight deaths (2.5%), within 30 days of embolization. CONCLUSIONS: Hepatic arterial embolization with small particles to cause terminal vessel blockade is an effective treatment method for patients with unresectable HCC. These data support our hypothesis that particles alone may be the critical component of catheter-directed embolotherapy. J Vasc Interv Radiol 2008; 19: Abbreviations: HCC hepatocellular carcinoma, HR hazard ratio, OR odds ratio, PV portal vein, PVA polyvinyl alcohol, TACE transcatheter arterial chemoembolization, TAE transcatheter arterial embolization TREATMENT of liver tumors by arte-irial embolization was initially proposed ling symptoms and local tumor growthtrials (11,12) reported a significant im- the late 1970s with the goal of control- When two randomized controlled by cutting off the tumor blood supplyprovement in survival of patients with (1 3). Our understanding that tumors hepatocellular carcinoma (HCC) treated From the Department of Surgery (M.A.M.), Indiana derive most of their blood supply fromwith transcatheter arterial chemoembo- (TACE) versus no treatment, University, Bloomington, Indiana; and Departments the hepatic artery, while normal liver islization of Radiology (A.M.C., J.S., L.A.B., C.T.S., G.I.G., supplied primarily by the portal venoustace emerged as the de facto standard K.T.B.), Biostatistics and Epidemiology (L.B.P.), and system (4), adds credence to the conceptof care for unresectable disease. Many Surgery (W.J., R.D., L.H.B., Y.F.), Memorial Sloan- Kettering Cancer Center, 1275 York Avenue, New of arterial embolization as a therapeuticdifferent chemotherapeutic agents have York, NY Received August 15, 2007; final procedure for vascular tumors. Series been used alone and in combination in revision received February 12, 2008; accepted February 14, Address correspondence to K.T.B.; from the late 1970s and early 1980s 5 7) ( TACE. Some practitioners use a set dose brown6@mskcc.org highlighted the use of intraarterial infu-osion of chemotherapeutic drugs directly others base the dose on tumor volume, chemotherapeutic agent, whereas None of the authors have identified a conflict of into the liver for primary and metastaticunderlying liver disease, or unique pa- comorbidities. Despite the routine interest. liver tumors, and from this, the additiontient SIR, 2008 of chemotherapy to arterial embolization was popularized 8 ( 10). based data to support it as the use of TACE, there are no evidence- DOI: /j.jvir standard 862

2 Volume 19 Number 6 Maluccio et al 863 treatment for unresectable HCC because no definitive benefit of any TACE regimen versus arterial embolization alone has been shown (13). Based on a current literature review, data are limited in terms of outcomes with the use of standardized technique and contemporary bland embolization particles for HCC. At the authors institution, transcatheter arterial embolization (TAE) is performed rather than TACE based on the hypothesis that the primary effect of TACE is a result of terminal vessel ischemia rather than chemotherapyinduced cytotoxicity. In our early experience with TAE with gelatin sponge particles and large polyvinyl alcohol (PVA) particles (14), we reported 1- and 2-year actual survival rates of 50% and 33%, respectively, with a median survival time of 11.7 months. These data were similar to those from contemporary series that used TACE with similar embolization agents (15,16). Since that time, we have standardized our technique to minimize interprocedural variability and improve patient selection, and we have maximized terminal vessel ischemia by adopting the use of small spherical embolic particles when they became commercially available in the late 1990s. The purposes of this study are to evaluate the overall survival of patients treated with TAE according to a standard technique and to evaluate patient and tumor variables that predict outcome. MATERIALS AND METHODS All data reported were collected prospectively and maintained in a Health Insurance Portability and Accountability Act compliant database. An institutional review board waiver was obtained to use patient information for this retrospective study. Patient and Disease Characteristics A total of 766 embolizations were performed on 322 patients at a single institution between 1997 and Patient characteristics, indication for treatment, and disease variables are shown in Tables 1 and. 2 Table 1 Demographic Information Variable Preprocedural Evaluation Value Sex Male 231 (72) Female 91 (28) Race White/not Hispanic 207 (64) Asian 81 (25) Other 34 (11) Age at embolization (y) Mean 69 Median 67 Range Okuda Stage I 173 (54) II 125 (39) III 4 (1) Missing 20 (6) Survival status Dead of disease 157 (49) Alive with disease 82 (25) Lost to follow-up 29 (9) No evidence of disease 11 (3) Dead other/unknown 43 (13) Note. Values in parentheses are percentages. Before treatment, cases were discussed at a multidisciplinary hepatobiliary disease management conference. All patients had preprocedural laboratory studies, including complete blood count, liver function tests, -fetoprotein measurement, and coagulation profile. Triple-phase computed tomography (CT) or gadolinium-enhanced magnetic resonance (MR) imaging was performed within 30 days before embolization. Our exclusion criteria are based on the M. D. Anderson criteria (17), including uncompensated cirrhosis (Child class C), untreated significant biliary obstruction, or a constellation of lactate hydrogenase level greater than 425 U/L, bilirubin level greater than 2.0 mg/dl, aspartate aminotransferase level greater than 100 U/L, and more than 50% of the liver replaced by tumor. Patients with extrahepatic disease were included if the disease was liver-dominant. No patients with extrahepatic disease were being treated with chemotherapy or radiation therapy at the time of embolization. Table 2 Characteristics at First Embolization Variable Data Collection Value Indication for embolization Primary treatment 312 (97) Bleeding 6 (2) Pain 4 (1) Reason for unresectability Parenchyma 193 (60) Bilateral disease 62 (19) Comorbidities 25 (8) Inflow or outflow 30 (9) Caval involvement 7 (2) Unknown 5 (2) PV occlusion None 247 (77) Right 39 (12) Left 16 (5) Bilateral 8 (2) Main 12 (4) Tumor-occupied liver (%) (42) (26) (14) (13) Unknown 18 (6) Extrahepatic disease Absent 180 (56) Present 67 (21) Indeterminate 75 (23) Number of tumors (34) 2 46 (14) 3 31 (10) 4 25 (8) 5/diffuse 110 (34) Note. Values in parentheses are percentages. Patient demographics, tumor characteristics, previous treatment, recurrence, and survival data were recorded. The number of tumors, size of the target lesion, volume and type of particles used, and vessels embolized were recorded. Alpha-fetoprotein levels were not recorded. Embolization We have described our technique previously (18). Hepatic angiography is performed from a common femoral approach with a 4- or 5-F angiographic catheter to establish hepatic arterial anatomy, location of lesion(s), and portal vein (PV) patency. All vessels supplying the target tumor are then embolized as selectively as possible

3 864 TAE with Only Particles for Unresectable HCC June 2008 JVIR with PVA, Embosphere microspheres (Biosphere Medical, Rockland, Massachusetts), or a combination of the two depending on availability and operator preference. Embolization is performed most often with use of a microcatheter and until there is complete stasis in the parent vessel. Segmental or subsegmental embolization is performed when possible. Staged procedures were used in patients with bilobar disease or large ( 10 cm) tumors. Whether a staged procedure was planned was not recorded (Table 3). Fifty-micrometer nonspherical PVA particles (Cook, Bloomington, Indiana) were used exclusively in the earlier years of the study ( ), and later m tris-acryl gelatin microspheres were more commonly used to impart terminal vessel blockade within the tumor. Larger PVA particles ( m) were used when an arteriovenous shunt was identified or in larger tumors ( 5 cm) to minimize the risk of particles passing into the systemic circulation (19). No chemotherapeutic agent or Ethiodol (Savage Laboratories, Melville, New York) was added. Patients were given antiemetic and antibiotic drugs before the procedure and for 24 hours afterward. Patients with a biliary-enteric anastomosis or biliary drainage catheter or stent, or who had undergone sphincterotomy, received piperacillin/tazobactam to minimize the risk of infectious complications (20). After embolization, patients were admitted for pain control and intravenous hydration until they were able to eat. If a patient developed a fever greater than 38.5 C with leukocytosis, blood cultures were drawn and intravenous antibiotics initiated or continued until the final cultures returned. Patients were not prescribed oral antibiotics on discharge unless the blood cultures were positive. Imaging Angiography was performed to identify the primary blood supply to the tumor(s) and after the embolization to document occlusion of target vessels and absence of additional tumor blood supply. Similar to TACE with Ethiodol, during bland embolization, contrast medium laden particles stain hypervascular tumor, verifying appropriate coverage with the embolic Table 3 Embolization Variables for 322 Patients Who Underwent 766 Embolization Procedures Variable Value Embolizations (43) 2 77 (24) 3 50 (16) 4 58 (18) Type of embolization Lobar 181 (24) Segmental 413 (54) Subsegmental 172 (22) Particulate agent PVA 371 (48) Embospheres 290 (38) Embospheres and PVA 105 (14) Note. Values in parentheses are percentages. agent. Apparent defects in lesion coverage prompt angiographic search for intra- or extrahepatic collateral supply. Contrast medium enhanced CT or MR and blood work were performed 4 6 weeks after embolization. When evidence of viable tumor was identified, repeat embolization was performed. Patients who had a poor radiographic response to embolotherapy after two consecutive sessions were not treated with embolization again. Additional procedures were recommended if there was evidence of tumor enhancement on the arterial phase of imaging. If no repeat intervention was required, patients underwent imaging every 3 months to assess durability of response and evaluate for the possibility of repeat embolization. Morbidity and Mortality The National Cancer Institute Common Terminology Criteria for Adverse Events (version 3.0) (21) were used to categorize complications. For interventional procedures, grade 1 complications are those that require no intervention and include asymptomatic vascular complications. Grade 2 events require bedside medical management and would include bacteremia and arrhythmias requiring medication. Grade 3 complications require additional intervention, such as vascular events requiring stent placement or cardiovascular compromise requiring extended therapy. Grade 4 complications result in chronic disability, and a Grade 5 complication is a periprocedural death. Data Analysis Survival time is defined as the time from the date of first embolization to the date of death or last follow-up. Survival curves were estimated with Kaplan-Meier methodology. Death from any cause was censored as an event. Kaplan-Meier estimates of survival time in various groups were compared with use of the log-rank test in the case of discrete variables or Cox proportional-hazards model in the case of continuous variables. A significance level of 15% in the univariate setting was used as the criterion to include a variable in the multivariate modeling procedure. Stepwise regression was used to determine variables incorporated into the multivariate Cox regression model. The associations between covariates were assessed with use of the Fisher exact test. RESULTS Overall Survival The median follow-up duration was 20 months (range, months) from the time of first embolization. Of the 322 patients, 200 (62%) have died: 157 died of disease or worsening underlying liver disease and 43 died of other or unknown causes. Ninetythree patients (29%) are alive, 82 with and 11 without evidence of viable disease. Twenty-nine patients (9%) were lost to follow-up. The median survival from the date of first embolization was 21 months (95% CI, months). The 1-, 2-, and 3-year overall survival rates for the 322 patients were 66%, 46%, and 33%, respectively (Fig 1). Patients without evidence of PV involvement survived significantly longer than those with PV involvement, with median survival times of 28 months and 11 months, respectively (P.001; Fig 2). Patients with disease in only the liver survived significantly longer than patients with suspicious lesions or biopsy-proven metastatic disease, with median survival times of 33.7 months, 14.5 months, and 7.9 months, respectively (P.001). Patients with solitary lesions survived significantly longer than patients with

4 Volume 19 Number 6 Maluccio et al 865 Figure 1. TAE. Overall survival of patients with HCC who underwent Figure 2. Overall survival of 159 patients with HCC without PV occlusion or extrahepatic disease (78 dead of disease). Median survival time was 40 months (95% CI, months) and median follow-up for survivors was 27 months. two to four lesions or multifocal disease ( 5 or diffuse), with median survival times of 35.4 months, 26.5 months, and 11.3 months, respectively (P.001). Previous randomized studies, including recent studies by Lo et al (11) and Llovet et al (12), excluded patients with evidence of extrahepatic disease or PV involvement by tumor. Of the 322 patients in the present study, 159 patients had no evidence of PV involvement and no evidence of extrahepatic disease. This select group of patients had a median survival of 40 months (95% CI, months) and 1-, 2-, and 3-year survival rates of 84%, 66%, and 51%, respectively (Fig 2). Prognostic Factors on Univariate Analysis Sex, age, race, and prior resection did not predict survival on univariate analysis (Table 4). Okuda stage, presence of extrahepatic disease, PV involvement, number of tumors, and size of largest lesion were all significant predictors of survival. Prognostic Factors on Multivariate Analysis Okuda stage (I vs II), presence of extrahepatic disease, number of tumors, and size of the largest tumor ( 5 cmvs 5 cm) were all independent predictors of survival on multivariate analysis. Patients with Okuda stage II liver disease had a twofold increased risk of death versus patients with well-compensated liver function (Okuda stage I disease; hazard ratio [HR], 2.1; P.001). Patients with extrahepatic disease at the time of the initial embolization had a threefold increased risk of death versus patients with disease confined to the liver (HR, 3.2; P.001). Patients with five or more tumors or diffuse disease had a twofold risk of death versus patients with solitary tumors or two to four discrete tumors (HR, 1.8; P.002). The presence of PV occlusion was not an independent predictor of survival on multivariate analysis (Table 5). Complications Ninety complications occurred in 75 patients (11.9% of 766 procedures). Grade 1 complications included 12 cases of prolonged postembolization syndrome in 10 patients. Two patients each experienced two episodes of prolonged postembolization syndrome. Two patients had asymptomatic external iliac artery dissections and one patient had a dissection of the common hepatic artery that required no treatment. One patient had a groin hematoma and another developed skin changes from nontarget embolization of the phrenic artery. Grade 2 complications included nine patients who developed bacteremia, 17 patients with 19 episodes of cardiac arrhythmia, two patients who had a vasovagal response, and five who experienced pulmonary edema. There were 21 grade 3 complications: three had stent placement for vascular dissection, three had myocardial infarctions, and nine had contrast agent induced nephropathy. Nontarget embolization occurred in six patients, three developed cholecystitis, and three developed pancreatitis. Nine patients experienced liver function decompensation (grade 4). All these patients were treated for their liver insufficiency with appropriate medications. There were eight deaths. Four patients developed progressive liver failure and died of liver failure within 30 days of their procedure. Two other patients died within 7 days of the procedure as a result of microspheres shunted into the pulmonary circulation. One patient had a fatal myocardial infarction, and one died at home of unknown causes 9 days after embolization. Based on the World Health Organization classification of procedural morbidity and mortality (22), there were 38 major complications in 766 procedures (5%). DISCUSSION Much of what we know about particle embolization for the treatment of

5 866 TAE with Only Particles for Unresectable HCC June 2008 JVIR Table 4 Univariate Survival Analysis Variable No. of Pts. Pts. Who Died Median Survival (95% CI) P Value* Sex Male ( ).35 Female ( ) Race White ( ).32 Asian ( ) Other ( ) Tumor-occupied liver (%) ( ) ( ) ( ) ( ) Okuda stage I ( ).001 II (9 18.1) Extrahepatic disease Absent ( ).001 Indeterminate ( ) Present ( ) PV occlusion Absent ( ).001 Right/left/bilateral ( ) Number of tumors ( ) ( ) 5/Diffuse ( ) Size before embolization (cm) ( ) ( ) ( ) * P values for discrete variables were computed with the log-rank test. P values for continuous variables were computed with a Cox regression model. Hazard ratio. Table 5 Multivariate Survival Analysis Variable No. of Pts. HR (95% CI) P Value Okuda stage I II ( ) Extrahepatic disease Absent Indeterminate/present ( ) No. of tumors ( ) 5/diffuse ( ) Lesion size before embolization (cm)* ( ) PV occlusion Absent Right/left/bilateral ( ) * Size represents the length of the tumor s greatest diameter. HCC is based on small prospective series (23) and randomized trials (24,25) evaluating particle embolization versus untreated controls. These studies were performed with gelatin sponge and/or coils that cause proximal vessel occlusion, and a significant proportion of patients had advanced liver disease (Okuda II/III or Child class B/C cirrhosis). To be most effective, we believe that particle embolization should result in terminal vessel occlusion to maximize ischemic tumor necrosis. In the early 1990s, gelatin sponge might have been the best available embolic agent, but that is no longer the case with the availability of small spherical embolic agents that cause terminal vessel blockade much more effectively than gelatin sponge or even nonspherical PVA particles. It is well known that proximal vessel occlusion results in near-instantaneous recruitment of intraparenchymal collateral flow, reconstituting the distal vasculature. Although more proximal occlusion might be adequate to prevent washout of an emulsion of chemotherapeutic agent and Lipiodol, it is suboptimal to produce ischemic tumor cell death. The results of early particle embolization studies are valid for embolization with gelatin sponge fragments, but they do not address the utility of bland embolization performed to effect terminal vessel blockade. Recent theories on the impact of arterial embolization induced ischemia and angiogenesis have shown mixed results in animal models. Extrapolating from studies of ischemia/reperfusion injury in the liver suggests that embolizationinduced ischemia would result in a compensatory increase in angiogenic factors that would promote tumor growth. Indeed, proximal ischemia induced by ligation of the hepatic artery in a rat model showed significant increases in vascular endothelial growth factor and matrix metalloproteinase 1 levels in the serum and liver tissue (26). Conversely, in a study of embolization with PVA in rats with liver tumors (27), only higher doses of PVA induced significant changes in angiogenesis factors within the tumors. A third preclinical model failed to show any difference in microvessel density or angiogenic factors after embolization (28). Selective and/or superselec-

6 Volume 19 Number 6 Maluccio et al 867 Table 6 Contemporary Trials of Chemoembolization for HCC (11,12,32 35) Survival (%) Trial Treatment Regimen Mean Sessions 1y 2y 3y Liver Disease Kawai et al (32,33) TACE NS 70% Child A Epirubicin (60 mg/m 2 ) and Lipiodol NS 64% Okuda I Groupe d Etude et de Traitement du Carcinome Hépatocellulaire (34) Lo et al (11) Llovet et al (12) Doxorubicin (40 mg/m 2 ) and Lipiodol NS 31% Okuda II TACE: cisplatin 70 mg and Lipiodol NS 100% Child A (N 47) 90% Okuda I TACE: cisplatin 30 mg and Lipiodol (N 40) TACE: doxorubicin mg/m 2 and Lipiodol (N 40) % Okuda I 53% Okuda II % Child A 65% Okuda I 35% Okuda II Patients with PVT excluded NS 79*/42 52*/21 NS 93% Child A 29% PVT Chen et al (35) TACE: epirubicin 50 mg/m 2, mitomycin 8 mg/m 2, and Lipiodol (N 216) Current series TAE (N 322) 2 84/66 66/46 51/33 67% Okuda I 33% Okuda II 23% PVT Note. NS not specified. * Among patients with Child class A liver disease. Among patients with Child class B liver disease. Patients without extrahepatic disease or PV invasion/overall series of 322 patients. tive techniques with terminal vessel blockade as described may in fact be the compromise between proximal occlusion with adaptive angiogenesis and adequate ischemia to the target tissue. Two recent randomized trials (11, 12) that evaluated TACE versus no treatment and showed a significant improvement in survival at 2 years are often credited with granting the legitimacy of TACE. The Barcelona Trial (12), published in 2002, set out to compare survival in patients treated with embolization (with gelatin sponge) or TACE (with doxorubicin plus gelatin sponge) with that of patients receiving supportive care. When a sequential analysis showed a significant improvement in survival in patients treated with TACE versus untreated controls, the study was terminated. At the time, there was no significant difference between the TAE and TACE groups. In addition, the study was analyzed on an intent-to-treat basis, and two patients in the embolization group did not receive any treatment, which, given the small group (N 37), makes the study even less powered to detect a difference between the two techniques. The opportunity to address the question of TAE versus TACE with doxorubicin was lost. Although the authors concluded that TACE should become the standard of care for unresectable HCC (12), their data leave open the question of whether TACE provided any additional benefit versus what might have been seen with bland embolization alone, even with only gelatin sponge. Metaanalyses on the subject have been performed to investigate the efficacy of regional therapies in the treatment of HCC (29,30). Camma et al (29) compared overall 2-year mortality odds ratios (ORs) in the context of different embolization procedures. To address the comparison of TAE versus no treatment, they pooled seven studies involving 509 patients and found a significant reduction in 2-year mortality rate favoring the treatment group. To address TACE versus controls (patients who received nonactive treatment), they pooled 13 studies involving 1,777 patients and illustrated a similar reduction in 2-year mortality rate in the treated patients. To address TACE versus TAE, they pooled 16 studies involving 1,930 patients and found no difference in 2-year mortality rates between these treatment groups. An evaluation of the randomized trials of arterial embolization compared with other forms of treatment (ie, arterial chemotherapy or tamoxifen) by Llovet et al (31) pooled TAE and TACE into the treatment groups and concluded that patients who were treated with arterial embolization had improved survival compared with those treated with other therapies. A sensitivity analysis pooling the TAE groups (three studies and 215 patients) failed to show the same overall survival improvement. However, this finding is not relevant to this current series for the following reasons: (i) The sensitivity analyses did not compare TAE versus TACE, nor has any significant difference in overall survival been borne out by any randomized trial; and (ii) the inclusion of patients from early series (1988 and 1994) biases the results in favor of TACE series, which include patients treated with more contemporary embolic agents.

7 868 TAE with Only Particles for Unresectable HCC June 2008 JVIR Our goal in the current study is not to suggest that TAE is superior to TACE. We want only to highlight the fact that, as a community, we began to focus more on the drugs used than on the embolization technique when there are no evidence-based data to demonstrate that the addition of chemotherapy adds any survival benefit versus particle embolization alone. Our survival figures are similar to those in many recent series of patients undergoing TACE with a variety of chemotherapeutic agents (Table 6) (11,12,32 35), despite the fact that, in many cases, our criteria include patients with more extensive disease. The addition of chemotherapy to embolization increases the cost of the procedure and may result in an increase in morbidity. Only one metaanalysis (29) evaluated the periprocedural morbidity and mortality of regional therapies. Periprocedural mortality was significantly higher with the administration of a chemotherapeutic agent (OR, 2.69), the number of courses of chemotherapy administered (OR, 1.50), and the treatment of patients with PV thrombosis (OR, 3.24). In contrast, the use of embolizing material alone had no impact on periprocedural mortality (OR, 1.36), and mortality was significantly lower in studies that used selective (segmental or subsegmental) techniques compared with those that used nonselective techniques (OR,.014). Our standard approach is to use selective or superselective embolization techniques whenever possible. Our univariate survival analysis shows that predictors of survival in TAE with small embolic agents are similar to those variables that predict survival in TACE series, and include Okuda stage, PV occlusion, number of tumors, presence of extrahepatic disease, and tumor size. On multivariate analysis, patients with multifocal disease ( 5 tumors or diffuse disease) and extrahepatic disease did poorly. In our series, multivariate analysis did not show PV occlusion to be an independent predictor of survival, as has been observed in other contemporary TACE series (32,33). However, PV occlusion was an exclusion criterion in the prospective randomized trials from Spain and Hong Kong (11,12), and portal involvement was found to be significant in predicting survival in a report by Takayasu et al (34) of 8,510 patients treated with TACE in Japan. On the basis of our data, we agree with those authors who conclude that PV involvement should not preclude hepatic embolization (36 38). These patients should be the target population for future trials investigating novel therapies. The metaanalyses on embolization illustrate the heterogeneity of patients enrolled in studies and the wide variation in the TACE protocols used. Without any standard technique, comparing studies is difficult. The strength of the present study is that it provides data on a large series of patients treated with particle embolization according to a standard embolization technique. By minimizing interprocedural variability, the data are more likely to reflect survival influenced by tumor and patient characteristics rather than technical considerations. This allows us to better select patients who may benefit from this regional therapy. The limitations of this study are that it is a single-armed study without a control group, outside the guidelines of a clinical protocol. A randomized controlled trial comparing TACE with particle embolization would be recommended to validate our results. Our goal for this study was to demonstrate evidence of the efficacy of particle embolization. We hope to encourage multiinstitutional studies investigating the impact of technical standards and common procedural endpoints as predictors of outcome. References 1. Tadavarthy SM, Knight L, Ovitt TW, Snyder C, Amplatz K Therapeutic transcatheter arterial embolization. Radiology 1974; 112: Allison DJ, Modlin IM, Jenkins WJ. Treatment of carcinoid liver metastases by hepatic-artery embolisation. Lancet 1977; 2: Wheeler PG, Melia W, Dubbins P, et al. Non-operative arterial embolisation in primary liver tumours. Br Med J 1979; 2: Breedis C, Young G. The blood supply of neoplasms in the liver. Am J Pathol 1954; 30: Misra NC, Jaiswal MS, Singh RV, Das B. Intrahepatic arterial infusion of combination of mitomycin-c and 5-fluorouracil in treatment of primary and metastatic liver carcinoma. Cancer 1977; 39: Patt YZ, Mavligit GM, Chuang VP, et al. Percutaneous hepatic arterial infusion (HAI) of mitomycin C and floxuridine (FUDR): an effective treatment for metastatic colorectal carcinoma in the liver. Cancer 1980; 46: Reed ML, Vaitkevicius VK, Al-Sarraf M, et al. The practicality of chronic hepatic artery infusion therapy of primary and metastatic hepatic malignancies: ten-year results of 124 patients in a prospective protocol. Cancer 1981; 47: Patt YZ, Chuang VP, Wallace S, Benjamin RS, Fuqua R, Mavligit GM. Hepatic arterial chemotherapy and occlusion for palliation of primary hepatocellular and unknown primary neoplasms in the liver. Cancer 1983; 51: Charnsangavej C, Chuang VP, Wallace S, Soo CS, Bowers T. Work in progress: transcatheter management of primary carcinoma of the liver. Radiology 1983; 147: Hirai K, Kawazoe Y, Yamashita K, et al. Arterial chemotherapy and transcatheter arterial embolization therapy for non-resectable hepatocellular carcinoma. Cancer Chemother Pharmacol 1989; 23(suppl):S37 S Lo CM, Ngan H, Tso WK, et al. Randomized controlled trial of transarterial lipiodol chemoembolization for unresectable hepatocellular carcinoma. Hepatology 2002; 35: Llovet JM, Real MI, Montana X, et al. Arterial embolisation or chemoembolisation versus symptomatic treatment in patients with unresectable hepatocellular carcinoma: a randomised controlled trial. Lancet 2002; 359: Brown DB, Geschwind JF, Soulen MC, Millward SF, Sacks D. Society of Interventional Radiology position statement on chemoembolization of hepatic malignancies. J Vasc Interv Radiol 2006; 17: Brown KT, Nevins AB, Getrajdman GI, et al. Particle embolization for hepatocellular carcinoma. J Vasc Interv Radiol 1998; 9: Jaeger HJ, Mehring UM, Castaneda F, et al. Sequential transarterial chemoembolization for unresectable advanced hepatocellular carcinoma. Cardiovasc Intervent Radiol 1996; 19: Pelletier G, Ducreux M, Gay F, et al. Treatment of unresectable hepatocellular carcinoma with lipiodol chemoembolization: a multicenter randomized trial. Groupe CHC. J Hepatol 1998; 29: Charnsangavej C. Chemoembolization of liver tumors. Semin Intervent Radiol 1993; 10:

8 Volume 19 Number 6 Maluccio et al Maluccio M, Covey AM, Gandhi R, et al. Comparison of survival rates after bland arterial embolization and ablation versus surgical resection for treating solitary hepatocellular carcinoma up to 7 cm. J Vasc Interv Radiol 2005; 16: Brown KT. Fatal pulmonary complications after arterial embolization with micro m tris-acryl gelatin microspheres. J Vasc Interv Radiol 2004; 15: Song SY, Chung JW, Han JK, et al. Liver abscess after transcatheter oily chemoembolization for hepatic tumors: incidence, predisposing factors, and clinical outcome. J Vasc Interv Radiol 2001; 12: National Cancer Institute. Common Terminology Criteria for Adverse Events, version 3.0. Bethesda, MD: National Cancer Institute, World Health Organization. WHO handbook for reporting results of cancer treatment. Geneva: WHO, Savastano S, Feltrin GP, Neri D, et al. Palliative treatment of hepatocellular carcinoma with transcatheter arterial embolization. Acta Radiol 1993; 34: Bruix J, Castells A, Montana X, et al. Phase II study of transarterial embolization in European patients with hepatocellular carcinoma: need for controlled trials. Hepatology 1994; 20: Bruix J, Llovet JM, Castells A, et al. Transarterial embolization versus symptomatic treatment in patients with advanced hepatocellular carcinoma: results of a randomized, controlled trial in a single institution. Hepatology 1998; 27: Guo WJ, Li J, Ling WL, et al. Influence of hepatic arterial blockage on blood perfusion and VEGF, MMP-1 expression of implanted liver cancer in rats. World J Gastroenterol 2002; 8: Gupta S, Kobayashi S, Phongkitkarun S, Broemeling LD, Kan Z. Effect of transcatheter hepatic arterial embolization on angiogenesis in an animal model. Invest Radiol 2006; 41: Li X, Feng GS, Zheng CS, Zhuo CK, Liu X. Influence of transarterial chemoembolization on angiogenesis and expression of vascular endothelial growth factor and basic fibroblast growth factor in rat with Walker-256 transplanted hepatoma: an experimental study. World J Gastroenterol 2003; 9: Camma C, Schepis F, Orlando A, et al. Transarterial chemoembolization for unresectable hepatocellular carcinoma: meta-analysis of randomized controlled trials. Radiology 2002, 224: Geschwind JF, Ramsey DE, Choti MA, Thuluvath PJ, Huncharek MS. Chemoembolization of hepatocellular carcinoma: results of a metaanalysis. Am J Clin Oncol 2003; 26: Llovet JM, Bruix J. Systematic review of randomized trials for unresectable hepatocellular carcinoma: chemoembolization improves survival. Hepatology 2003; 37: Kawai S, Tani M, Okamura J, et al. Prospective and randomized clinical trial for the treatment of hepatocellular carcinoma a comparison between L-TAE with Farmorubicin and L-TAE with Adriamycin: preliminary results (second cooperative study). Cooperative Study Group for Liver Cancer Treatment of Japan. Cancer Chemother Pharmacol 1994; 33(suppl):S97 S Kawai S, Tani M, Okamura J, et al. Prospective and randomized trial of lipiodol-transcatheter arterial chemoembolization for treatment of hepatocellular carcinoma: a comparison of epirubicin and doxorubicin (second cooperative study). The Cooperative Study Group for Liver Cancer Treatment of Japan. Semin Oncol 1997; 24(2 suppl 6):S6-38 S Groupe d Etude et de Traitement du Carcinome Hépatocellulaire. A comparison of lipiodol chemoembolization and conservative treatment for unresectable hepatocellular carcinoma. N Engl J Med 1995; 332: Chen MS, Li JQ, Zhang YQ, et al. High-dose iodized oil transcatheter arterial chemoembolization for patients with large hepatocellular carcinoma. World J Gastroenterol 2002; 8: Lee HS, Kim JS, Choi IJ, Chung JW, Park JH, Kim CY. The safety and efficacy of transcatheter arterial chemoembolization in the treatment of patients with hepatocellular carcinoma and main portal vein obstruction: a prospective controlled study. Cancer 1997; 79: Georgiades CS, Hong K, D Angelo M, Geschwind JF. Safety and efficacy of transarterial chemoembolization in patients with unresectable hepatocellular carcinoma and portal vein thrombosis. J Vasc Interv Radiol 2005; 16: Takayasu K, Arii S, Ikai I, et al. Prospective cohort study of transarterial chemoembolization for unresectable hepatocellular carcinoma in 8510 patients. Gastroenterology 2006; 131: