Local Recurrence After Chemoembolization of Hepatocellular Carcinoma: Uptake of Gadoxetic Acid as a New Prognostic Factor

Transcription

1 Gastrointestinal Imaging Original Research Ishimaru et al. Gadoxetic Acid Uptake as Predictor of Hepatocellular Carcinoma Recurrence Gastrointestinal Imaging Original Research Hideki Ishimaru 1,2,3 Kazuaki Nakashima 1 Takayuki Sakugawa 1 Ayami Sakamoto 1 Yohjiro Matsuoka 1 Kazuto Ashizawa 4 Masataka Uetani 3 Ishimaru H, Nakashima K, Sakugawa T, et al. Keywords: gadoxetic acid enhanced MRI, hepatocellular carcinoma, prognostic factor, transcatheter chemoembolization DOI: /AJR Received February 24, 2013; accepted after revision September 2, Department of Radiology, National Hospital Organization Nagasaki Medical Center, Omura, Japan. 2 Clinical Research Center, National Hospital Organization Nagasaki Medical Center, Omura, Japan. 3 Department of Radiology, Nagasaki University Graduate School of Biomedical Sciences, Sakamoto, Nagasaki , Japan. Address correspondence to H. Ishimaru (ishi_maru@yahoo.co.jp). 4 Department of Clinical Oncology and Radiology, Nagasaki University Graduate School of Biomedical Sciences, Nagasaki, Japan. AJR 2014; 202: X/14/ American Roentgen Ray Society Local Recurrence After Chemoembolization of Hepatocellular Carcinoma: Uptake of Gadoxetic Acid as a New Prognostic Factor OBJECTIVE. The purpose of this article is to investigate whether there is a difference in susceptibility to transcatheter arterial chemoembolization between hepatocellular carcinomas (HCCs) showing high uptake and those showing low uptake of gadoxetic acid in the hepatobiliary phase of MRI. MATERIALS AND METHODS. One hundred HCCs that achieved optimal chemoembolization, as assessed by immediate CT in 60 patients, were classified as having high (n = 19) or low (n = 81) uptake of gadoxetic acid on MRI performed before chemoembolization. The local recurrence rates were estimated using the Kaplan-Meier method, and differences between the groups were compared using the log-rank test. The following factors were also correlated with the local recurrence rate using the Cox proportional hazards model for a univariate analysis: high uptake of gadoxetic acid, number of feeding arteries, extrahepatic arterial supply, Child-Pugh class, clinical tumor stage, size, location, and iodized oil accumulation in the noncancerous tissue surrounding the lesion. Parameters that were significant at p < 0.05 were entered into a multivariate model. RESULTS. The 1- and 3-year local recurrence free rates were 95% in high-uptake HCCs and 66% and 54%, respectively, in low-uptake HCCs (log-rank test, p < 0.01). The low uptake of gadoxetic acid was the only significant predictor of early local recurrence (hazard ratio = 9.24; p = 0.03) by multivariate analysis. CONCLUSION. HCCs showing high uptake of gadoxetic acid appear to be susceptible to chemoembolization. G adoxetic acid is a new hepatobiliary-specific contrast medium for MRI, of which approximately 50% is taken up by hepatocytes via the uptake transporter organic anion-transporting polypeptide 8 (OATP8, also referred to as OATP1B3) and then excreted into the bile ducts [1]. The biphasic nature of gadoxetic acid enables us to obtain dynamic contrast-enhanced imaging and hepatobiliary phase imaging, which depicts lesions without hepatobiliary function, including typical hepatocellular carcinomas (HCCs), as hypointense areas relative to the background of the liver parenchyma. OATP8 is also the uptake transporter of gadoxetic acid in human HCC cells, and approximately 5 10% of HCCs take up gadoxetic acid via the overexpression of OATP8 and show unique hyperintensity in the hepatobiliary phase [2]. Furthermore, OATP8 is known to transport certain anticancer drugs, such as antifolate methotrexate [3], paclitaxel, and docetaxel [4]. The down-regulation of OATP8 in typical HCCs is thought to influence the uptake of anticancer drugs and result in chemoresistance [5]. It has been speculated that HCCs showing upregulation of OATP8 may take up these drugs in addition to gadoxetic acid, and hepatobiliary phase imaging may therefore be used to predict the potential efficacy of anticancer drugs transported by OATP8 [6]. However, the correlation between gadoxetic acid uptake and the efficacy of anticancer drugs remains unclear. Surgical resection has been considered a choice for the long-term control of HCC. However, no more than 40% of patients are surgical candidates at the time of diagnosis [7]. Transarterial chemoembolization (TACE) is the most widely used primary treatment for unresectable HCC because of its survival benefit [8], because the arterial obstruction induces ischemic tumor necrosis and prevents washout of injected anticancer drugs from a tumor [9]. The current study 744 AJR:202, April 2014

2 Gadoxetic Acid Uptake as Predictor of Hepatocellular Carcinoma Recurrence aimed to investigate whether there is a difference in susceptibility to TACE between HCCs showing high uptake of gadoxetic acid and HCCs showing low uptake, because the degree of uptake of gadoxetic acid may be useful in the guidance of therapy and predicting prognosis. Materials and Methods Patients Our retrospective study was approved by the institutional review board, and informed consent was waived. In our hospital, patients who were not suitable for curative therapies, such as resection, liver transplantation, or percutaneous intervention, were considered for TACE. Our study group included all patients with HCC who were undergoing treatment for the first time between July 2008 and June 2012 and who had undergone gadoxetic acid enhanced MRI before TACE. The requirements for inclusion in this retrospective study were as follows: no previous treatment for HCC, liver cirrhosis classified as Child- Pugh class A or B, Barcelona Clinic Liver Cancer (BCLC) stage A or B, and no vascular invasion. The exclusion criteria included the lack of gadoxetic acid enhanced MRI before TACE and a nodule with a marginal defect of iodized oil on CT immediately after TACE. The study included 43 men and 17 women aged years old (median age, 68.4 years). Overall, 19 patients had a solitary nodule, and 41 had multiple nodules. Twenty-seven patients (45%) had advanced stage HCC (BCLC stage B). Nineteen patients (32%) had solitary nodules (BCLC stage A); however, two of the 19 patients refused surgery, and the remaining 17 were regarded as unsuitable for surgery because of at least one of the following reasons: thrombocytopenia (platelet count < 100,000 cells/mm 3 ; n = 13), poor hepatic reserve (n = 8), massive ascites (n = 3), or advanced age (> 80 years old; n = 5). In these 19 patients and 14 patients with two or three nodules (BCLC stage A), the nodules were not suitable for radiofrequency ablation for at least one of the following reasons: large tumor size (> 3 cm; n = 9); patient s thrombocytopenia (platelet count < 100,000 cells/mm 3 ; n = 19); proximity to large vessels, gastrointestinal tract, or gallbladder (n = 4); poor visualization on sonography (n = 2); or massive ascites (n = 3). Diagnosis Except for three nodules in three patients with pathologic diagnosis by needle biopsy, the diagnosis of HCC was made on the basis of at least two coincident imaging findings on CT and MRI compatible with HCC in high-risk patients [10, 11]. All nodules were hypervascular tumors with typical imaging findings, including the presence of hyperenhancement on arterial phase CT and MRI; this mass appears hypoenhancing relative to the liver on delayed phase CT and MRI. For lesions with homogeneous high uptake of gadoxetic acid, a hypointense rim on hepatobiliary phase imaging suggestive of a pseudocapsule was required for inclusion in this study to rule out focal nodular hyperplasia like nodules [2]. Gadoxetic Acid Enhanced MRI Gadoxetic acid enhanced MRI was performed before TACE (mean time before TACE, 33.9 days; range, 0 93 days) for the characterization and pretreatment staging of HCC. MRI was performed with a 1.5-T system (Intera Achieva Nova Dual 1.5 T, Philips Healthcare). Contrast-enhanced dynamic studies and hepatobiliary phase imaging were performed using 3D T1-weighted field-echo images with fat suppression (T1-high resolution isotropic volumetric excitation; FOV, 350 mm; TR/TE, 3.6/1.7; flip angle, 10 ; 4-mm thickness with 2-mm overlap; matrix). For the dynamic study, a dose of 0.1 ml/kg of body weight (0.25 mmol/ml) of gadoxetic acid (Primovist, Bayer Schering Pharma) was IV injected at a flow rate of 1.0 ml/s, followed by a 20-mL saline flush. We used the test injection method (0.5 ml of gadoxetic acid plus a 10-mL saline flush) to determine the optimal arterial dominant phase, which was determined as the peak time of enhancement in the abdominal aorta plus an additional 8 seconds. After imaging in the arterial phase, the portal phase and equilibrium phase images were obtained 20 and 60 seconds after completion of the previous imaging phase, respectively. The hepatobiliary phase imaging was obtained 20 minutes after injection in all patients. Image Analyses The imaging analysis was performed on the basis of the consensus obtained by two experienced radiologists who were blinded to the outcome of TACE. The mean signal intensity (SI) of the tumor and surrounding background liver was measured by placing regions of interest on hepatobiliary phase images. The region of interest of the tumor was placed at the section level of the largest tumor diameter devoid of necrosis. The relative enhancement ratio was calculated on unenhanced and contrast-enhanced MRI using the following formula: (SI nodule / SI liver ) hepatobiliary phase / (SI nodule / SI liver ) unenhanced, where SI nodule represents the SI of the nodule, and SI liver represents the SI of the liver parenchyma [12]. According to previous reports [12, 13], the HCC nodules were retrospectively classified into two types according to relative enhancement ratio: the tumor with a relative enhancement ratio of more than 1 was defined as high-uptake HCC, and the other tumors (relative enhancement ratio < 1) were defined as lowuptake HCCs. TACE Every TACE procedure was performed by one of two board-certified interventional radiologists and one of four interventional radiologists in training. Before chemoembolization, rotational digital subtraction angiography of the common, proper, or right or left hepatic artery was performed using a 4.0-French catheter introduced via the femoral artery for evaluation of the feeding arteries. In cases of subphrenic HCC, located in segments VII and VIII [14], selective arteriography of the inferior phrenic artery was performed first to investigate the extrahepatic blood supply to the tumor (Fig. 1). Then, a 1.8-French tip (Carnelian Pixie, Tokai Medical Products) microcatheter was inserted as close to the tumor as possible. The anticancer-in-oil emulsion was prepared using a pumping method, whereby 1 10 ml of iodized oil (Lipiodol, Terumo), 5 10 ml of iopamidol 300 (Oiparomin, Konica-Minolta), mg of epirubicin (Epirubicin, Nippon-Kayaku), and 8 10 mg of mitomycin-c (Mitomycin, Kyowa Hakko) were mixed. The amount of the anticancer-in-oil emulsion was determined on the basis of the number and the diameter of the lesions, as well as the degree of tumor vascularity. These emulsions were mixed by repeated pumping through a three-way stopcock. TACE was performed by injecting the anticancer-in-oil emulsion followed by gelatin sponge particles measuring 1 mm in diameter (Gelpart, Nippon-Kayaku). When the tumor was supplied by two or more different feeding arteries, each feeding artery was embolized separately (Figs. 1 and 2). In cases of multiple lesions, we selected lesions to be treated in order of tumor size, starting with the largest. Then we tried to treat as much lesion as possible for as long a time as the patient would accept. The maximum number of treated nodules in one session was seven. TACE is routinely performed using a combined angiography and CT system (Multistar Plus/TOP, Siemens Healthcare), which makes it possible to evaluate lipiodol accumulation in HCC during angiographic procedures. Iodized oil accumulation was evaluated on CT immediately after the procedure (Figs. 1 and 2); whenever the CT showed a nontherapeutic area in any part of the tumor, we subsequently attempted additional TACE via other feeding arteries during the same angiographic session. The nodules showing a marginal defect indicating residual tumor on CT at the end of the procedure despite such efforts were excluded from AJR:202, April

3 Ishimaru et al. this study. The nodules that showed intratumoral heterogeneous deposit of iodized oil indicating tumor necrosis or intratumoral fat component were included in this study. For the analyses, the total number of injection points was counted as the number of feeding arteries, and the presence of an extrahepatic tumor supply was evaluated. Iodized oil accumulation in the noncancerous tissue completely surrounding the lesion was also assessed by the immediate CT and used in the analyses. Definition of Local Recurrence The patients were followed up by either dynamic CT or dynamic MRI examinations every 3 6 months until recurrent HCC was confirmed A D G (Figs. 1 and 2). Local recurrence was determined to be present when a portion showing early enhancement on arterial phase and washout on delayed phase was observed within or at the margins of the original mass (mean follow-up period after TACE, 16.4 months; range, 3 54 months). Statistical Analyses The comparisons between groups were conducted using the chi-square test for categoric variables. Patient age, lesion size, and the number of feeding arteries were statistically evaluated using the Student t test. The primary outcome was the local recurrence free interval (calculated in months). The B E local recurrence rate of each group was estimated using the Kaplan-Meier method, and differences between the groups were compared using the logrank test. In addition to intratumoral high uptake of gadoxetic acid, the other potential prognostic variables, such as number of feeding arteries and the presence of extrahepatic arterial supply, were analyzed in a univariate analysis using the Cox proportional hazards model for the prediction of local recurrence. Other traditional prognostic factors were also examined, including Child-Pugh class, BCLC tumor staging, tumor size, central Fig year-old man with right lobe hepatocellular carcinoma showing low uptake of gadoxetic acid. A, Arterial phase of contrast-enhanced dynamic MRI shows 3.5-cm-diameter hypervascular tumor in right hepatic lobe. B, Axial T1-weighted image obtained during hepatobiliary phase shows intratumoral low uptake of gadoxetic acid. C, Right inferior phrenic angiography shows hypervascular tumor and feeding artery (arrow and inset). D, Right hepatic angiography shows hypervascular tumor and three tumor-feeding arteries arising from right posterior and right anterior hepatic artery (arrows and insets). E, Unenhanced axial CT scan obtained immediately after transarterial chemoembolization shows tumor with good accumulation of iodized oil. There was iodized oil accumulation in medial noncancerous tissue but poor accumulation in anterior noncancerous tissue. F and G, Arterial (F) and equilibrium (G) phases of dynamic CT scan obtained at 21-month follow-up shows local recurrence (arrows) adjacent to accumulated iodized oil. C F 746 AJR:202, April 2014

4 Gadoxetic Acid Uptake as Predictor of Hepatocellular Carcinoma Recurrence lesions located within 0.5 cm of the first or second branches of the portal vein [15], and iodized oil accumulation in the noncancerous tissue completely surrounding the lesion [16]. All variables with p < 0.05 in the univariate analyses were entered in a multivariate analysis using forced-entry model. The statistical analyses were performed using commercially available software (StatMate IV, ATMS). A p < 0.05 was considered to indicate statistical significance. Results Overall, 19 nodules were classified as high-uptake HCCs, and 81 nodules were classified as low-uptake HCCs. Table 1 shows the tumor characteristics. The mean (± SD) relative enhancement ratios for highuptake and low-uptake HCCs were 1.12 ± 0.10 and 0.79 ± 0.08, respectively. Otherwise, tumor characteristics were similar between the two groups. High-uptake HCC nodules occurred in 12 patients: one patient had single high-uptake HCC, one had two high-uptake HCCs, and 10 patients had both high- and low-uptake HCCs. The remaining 48 patients had single or multiple low-uptake HCCs. Table 2 shows the profiles of these 60 patients. The percentage of patients positive for prothrombin induced by vitamin K absence or agonist II (PIVKA II) and the percentage of patients with BCLC stage B HCC were higher for patients with high-uptake HCCs. Otherwise, the patient characteristics were similar between the two groups. Cumulative overall local recurrence free rates in all nodules were 70% at 1 year and 63% at 3 years. Kaplan-Meier curves were generated to show the relationship between local recurrence free rates and intratumoral high uptake of gadoxetic acid (Fig. 3). In lowuptake HCCs, the 1- and 3-year local recurrence free rates were 66% and 54%, respectively. In contrast, the 1- and 3-year local recurrence free rates in high-uptake HCCs were 95%, and there was a statistically significant difference between the rates for each type of HCC according to the log-rank test (p < 0.01). The univariate analysis of the parameters in the Cox model showed that tumor size (hazard ratio [HR] = 1.31; p < 0.01) and the number of feeding arteries (HR = 1.55; p < 0.01) were continuous variables that served as significant predictors of early local recurrence, in addition to intratumoral high uptake of gadoxetic acid (HR = 10.09; p = 0.02) and iodized oil accumulation in the noncancerous tissue completely surrounding the lesion (HR = 2.44; p = 0.04) (Table 3). TABLE 1: Profiles of the 100 Enrolled Hepatocellular Carcinoma (HCC) Nodules Characteristic High-Uptake HCCs (n = 19) Low-Uptake HCCs (n = 81) p Relative enhancement ratio, mean ± SD 1.12 ± ± 0.08 Size (cm), mean ± SD (range) 2.5 ± 1.3 ( ) 2.9 ± 1.8 ( ) 0.47 Solitary lesions in the patients One of multiple lesions Central location No. of feeding arteries, mean ± SD 1.57 ± ± Extrahepatic arterial supply Iodized oil accumulation in the noncancerous tissue completely surrounding the lesion Note Unless otherwise noted, data are numbers of nodules. TABLE 2: Profiles of 60 Patients With Hepatocellular Carcinoma (HCC) Characteristic Patients With High- Uptake HCCs (n = 12) a Patients With Low- Uptake HCCs (n = 48) p Age (y), mean ± SD 66.0 ± ± Sex Male Female 3 14 Child-Pugh cirrhosis class A B 4 24 Cause of cirrhosis Hepatitis B virus Hepatitis C virus Alcoholic Other Positive for α-fetoprotein Positive for prothrombin induced by vitamin K absence or agonist II Barcelona Clinic Liver Cancer stage 0.04 A 3 30 B 9 18 Note Unless otherwise noted, data are number of patients. a Nine patients who had both high- and low-uptake HCCs were included. The four variables with a p value less than 0.05 in the univariate analysis were entered into a multivariate analysis, which found that the absence of intratumoral high uptake of gadoxetic acid (HR = 9.24; p = 0.03) was significantly associated with local recurrence (Table 4). Discussion HCC shows frequent multicentricity even at the time of the first diagnosis, as well as frequent recurrence after surgery or local ablation, especially in hepatitis C virus related liver cirrhosis [17]. In Japan, superselective TACE is the most commonly performed AJR:202, April

5 Ishimaru et al. TABLE 3: Predicting Local Recurrence With a Univariate Cox Proportional Hazard Analysis Variable Hazard Ratio (95% CI) p Child-Pugh Class A No 1.77 ( ) 0.11 Barcelona Clinic Liver Cancer stage A No 1.23 ( ) 0.58 Tumor size (cm) 1.31 ( ) < 0.01 Intratumoral high uptake of gadoxetic acid No ( ) 0.02 Central location 1.74 ( ) 0.18 No No. of feeding arteries 1.55 ( ) < 0.01 Extrahepatic arterial supply 1.73 ( ) 0.37 No Iodized oil accumulation in the noncancerous tissue completely surrounding the lesion No 2.44 ( ) 0.04 TABLE 4: Predicting Local Recurrence With a Multivariate Cox Proportional Hazard Analysis Variable Hazard Ratio (95% CI) p Tumor size (cm) 1.20 ( ) 0.11 Intratumoral high uptake of gadoxetic acid No 9.24 ( ) 0.03 No. of feeding arteries 1.27 ( ) 0.21 Iodized oil accumulation in the noncancerous tissue completely surrounding the lesion No 1.66 ( ) 0.27 treatment method for HCC, because this technique has an improved survival benefit [8], a less-invasive nature, and increased effectiveness [17]. The reported complete response ratios for hypervascular HCCs that are less than 5 cm are approximately 30 60% after superselective TACE with iodized oil [18 20], and the local control rates of our series (63% at 3 years) were within the range of those of recently published studies [18 20]. Furthermore, the tumor size [19], degree of iodized oil accumulation [16, 19, 21], and location [15, 21] are well-known factors that affect the local control rate. The results of the current study show that intratumoral high uptake of gadoxetic acid may serve as a novel prognostic factor for good local control after TACE. One possible explanation for the difference in the local control rate between groups in this study may be a difference in chemosensitivity to epirubicin or mitomycin C. Before the arrival of sorafenib, both singleagent and multidrug chemotherapy regimens showed a lack of efficacy in most cases [22, 23] and therefore have not been a promising option for the treatment of this primary cancer of the liver. For most chemotherapeutic drugs, the exact mechanisms of cellular uptake are poorly understood [24]. However, some anticancer drugs, such as methotrexate [3] and paclitaxel [4], have been shown to be a substrate for OATP8, and some authors have speculated that HCCs with upregulation of OATP8 may take up these drugs in addition to gadoxetic acid [6]. Therefore, preoperative gadoxetic acid enhanced MRI may be used to predict the potential efficacy of anticancer drugs transported by OATP8 [6]. Recently, Iwakiri et al. [25] showed that epirubicin was taken up, at least in part, by OATPs in primary cultures of rat hepatocytes. Because OATPs in the rat liver, including Oatp1a1 (Oatp1), Oatp1a4 (Oatp2), and Oatp1b2 (Oatp4), function similarly to OATP2 and OATP8 in the human liver [26], we surmise that overexpression of OATP8 in HCCs may increase cellular uptake of epirubicin and the chemosensitivity of HCCs with high uptake of gadoxetic acid. However, further studies are necessary to investigate whether epirubicin is a substrate for OATP8 in human HCCs. When reviewing the literature, we did not find any reports examining the uptake transporter of mitomycin C. A more probable explanation for the difference in the local control rate between groups in this study is that there may have been a difference in tumor biology between groups. A recent study by Kitao et al. [27] found that hyperintense HCCs on hepatobiliary phase, which are nearly equal to our high-uptake HCCs, showed a significantly lower recurrence rate (including all local recurrence and intrahepatic and extrahepatic metastasis) after surgical resection than did hypointense HCCs, which are nearly equal to our low-uptake HCCs. They also showed on surgical specimens that hypointense HCCs had a significantly higher rate of portal vein invasion than did hyperintense HCCs. Moreover, hypointense HCCs included poorly differentiated HCCs (42/158 [27%]). Other previous pathologic analyses [6, 28] also report- 748 AJR:202, April 2014

6 Gadoxetic Acid Uptake as Predictor of Hepatocellular Carcinoma Recurrence Percentage A D G B E Fig year-old man with right lobe hepatocellular carcinoma (HCC) showing high uptake of gadoxetic acid. A and B, Arterial (A) and equilibrium (B) phases of contrast-enhanced dynamic MRI show large hypervascular tumor in right hepatic lobe. C, Axial T1-weighted image obtained during hepatobiliary phase shows intratumoral high uptake of gadoxetic acid. D, Right hepatic angiography shows hypervascular tumor and three tumor-feeding arteries arising from right posterior and right anterior hepatic artery (arrows and insets). E, Unenhanced axial CT scan obtained immediately after transarterial chemoembolization shows tumor with slightly heterogeneous accumulation of iodized oil. There was iodized oil accumulation in anterior and medial noncancerous tissue but poor accumulation in dorsal noncancerous tissue. F and G, Arterial (F) and equilibrium (G) phases of dynamic MRI scan obtained at 33-month follow-up show no local recurrence but newly developed HCC in superior posterior segment. Follow-Up Period (mo) High-uptake HCCs Low-uptake HCCs Fig. 3 Kaplan-Meier curves show local recurrence free rates for low-uptake and high-uptake hepatocellular carcinomas (HCCs). ed that some HCCs appearing hypointense to the surrounding liver parenchyma were classified as poorly differentiated HCCs and that, in contrast, all HCCs exhibiting uptake of gadoxetic acid were classified as well or moderately differentiated HCCs. To the best of our knowledge, there was only one case of poorly differentiated HCC showing uptake of gadoxetic acid on hepatobiliary phase [29]. Intrahepatic metastases can occur through the portal vein as an efferent tumor vessel, and microscopic daughter nodules are frequently found in the same portal branch area or an adjacent area [30]. Furthermore, aggressive and poorly differentiated HCCs more frequently tend to develop intrahepatic metastases via the portal vein [31]. In our chemoembolization method, C F AJR:202, April

7 Ishimaru et al. which preserves nontumorous liver tissue as much as possible, the therapeutic effect may have been limited and, therefore, was insufficient to control microscopic daughter nodules adjacent to the main tumor. When untreated microscopic daughter nodules grow in size and are adjacent to the areas of iodized oil retention, these lesions would likely be counted as evidence of local recurrence after TACE. Thus, the low-uptake HCCs in our study may have included some poorly differentiated HCCs that were resistant to TACE, and the differences in the local control rates between low- and high-uptake HCCs may have been affected by the increased tendency of low-uptake HCCs to metastasize to adjacent areas. In this series, the PIVKA II positive rate, which correlates with the frequency of intrahepatic metastasis [32], was higher in patients with high-uptake HCCs than in those without high-uptake HCCs. In this study population, the percentage of BCLC stage B tumors was unexpectedly higher in patients with high-uptake HCCs than in those without high-uptake HCCs. This patient selection bias may explain the higher positive rate for PIVKA II in patients with high-uptake HCCs. However, the bias would not favor the local control rate of high-uptake HCCs. The current study has some limitations. First, because needle biopsy was performed for only three of the 100 nodules (3%), the rates of poorly differentiated HCCs, which are resistant to locoregional treatments, such as radiofrequency ablation, TACE, and combined radiofrequency ablation and TACE [33], were unknown in both groups. However, according to the guidelines on the management of HCC by the American Association for the Study of Liver Diseases [11], a lesion more than 1 cm in diameter with typical imaging appearances (i.e., hypervascular in the arterial phase with washout in the portal venous or delayed phase) should be treated as HCC without biopsy. Moreover, the incidence of needle tract tumor seeding after biopsy of a HCC is reported to be 2.7% [34]. Biopsy for all nodules before TACE would be intolerable in patients with advanced stage HCCs. Therefore, we think that our data are important to clarify that we can predict good local control after TACE by intratumoral high uptake of gadoxetic acid without biopsy. Second, we did not analyze the difference in survival between two groups because the number of patients was small (only 60 patients), and the follow-up period averaged 16.4 months. In conclusion, TACE can provide good local control within HCCs showing high uptake of gadoxetic acid, although the exact mechanism for this response to chemoembolization is not fully understood. Acknowledgments We acknowledge the essential contributions of Hiroshi Yatsuhashi and Shinya Nagaoka of Clinical Research Center and every member of Department of Hepatology at Nagasaki Medical Center. References 1. Seale MK, Catalano OA, Saini S, Hahn PF, Sahani DV. Hepatobiliary-specific MR contrast agents: role in imaging the liver and biliary tree. RadioGraphics 2009; 29: Lee JM, Zech CJ, Bolondi L, et al. Consensus report of the 4th International Forum for Gadolinium-Ethoxybenzyl-Diethylenetriamine Pentaacetic Acid Magnetic Resonance Imaging. Korean J Radiol 2011; 12: Abe T, Unno M, Onogawa T, et al. LST-2, a human liver-specific organic anion transporter, determines methotrexate sensitivity in gastrointestinal cancers. Gastroenterology 2001; 120: Smith NF, Acharya MR, Desai N, Figg WD, Sparreboom A. Identification of OATP1B3 as a highaffinity hepatocellular transporter of paclitaxel. Cancer Biol Ther 2005; 4: Takano M, Otani Y, Tanda M, Kawami M, Nagai J, Yumoto R. Paclitaxel-resistance conferred by altered expression of efflux and influx transporters for paclitaxel in the human hepatoma cell line, HepG2. Drug Metab Pharmacokinet 2009; 24: Narita M, Hatano E, Arizono S, et al. Expression of OATP1B3 determines uptake of Gd-EOB-DT- PA in hepatocellular carcinoma. J Gastroenterol 2009; 44: Poon RT, Fan ST, Tsang FH, Wong J. Locoregional therapies for hepatocellular carcinoma: a critical review from the surgeon s perspective. Ann Surg 2002; 235: 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: Shin SW. The current practice of transarterial chemoembolization for the treatment of hepatocellular carcinoma. Korean J Radiol 2009; 10: Bruix J, Sherman M, Llovet JM, et al. Clinical management of hepatocellular carcinoma: conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001; 35: Bruix J, Sherman M; American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53: Tsuboyama T, Onishi H, Kim T, et al. Hepatocellular carcinoma: hepatocyte-selective enhancement at gadoxetic acid-enhanced MR imaging correlation with expression of sinusoidal and canalicular transporters and bile accumulation. Radiology 2010; 255: Kim JY, Kim MJ, Kim KA, Jeong HT, Park YN. Hyperintense HCC on hepatobiliary phase images of gadoxetic acid-enhanced MRI: correlation with clinical and pathological features. Eur J Radiol 2012; 81: Basile A, Tsetis D, Montineri A, et al. MDCT anatomic assessment of right inferior phrenic artery origin related to potential supply to hepatocellular carcinoma and its embolization. Cardiovasc Intervent Radiol 2008; 31: Murakami T, Ishimaru H, Sakamoto I, et al. Percutaneous radiofrequency ablation and transcatheter arterial chemoembolization for hypervascular hepatocellular carcinoma: rate and risk factors for local recurrence. Cardiovasc Intervent Radiol 2007; 30: Nishimine K, Uchida H, Matsuo N, et al. Segmental transarterial chemoembolization with Lipiodol mixed with anticancer drugs for nonresectable hepatocellular carcinoma: follow-up CT and therapeutic results. Cancer Chemother Pharmacol 1994; 33(suppl):S60 S Matsui O, Miyayama S, Sanada J, et al. Interventional oncology: new options for interstitial treatments and intravascular approaches: superselective TACE using iodized oil for HCC rationale, technique and outcome. J Hepatobiliary Pancreat Sci 2010; 17: Matsui O, Kadoya M, Yoshikawa J, et al. Small hepatocellular carcinoma: treatment with subsegmental transcatheter arterial embolization. Radiology 1993; 188: Takayasu K, Muramatsu Y, Maeda T, et al. Targeted transarterial oily chemoembolization for small foci of hepatocellular carcinoma using a unified helical CT and angiography system: analysis of factors affecting local recurrence and survival rates. AJR 2001; 176: Miyayama S, Matsui O, Yamashiro M, et al. Ultraselective transcatheter arterial chemoembolization with a 2-f tip microcatheter for small hepatocellular carcinomas: relationship between local tumor recurrence and visualization of the portal vein with iodized oil. J Vasc Interv Radiol 2007; 18: Park SH, Cho YK, Ahn YS, Park YO, Kim JK, 750 AJR:202, April 2014

8 Gadoxetic Acid Uptake as Predictor of Hepatocellular Carcinoma Recurrence Chung JW. Local recurrence of hepatocellular carcinoma after segmental transarterial chemoembolization: risk estimates based on multiple prognostic factors. Korean J Radiol 2007; 8: Schafer DF, Sorrell MF. Hepatocellular carcinoma. Lancet 1999; 353: Cao H, Phan H, Yang LX. Improved chemotherapy for hepatocellular carcinoma. Anticancer Res 2012; 32: Gillet JP, Gottesman MM. Mechanisms of multidrug resistance in cancer. Methods Mol Biol 2010; 596: Iwakiri T, Okumura M, Hidaka M, et al. Inhibition of carrier-mediated uptake of epirubicin reduces cytotoxicity in primary culture of rat hepatocytes. J Appl Toxicol 2008; 28: Chandra P, Brouwer KL. The complexities of hepatic drug transport: current knowledge and emerging concepts. Pharm Res 2004; 21: Kitao A, Matsui O, Yoneda N, et al. Hypervascular hepatocellular carcinoma: correlation between biologic features and signal intensity on gadoxetic acid-enhanced MR images. Radiology 2012; 265: Motosugi U, Ichikawa T, Sou H, et al. Distinguishing hypervascular pseudolesions of the liver from hypervascular hepatocellular carcinomas with gadoxetic acid-enhanced MR imaging. Radiology 2010; 256: Kim SH, Kim SH, Lee J, et al. Gadoxetic acidenhanced MRI versus triple-phase MDCT for the preoperative detection of hepatocellular carcinoma. AJR 2009; 192: Toyosaka A, Okamoto E, Mitsunobu M, Oriyama T, Nakao N, Miura K. Pathologic and radiographic studies of intrahepatic metastasis in hepatocellular carcinoma: the role of efferent vessels. HPB Surg 1996; 10:97 103; discussion, Ng IO, Guan XY, Poon RT, Fan ST, Lee JM. Determination of the molecular relationship between multiple tumour nodules in hepatocellular carcinoma differentiates multicentric origin from intrahepatic metastasis. J Pathol 2003; 199: Miyaaki H, Nakashima O, Kurogi M, Eguchi K, Kojiro M. Lens culinaris agglutinin-reactive alpha-fetoprotein and protein induced by vitamin K absence II are potential indicators of a poor prognosis: a histopathological study of surgically resected hepatocellular carcinoma. J Gastroenterol 2007; 42: Livraghi T, Goldberg SN, Lazzaroni S, et al. Hepatocellular carcinoma: radio-frequency ablation of medium and large lesions. Radiology 2000; 214: Silva MA, Hegab B, Hyde C, Guo B, Buckels JA, Mirza DF. Needle track seeding following biopsy of liver lesions in the diagnosis of hepatocellular cancer: a systematic review and meta-analysis. Gut 2008; 57: FOR YOUR INFORMATION The comprehensive book based on the ARRS 2013 annual meeting categorical course on Body MRI is now available! For more information or to purchase a copy, see AJR:202, April