- Email: [email protected]
Correlation between Tumor Perfusion and Lipiodol Deposition in Hepatocellular Carcinoma after Transarterial Chemoembolization
Correlation between Tumor Perfusion and Lipiodol Deposition in Hepatocellular Carcinoma after Transarterial Chemoembolization Lin Yang, MD, Xiao Ming Zhang, MD, Xiang Ping Zhou, MD, Wei Tang, MD, Yong Song Guan, MD, Zhao Hua Zhai, MD, and Guo Li Dong, MD
PURPOSE: To study the correlation of tumor perfusion with lipiodol deposition in hepatocellular carcinoma (HCC) after transarterial chemoembolization with multidetector computed tomography (MDCT) perfusion imaging. MATERIALS AND METHODS: MDCT perfusion imaging was performed in 24 patients with HCC 1 to 7 days before chemoembolization. The computed tomography (CT) perfusion parameters, such as hepatic arterial perfusion (HAP), hepatic portal perfusion (HPP), total liver perfusion (TLP), and hepatic arterial perfusion index (HAPI), were calculated with the slope method. The follow-up CT scans (noncontrast) were performed 4 weeks after chemoembolization to analyze lipiodol deposition. The lipiodol deposition in the tumor was classified into three grades and compared with CT perfusion parameters before chemoembolization. RESULTS: The HAP and TLP of tumors before chemoembolization were correlated with the grades of lipiodol deposition in tumors after chemoembolization (r ⴝ 0.768, P < .0001 and r ⴝ 0.616, P ⴝ .001, respectively). However, the HPP and HAPI of the tumors were not related to the grades of iodized oil deposition (r ⴝ 0.227, P ⴝ .286 and r ⴝ 0.111, P ⴝ .607, respectively). Higher HAP was correlated with better lipiodol deposition, and lower HAP was correlated with poorer lipiodol deposition. CONCLUSIONS: MDCT perfusion imaging has the potential to help select more appropriate patients with HCC for chemoembolization. J Vasc Interv Radiol 2010; 21:1841–1846 Abbreviations: DSA ⫽ digital subtraction angiography, FOV ⫽ field of view, HAP ⫽ hepatic arterial perfusion, HAPI ⫽ hepatic arterial perfusion index, HCC ⫽ hepatocellular carcinoma, HPP ⫽ hepatic portal perfusion, MDCT ⫽ multidetector computed tomography, ROI ⫽ region of interest, SD ⫽ standard deviation, TLP ⫽ total liver perfusion
HEPATOCELLULAR carcinoma (HCC) is a common malignant tumor with a
From Sichuan Key Laboratory of Medical Imaging (L.Y., X.M.Z., W.T., Z.H.Z., G.L.D), Department of Radiology, Affiliated Hospital of North Sichuan Medical College, Nanchong, Sichuan 637000, P. R. China and Department of Radiology (X.P.Z., Y.S.G.), West China Hospital, Sichuan University, Chengdu, Sichuan P. R. China. Received August 21, 2009; final revision received June 30, 2010; accepted August 2, 2010. Address correspondence to L.Y.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2010 DOI: 10.1016/j.jvir.2010.08.015
very high mortality rate, and it has become a major health problem worldwide (1). The majority of patients with HCC are unable to receive surgical resection or transplantation because of their poor hepatic function and the typically advanced nature of the disease at presentation. Transarterial chemoembolization has proven to be effective for unresectable HCC (2– 4). Previous findings indicate that the grades of iodized oil deposition in the liver tumor correlated well with the antitumor effect, and the tumor hemodynamics correlated with lipiodol deposition in tumor tissue (5,6). Therefore, the evaluation of liver tu-
mor hemodynamics will help to select more appropriate patients with liver tumors to undergo chemoembolization. Computed tomography (CT) perfusion imaging, which has emerged in recent years, has the ability to provide quantitative information about both arterial and portal perfusion of the liver tumor (7,8). Multidetector computed tomography (MDCT) scanners support dynamic scanning with a larger range and are therefore thought to have the ability to provide a larger amount of perfusion-related data than conventional single-slice CT (9,10). We hypothesized that CT perfusion
1841
1842
•
Correlation between Tumor Perfusion and Lipiodol Deposition
December 2010
JVIR
Table 1 Baseline Patient Demographics and Tumor Characteristics
No.
Age Gender (yr)
Tumor Number
Chemoembolization Basis
Hepatic Arterial Perfusion (mL/ min/mLl)
Hepatic Portal Perfusion (mL/ min/mL)
Total Liver Perfusion (mL/ min/mL)
Hepatic Arterial Perfusion Index (%)
Lipiodol Deposition Grade
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Male Male Male Male Male Male Male Male Male Female Male Male Male Female Male Female Male Male Male Male Male Male Female Male
Multiple Multiple Solitary Solitary Multiple Solitary Multiple Solitary Solitary Solitary Solitary Multiple Multiple Solitary Solitary Multiple Multiple Multiple Solitary Multiple Multiple Solitary Multiple Solitary
Lobar Lobar Lobar Lobar Whole liver Lobar Whole liver Lobar Lobar Lobar Lobar Whole liver Whole liver Whole liver Lobar Lobar Whole liver Whole liver Lobar lobar Whole liver Lobar Whole liver Lobar
0.364 0.254 0.273 0.558 0.400 1.109 0.345 0.565 0.195 0.639 0.180 0.128 0.177 0.238 0.624 0.348 0.430 0.436 0.513 0.669 0.254 0.600 0.512 0.158
0.391 0.093 0.360 1.220 0.156 0.305 0.144 0.947 0.119 0.265 0.274 0.051 0.316 0.080 0.000 0.557 0.095 0.188 0.458 0.393 0.715 0.294 0.226 0.573
0.755 0.347 0.63.3 1.77.8 0.556 1.413 0.489 1.512 0.314 0.900 0.454 0.179 0.493 0.318 0.624 0.905 0.525 0.624 0.971 1.062 0.969 0.894 0.738 0.731
48.2 73.2 43.1 31.4 71.9 78.5 70.6 37.4 62.1 71.0 39.6 71.5 35.9 74.8 100.0 38.5 81.9 69.9 52.8 63.0 26.2 67.1 69.4 21.6
I III II I I I II I III I III III II II I III II I II I I I I III
28 43 50 63 45 56 51 51 50 43 70 58 42 53 51 40 40 51 70 48 56 55 70 55
parameters would correlate with lipiodol deposition in the liver tumor after chemoembolization and that this may help in selecting more appropriate patients with HCC to undergo the chemoembolization procedure. We therefore conducted this study to investigate the correlation between tumor perfusion and lipiodol deposition in HCC after chemoembolization.
Pugh A and B. In this group, 12 patients had a solitary tumor, and 12 patients had more than one tumor (multiple tumors) (Table 1). The diameter of the tumors was 9.1 ⫾ 5.0 cm on CT images. All of the 24 patients were scheduled to undergo chemoembolization therapy and were recruited in this study.
MATERIALS AND METHODS
MDCT Imaging
Patients
MDCT was performed in all patients 1 to 7 days (average, 3.1 ⫾ 1.3 days) before chemoembolization with 16-slice helical CT (Philips Brilliance 16; Philips Medical Systems, Best, The Netherlands). The CT perfusion imaging before chemoembolization was performed at an appropriate level through the tumor based on unenhanced images. In patients with multiple tumors, only one tumor near the portal vein trunk was studied. Each subject was injected intravenously with 50 mL contrast material (Ultravist 300; Schering, Berlin, Germany) with a power injector with a flow rate of 6 mL/sec. Acquisition started 6 seconds after the contrast agent injection
This prospective study was approved by our Institutional Review Board, and patient informed consent was waived. From June 2006 to May 2008, 24 patients with unresectable HCC proven by biopsy were admitted to our hospital. Of the 24 patients, 20 were men and 4 were women, with an average age of 51.6 ⫾ 10.2 years (range, 28 –70 years). All of the 24 patients had chronic viral hepatitis B infection and liver cirrhosis. The mean value of the Child-Pugh score of the 24 patients was 6.2 ⫾ 1.6 points. Twentyone (88%) and three (12%) patients were, respectively, classified as Child-
and lasted for 90 seconds. Patients were asked to hold their breath as long as possible during the CT perfusion imaging. The CT perfusion images were then obtained at the selected slice (eight sections with 3-mm section width) of the tumor. This sequence provided a total of 480 axial images for each patient. The scan parameters for this perfusion sequence were as follows: field of view (FOV) of 320 mm, matrix size of 1024 ⫻ 1024, scan speed of 1.0 sec/rot, tube voltage of 120 kV, and tube current of 150 mA. Finally, dynamic-enhanced scanning (120 kV, 300 mAs) was performed as in our normal workup routine at 25 seconds and 60 seconds after intravenous injection of a supplemental dose of 80 mL of the same contrast media at a rate of 3.0 mL/sec. Chemoembolization Procedure Two radiologists with more than 20 and 6 years of experience in interventional radiology, respectively, performed the chemoembolization procedure. Diagnostic digital subtraction angiography (DSA) was performed in
Volume 21
Number 12
Yang et al
Figure 1. The ROI was outlined on the pretreatment CT image. The ROI of the aorta (A1), portal vein (V1), and spleen (B1) were placed in the center (a). The ROI of the tumor (T1) was placed at least 1 cm away from the tumor border (b).
a DSA unit (LCV plus; GE Medical Systems, Milwaukee, Wisconsin) with the use of a 5-F RH angiographic catheter (Terumo, Fujinomiya, Japan) to assess the location and number of vessels feeding the tumors and to determine the optimal catheter position for chemoembolization. Sometimes, a 3-F microcatheter was also used to perform the catheterization. In the 24 patients, DSA was followed by the chemoembolization procedure. After catheterization, chemoembolization was performed by the administration of 5-fluorouracil (1000 mg–1500 mg) and hydroxycamptothecine (30 mg– 40 mg), followed by lipiodol (3–20 mL) (Lipiodol UltraFluid, Laboratoire Guerbet, Aulnay-Sous-Bois, France) with an Adriamycin (40 mg–50 mg) emulsion and gelfoam particles (with a diameter of 1 mm, 20 – 60 particles) (11). The volume of the embolus was determined by taking into account the hepatic serum functional indices (serum albumin level, serum bilirubin level, and prothrombin time ratio) and the diameter of the lesion. The chemoembolization procedure was stopped when the tumor stain disappeared or decreased apparently or when the patient could no longer tolerate the procedure. The degree of stasis was obtained at the physician’s discretion. CT Image Analysis The follow-up CT scans (noncontrast) were performed 4 weeks after chemoembolization to analyze lipiodol deposition. The raw data of CT perfusion imaging and follow-up CT
scans were transferred to a Philips Brilliance workstation (version 2.0.11, Philips Medical Systems, Best, The Netherlands). One radiologist with 8 years of experience in abdominal CT and 5 years of experience in perfusion CT reviewed the perfusion CT images and measured the CT perfusion parameters. The region-of-interest (ROI) was drawn manually in the center of the aorta, portal vein, and spleen. The following rules were applied when demarcating the ROI: (a) the ROI was placed at least 1 cm away from the tumor border, (b) the ROI did not include the obviously necrotic tissue defined on conventional CT images, and (c) ROI did not include vessels (Fig 1). CT perfusion values of hepatic arterial perfusion (HAP), hepatic portal perfusion (HPP), total liver perfusion (TLP) and hepatic arterial perfusion index (HAPI) were calculated with the slope method. In calculating perfusion values, the time of maximum enhancement within a splenic ROI was used to separate the arterial and portal phases of liver enhancement as previously described by Miles et al (7). In short, HAP (mL/min/mL) was determined by dividing the peak gradient of the liver time-attenuation curve before the peak splenic attenuation (arterial or preportal phase) by the peak aortic CT number increase, because HPP is negligible during this phase. The HPP was calculated by dividing the peak gradient of the liver time-attenuation curve after the peak splenic attenuation by the peak portal trunk CT number increase. The TLP was calculated by adding the HAP and HPP values, and
•
1843
the HAPI equals HAP divided by TLP. All ROIs contained at least 20 pixels (256 ⫻ 256 matrix) to minimize the effects of noise. The quantitative maps of CT perfusion parameters were automatically created by the perfusion software (Fig 2). The quality of the perfusion acquisition was estimated from the attenuation curves over the time obtained in the different regions of interest. If reliable measurements could be performed at the selected slice, the quality was rated as good, and if no reliable measurements could be obtained, the quality was rated as poor. To minimize the calculating bias, all perfusion parameters were calculated three times, and the average was calculated for every patient. The lipiodol deposition in the tumor after chemoembolization was depicted on the unenhanced CT images. Two radiologists with 8 and 6 years of abdominal CT experience, respectively, independently reviewed the CT images to observe the lipiodol deposition in the tumor after chemoembolization. The radiologists performing the lipiodol area measurements were blinded to the findings of the prechemoembolization perfusion imaging. Any discrepancy between the two rates was resolved by consensus. The lipiodol accumulation percentage was calculated from the following measurements: (a) the area of tumor with lipiodol accumulation and (b) the area of the tumor as a whole. The lipiodol deposition in the tumor after chemoembolization was divided into three grades (12–14). Grade I was defined as the iodized oil remaining and dispersing throughout the tumor or the area of iodized oil deposition over 60% of the tumor. Grade II was defined as the area of iodized oil deposition over 20%– 60% of the tumor. Grade III was defined as no accumulation or only a vague accumulation of iodized oil in the tumor or an area of iodized oil deposition less than 20% of the tumor (Fig 3). Statistical Analysis SPSS 12.0 (SPSS. Inc., Chicago, Illinois) was applied to conduct statistical analysis. The CT perfusion parameters were expressed as the mean ⫾ standard deviation (SD). The correlation between the CT perfusion parameters of HCC before chemoembolization
1844
•
Correlation between Tumor Perfusion and Lipiodol Deposition
Figure 2. The axial perfusion images before transarterial chemoembolization were created by the maximum slope method. The tumor shows an increased hepatic arterial perfusion and decreased hepatic portal perfusion compared with the normal parenchyma. (a) HAP image, (b) HPP image, (c) TLP image, and (d) HAPI image.
and the grades of lipiodol deposition in the tumor after chemoembolization were assessed with Spearman correlation analysis. The P value cutoff for statistical significance was set at .05 (two-tailed level of significance).
RESULTS All of the analyzed tumors have the typical arterial enhancement and portal venous washout characteristics on multiphase CT imaging. Before chemoembolization, the CT perfusion parameters of the tumors were calculated (Table 2), and quantitative maps were created in all 24 patients. All patients had good ratings of perfusion acquisition. Four weeks after chemoembolization, the diameter of the tumors on the CT images was 9.0 ⫾ 5.4 cm, which did not significantly change when compared with that before chemoembolization (P ⫽ .795). The lipiodol deposition in the tumor after chemoembolization was depicted on the axial CT images. There was no correlation between tumor size
and quality of lipiodol deposition. According to the criterion explained in the materials and methods section, 12, six, and six cases, respectively, were rated as grade I, grade II, and grade III iodized oil deposition. The HAP and TLP values of the tumors before chemoembolization showed correlation with the grades of the lipiodol deposition in the tumors after chemoembolization (r ⫽ 0.768, P ⬍ .0001 and r ⫽ 0.616, P ⫽ .001, respectively), whereas the HPP and HAPI values did not demonstrate a correlation with the grades of lipiodol deposition in the tumor (r ⫽ 0.227, P ⫽ .286 and r ⫽ 0.111, P ⫽ .607, respectively) (Table 2).
DISCUSSION CT perfusion imaging is a new application for the quantification of tissue perfusion using dynamic CT scanning, which is performed after intravenous bolus administration of an iodinated contrast agent. This technique can quantify perfusion in abso-
December 2010
JVIR
lute units and at high spatial resolution (7,8,15). In this study, we studied liver tumor perfusion with CT perfusion imaging, and we investigated the correlation between tumor perfusion and lipiodol deposition in the tumor tissue after chemoembolization. We found that the values of HAP and TLP of the tumors correlated well with the grades of lipiodol deposition in the tumors after chemoembolization, whereas the HPP and HAPI values of the tumors were not related to the grades of lipiodol deposition in tumor. Although the data in our study need further validation in larger cohorts, they indicate the potential role of MDCT perfusion imaging in HCC management. According to our data, MDCT perfusion imaging may be performed in HCC patients who are being considered for chemoembolization. Increased HAP on MDCT correlated well with lipiodol deposition after chemoembolization and may have a role in selecting more appropriate patients for transarterial locoregional therapies. The correlation of the hypervascularity of HCC on angiography has been documented (5,6). Kanematsu et al (5) studied 200 patients who had HCC and found that HCC in the form of hypervascular masses as seen on the routine angiography showed a good accumulation of lipiodol when compared with hypovascular HCC. In 79 of the 200 patients, CT examination was carried out before and 1 month after chemoembolization. Among those 79 patients, 70 showed a good accumulation of lipiodol in the tumor, and nine showed poor deposition. A decrease in the size of the tumor was evident in 30 of the 70 patients with good deposition, and the tumor did not decrease in size in the nine patients with poor deposition. Kenji et al (6) studied the clinical and pathologic characteristics of 14 cases of HCC that achieved total tumor necrosis in response to transcatheter arterial embolization. On lipiodol CT, lipiodol was densely and homogeneously retained within the whole tumor. Histologic examination found a trabecular pattern with broad blood spaces in which lipiodol was positive with Sudan III staining. They found that deposition of lipiodol throughout the tumor was essential, and the cases showed arterial hypervascularity. These studies in-
Volume 21
Number 12
Yang et al
•
1845
Figure 3. The grades of lipiodol deposition in the tumor after chemoembolization on unenhanced CT images. Grade I was defined as the lipiodol depositing very completely in the tumor (a). Grade II was defined as the lipiodol depositing in some area of the tumor (b). Grade III was defined as only a vague accumulation of lipiodol in the tumor (c).
Table 2 Correlation between CT Perfusion Parameters and Lipiodol Deposition in HCC after Chemoembolization Parameters
Grade I
Grade II
Grade III
R square
r
P Value
Hepatic arterial perfusion (mL/min/mL) Hepatic portal perfusion (mL/min/mL) Total liver perfusion (mL/min/mL) Hepatic arterial perfusion index (%)
0.561 ⫾ 0.213 0.425 ⫾ 0.356 0.985 ⫾ 0.390 61.2 ⫾ 21.4
0.329 ⫾ 0.126 0.242 ⫾ 0.157 0.572 ⫾ 220 59.9 ⫾ 18.6
0.211 ⫾ 0.079 0.278 ⫾ 0.235 0.488 ⫾ 0.276 51.1 ⫾ 20.9
0.462 0.059 0.322 0.039
0.768 0.227 0.616 0.111
⬍ .0001 .286 .001 .607
dicated that hypervascular HCC showed good lipiodol deposition after chemoembolization. In recent years, CT perfusion imaging has been developed to monitor liver tumor interventional therapy. Tsushima et al (16) studied the hemodynamic parameters of 36 malignant liver tumors in 28 patients. In four patients who underwent chemoembolization, CT perfusion imaging was performed before and after chemoembolization. They found that liver tumors were seen as hypervascular lesions on arterial perfusion CT, and arterial perfusion of the liver tumors was obviously decreased after successful chemoembolization. In our study, we focused on the correlation of tumor perfusion and lipiodol deposition in HCC after transarterial chemoembolization. We found that the HAP and TLP of the tumors before chemoembolization had significant correlation with the grades of lipiodol deposition in the tumors after chemoembolization. The ethiodol-retention grading scale was reported in the literature (12–14), but it has not been unified. Matsuo et al (12) classified the lipiodol CT findings into four types: type I, homogeneous accumulation; type II, partial
defect; type III, sporadic accumulation; and type IV, punctate or no accumulation. According to the studies of Catalano et al (13,14), ethiodol uptake was classified as: 0, absent; I, lower than 10% of the tumor volume; II, lower than 50%; III, higher than 50%; and IV, homogeneous. In our study, because the group was limited, the ethiodol uptake was classified into three types. This study has several limitations. First, the disadvantage of CT perfusion imaging with a single-slice CT scanner is the limited sample volume, which makes the choice of location for the principal investigation such that only a limited section of the organ of interest can be studied (16). Using an MDCT scanner, more tumor tissue can be examined, which may help overcome these limitations (9,17,18). Second, perfusion CT study will increase radiation exposure. Further research is needed to reduce the total radiation dose. Third, patient motion out of the image plane caused by breathing will lead to loss of data for perfusion CT. Respiratory gating and careful instruction of the patient may ameliorate some of the motion problems. Fourth, the clearance of lipiodol may occur
with time. To study this parameter, the timeframe between chemoembolization and the measurement of lipiodol deposition would need to be considered. Finally, there are a few factors that effect lipiodol uptake (eg, catheter position, adequate antegrade flow, and chemoembolization technique). Selectivity of chemoembolization mainly depends on the identification and catheterization of the tumor-feeding arteries. Postchemoembolization CT may demonstrate poor lipiodol uptake regardless of the tumor vascular characteristics if there is poor antegrade flow to the tumor or if the tumor-feeding vessels are excluded from the treatment field secondary to catheter position. In this study, the tumor-feeding artery and appropriate catheter position were evaluated only by angiography. Recently, it has been reported that C-arm CT is superior to angiography for identifying tumorfeeding arteries during superselective chemoembolization for HCC (19). It will be used to evaluate if tumors are covered completely by local treatment and to change the catheter position if necessary (20). This study focuses on MDCT perfusion imaging and its ability to predict postchemoembolization
1846
•
Correlation between Tumor Perfusion and Lipiodol Deposition
lipiodol deposition in the tumor. Clinical outcomes, such as imaging response and survival, have not been studied and should be the focus of future analyses. Our results indicate that tumor perfusion correlated with the grades of lipiodol deposition in HCC after chemoembolization. MDCT perfusion imaging has the potential to help select more appropriate patients with HCC for chemoembolization. References 1. Lau WY, Lai EC. Hepatocellular carcinoma: current management and recent advances. Hepatobiliary Pancreat Dis Int 2008; 7:237–257. 2. Marelli L, Stigliano R, Triantos C, et al. Transarterial therapy for hepatocellular carcinoma: which technique is more effective? A systematic review of cohort and randomized studies. Cardiovasc Intervent Radiol 2007; 30:6 –25. 3. Lo CM, Ngan H, Tso WK, et al. Randomized controlled trial of transarterial lipiodol chemoembolization for unresectable hepatocellular carcinoma. Hepatology 2002; 35:1164 –1171. 4. 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; 359:1734 – 1739. 5. Kanematsu T, Furuta T, Takenaka K, et al. A 5-year experience of lipiodolization. Hepatology 1989; 10:98 –102.
6. Kenji J, Hyodo I, Tanimizu M, et al. Total necrosis of hepatocellular carcinoma with a combination therapy of arterial infusion of chemotherapeutic lipiodol and transcatheter arterial embolization: report of 14 cases. Semin Oncol 1997; 24:S671–S680. 7. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993; 188:405– 411. 8. Fournier LS, Cuenod CA, de Bazelaire C, et al. Early modifications of hepatic perfusion measured by functional CT in a rat model of hepatocellular carcinoma using a blood pool contrast agent. Eur Radiol 2004; 14:2125–2133. 9. Kobayashi T, Hayashi T, Funabasama S, et al. Three-dimensional perfusion imaging of hepatocellular carcinoma using 256-slice multidetector-row computed tomography. Radiat Med 2008; 26:557–561. 10. Youn SW, Kim JH, Weon YC, et al. Perfusion CT of the brain using 40mm-wide detector and toggling table technique for initial imaging of acute stroke. AJR Am J Roentgenol 2008; 191: w120 –w126. 11. Ohishi H, Uchida H, Yoshimura H, et al. Hepatocellular carcinoma detected by iodized oil. Radiology 1985; 154:19 –24. 12. Matsuo N, Uchida H, Sakaguchi H, et al. Optimal lipiodol volume in transcatheter arterial chemoembolotherapy for hepatocellular carcinoma: study based on lipiodol accumulation patterns and histopathologic findings. Semin Oncol 1997; 24:S661–S670.
December 2010
JVIR
13. Catalano O, Esposito M, Sandomenico F, et al. Multiphasic helical computerized tomography of hepatocarcinoma. Assessment after chemoembolization. Radiol Med 2000; 99:456 – 460. 14. Catalano O, Cusati B, Esposito M, et al. The correlation between Doppler echography with a contrast medium and CT in the study of a hepatocarcinoma submitted to chemoembolization. Radiol Med 1998; 95:608 – 613. 15. Pandharipande PV, Krinsky GA, Rusinek H, et al. Perfusion imaging of the liver: current challenges and future goals. Radiology 2005; 234:661– 673. 16. Tsushima Y, Funabasama S, Aoki J, et al. Quantitative perfusion map of malignant liver tumors, created from dynamic computed tomography data. Acad Radiol 2004; 11:215–223. 17. Ippolito D, Sironi S, Pozzi M, et al. Perfusion CT in cirrhotic patients with early stage hepatocellular carcinoma: assessment of tumor-related vascularization. Eur J Radiol 2008; 73:148 –152. 18. Laghi A. Multidetector CT (64 Slices) of the liver: examination techniques. Eur Radiol 2007; 17:675– 683. 19. Iwazawa J, Ohue S, Mitani T, et al. Identifying feeding arteries during TACE of hepatic tumors: comparison of C-arm CT and digital subtraction angiography. AJR Am J Roentgenol 2009; 192:1057–1063. 20. Huppert PE, Firlbeck G, Meissner OA, et al. C-arm CT for chemo-embolization of liver tumors. Radiologe 2009; 49:830 – 836.
PURPOSE: To study the correlation of tumor perfusion with lipiodol deposition in hepatocellular carcinoma (HCC) after transarterial chemoembolization with multidetector computed tomography (MDCT) perfusion imaging. MATERIALS AND METHODS: MDCT perfusion imaging was performed in 24 patients with HCC 1 to 7 days before chemoembolization. The computed tomography (CT) perfusion parameters, such as hepatic arterial perfusion (HAP), hepatic portal perfusion (HPP), total liver perfusion (TLP), and hepatic arterial perfusion index (HAPI), were calculated with the slope method. The follow-up CT scans (noncontrast) were performed 4 weeks after chemoembolization to analyze lipiodol deposition. The lipiodol deposition in the tumor was classified into three grades and compared with CT perfusion parameters before chemoembolization. RESULTS: The HAP and TLP of tumors before chemoembolization were correlated with the grades of lipiodol deposition in tumors after chemoembolization (r ⴝ 0.768, P < .0001 and r ⴝ 0.616, P ⴝ .001, respectively). However, the HPP and HAPI of the tumors were not related to the grades of iodized oil deposition (r ⴝ 0.227, P ⴝ .286 and r ⴝ 0.111, P ⴝ .607, respectively). Higher HAP was correlated with better lipiodol deposition, and lower HAP was correlated with poorer lipiodol deposition. CONCLUSIONS: MDCT perfusion imaging has the potential to help select more appropriate patients with HCC for chemoembolization. J Vasc Interv Radiol 2010; 21:1841–1846 Abbreviations: DSA ⫽ digital subtraction angiography, FOV ⫽ field of view, HAP ⫽ hepatic arterial perfusion, HAPI ⫽ hepatic arterial perfusion index, HCC ⫽ hepatocellular carcinoma, HPP ⫽ hepatic portal perfusion, MDCT ⫽ multidetector computed tomography, ROI ⫽ region of interest, SD ⫽ standard deviation, TLP ⫽ total liver perfusion
HEPATOCELLULAR carcinoma (HCC) is a common malignant tumor with a
From Sichuan Key Laboratory of Medical Imaging (L.Y., X.M.Z., W.T., Z.H.Z., G.L.D), Department of Radiology, Affiliated Hospital of North Sichuan Medical College, Nanchong, Sichuan 637000, P. R. China and Department of Radiology (X.P.Z., Y.S.G.), West China Hospital, Sichuan University, Chengdu, Sichuan P. R. China. Received August 21, 2009; final revision received June 30, 2010; accepted August 2, 2010. Address correspondence to L.Y.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2010 DOI: 10.1016/j.jvir.2010.08.015
very high mortality rate, and it has become a major health problem worldwide (1). The majority of patients with HCC are unable to receive surgical resection or transplantation because of their poor hepatic function and the typically advanced nature of the disease at presentation. Transarterial chemoembolization has proven to be effective for unresectable HCC (2– 4). Previous findings indicate that the grades of iodized oil deposition in the liver tumor correlated well with the antitumor effect, and the tumor hemodynamics correlated with lipiodol deposition in tumor tissue (5,6). Therefore, the evaluation of liver tu-
mor hemodynamics will help to select more appropriate patients with liver tumors to undergo chemoembolization. Computed tomography (CT) perfusion imaging, which has emerged in recent years, has the ability to provide quantitative information about both arterial and portal perfusion of the liver tumor (7,8). Multidetector computed tomography (MDCT) scanners support dynamic scanning with a larger range and are therefore thought to have the ability to provide a larger amount of perfusion-related data than conventional single-slice CT (9,10). We hypothesized that CT perfusion
1841
1842
•
Correlation between Tumor Perfusion and Lipiodol Deposition
December 2010
JVIR
Table 1 Baseline Patient Demographics and Tumor Characteristics
No.
Age Gender (yr)
Tumor Number
Chemoembolization Basis
Hepatic Arterial Perfusion (mL/ min/mLl)
Hepatic Portal Perfusion (mL/ min/mL)
Total Liver Perfusion (mL/ min/mL)
Hepatic Arterial Perfusion Index (%)
Lipiodol Deposition Grade
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Male Male Male Male Male Male Male Male Male Female Male Male Male Female Male Female Male Male Male Male Male Male Female Male
Multiple Multiple Solitary Solitary Multiple Solitary Multiple Solitary Solitary Solitary Solitary Multiple Multiple Solitary Solitary Multiple Multiple Multiple Solitary Multiple Multiple Solitary Multiple Solitary
Lobar Lobar Lobar Lobar Whole liver Lobar Whole liver Lobar Lobar Lobar Lobar Whole liver Whole liver Whole liver Lobar Lobar Whole liver Whole liver Lobar lobar Whole liver Lobar Whole liver Lobar
0.364 0.254 0.273 0.558 0.400 1.109 0.345 0.565 0.195 0.639 0.180 0.128 0.177 0.238 0.624 0.348 0.430 0.436 0.513 0.669 0.254 0.600 0.512 0.158
0.391 0.093 0.360 1.220 0.156 0.305 0.144 0.947 0.119 0.265 0.274 0.051 0.316 0.080 0.000 0.557 0.095 0.188 0.458 0.393 0.715 0.294 0.226 0.573
0.755 0.347 0.63.3 1.77.8 0.556 1.413 0.489 1.512 0.314 0.900 0.454 0.179 0.493 0.318 0.624 0.905 0.525 0.624 0.971 1.062 0.969 0.894 0.738 0.731
48.2 73.2 43.1 31.4 71.9 78.5 70.6 37.4 62.1 71.0 39.6 71.5 35.9 74.8 100.0 38.5 81.9 69.9 52.8 63.0 26.2 67.1 69.4 21.6
I III II I I I II I III I III III II II I III II I II I I I I III
28 43 50 63 45 56 51 51 50 43 70 58 42 53 51 40 40 51 70 48 56 55 70 55
parameters would correlate with lipiodol deposition in the liver tumor after chemoembolization and that this may help in selecting more appropriate patients with HCC to undergo the chemoembolization procedure. We therefore conducted this study to investigate the correlation between tumor perfusion and lipiodol deposition in HCC after chemoembolization.
Pugh A and B. In this group, 12 patients had a solitary tumor, and 12 patients had more than one tumor (multiple tumors) (Table 1). The diameter of the tumors was 9.1 ⫾ 5.0 cm on CT images. All of the 24 patients were scheduled to undergo chemoembolization therapy and were recruited in this study.
MATERIALS AND METHODS
MDCT Imaging
Patients
MDCT was performed in all patients 1 to 7 days (average, 3.1 ⫾ 1.3 days) before chemoembolization with 16-slice helical CT (Philips Brilliance 16; Philips Medical Systems, Best, The Netherlands). The CT perfusion imaging before chemoembolization was performed at an appropriate level through the tumor based on unenhanced images. In patients with multiple tumors, only one tumor near the portal vein trunk was studied. Each subject was injected intravenously with 50 mL contrast material (Ultravist 300; Schering, Berlin, Germany) with a power injector with a flow rate of 6 mL/sec. Acquisition started 6 seconds after the contrast agent injection
This prospective study was approved by our Institutional Review Board, and patient informed consent was waived. From June 2006 to May 2008, 24 patients with unresectable HCC proven by biopsy were admitted to our hospital. Of the 24 patients, 20 were men and 4 were women, with an average age of 51.6 ⫾ 10.2 years (range, 28 –70 years). All of the 24 patients had chronic viral hepatitis B infection and liver cirrhosis. The mean value of the Child-Pugh score of the 24 patients was 6.2 ⫾ 1.6 points. Twentyone (88%) and three (12%) patients were, respectively, classified as Child-
and lasted for 90 seconds. Patients were asked to hold their breath as long as possible during the CT perfusion imaging. The CT perfusion images were then obtained at the selected slice (eight sections with 3-mm section width) of the tumor. This sequence provided a total of 480 axial images for each patient. The scan parameters for this perfusion sequence were as follows: field of view (FOV) of 320 mm, matrix size of 1024 ⫻ 1024, scan speed of 1.0 sec/rot, tube voltage of 120 kV, and tube current of 150 mA. Finally, dynamic-enhanced scanning (120 kV, 300 mAs) was performed as in our normal workup routine at 25 seconds and 60 seconds after intravenous injection of a supplemental dose of 80 mL of the same contrast media at a rate of 3.0 mL/sec. Chemoembolization Procedure Two radiologists with more than 20 and 6 years of experience in interventional radiology, respectively, performed the chemoembolization procedure. Diagnostic digital subtraction angiography (DSA) was performed in
Volume 21
Number 12
Yang et al
Figure 1. The ROI was outlined on the pretreatment CT image. The ROI of the aorta (A1), portal vein (V1), and spleen (B1) were placed in the center (a). The ROI of the tumor (T1) was placed at least 1 cm away from the tumor border (b).
a DSA unit (LCV plus; GE Medical Systems, Milwaukee, Wisconsin) with the use of a 5-F RH angiographic catheter (Terumo, Fujinomiya, Japan) to assess the location and number of vessels feeding the tumors and to determine the optimal catheter position for chemoembolization. Sometimes, a 3-F microcatheter was also used to perform the catheterization. In the 24 patients, DSA was followed by the chemoembolization procedure. After catheterization, chemoembolization was performed by the administration of 5-fluorouracil (1000 mg–1500 mg) and hydroxycamptothecine (30 mg– 40 mg), followed by lipiodol (3–20 mL) (Lipiodol UltraFluid, Laboratoire Guerbet, Aulnay-Sous-Bois, France) with an Adriamycin (40 mg–50 mg) emulsion and gelfoam particles (with a diameter of 1 mm, 20 – 60 particles) (11). The volume of the embolus was determined by taking into account the hepatic serum functional indices (serum albumin level, serum bilirubin level, and prothrombin time ratio) and the diameter of the lesion. The chemoembolization procedure was stopped when the tumor stain disappeared or decreased apparently or when the patient could no longer tolerate the procedure. The degree of stasis was obtained at the physician’s discretion. CT Image Analysis The follow-up CT scans (noncontrast) were performed 4 weeks after chemoembolization to analyze lipiodol deposition. The raw data of CT perfusion imaging and follow-up CT
scans were transferred to a Philips Brilliance workstation (version 2.0.11, Philips Medical Systems, Best, The Netherlands). One radiologist with 8 years of experience in abdominal CT and 5 years of experience in perfusion CT reviewed the perfusion CT images and measured the CT perfusion parameters. The region-of-interest (ROI) was drawn manually in the center of the aorta, portal vein, and spleen. The following rules were applied when demarcating the ROI: (a) the ROI was placed at least 1 cm away from the tumor border, (b) the ROI did not include the obviously necrotic tissue defined on conventional CT images, and (c) ROI did not include vessels (Fig 1). CT perfusion values of hepatic arterial perfusion (HAP), hepatic portal perfusion (HPP), total liver perfusion (TLP) and hepatic arterial perfusion index (HAPI) were calculated with the slope method. In calculating perfusion values, the time of maximum enhancement within a splenic ROI was used to separate the arterial and portal phases of liver enhancement as previously described by Miles et al (7). In short, HAP (mL/min/mL) was determined by dividing the peak gradient of the liver time-attenuation curve before the peak splenic attenuation (arterial or preportal phase) by the peak aortic CT number increase, because HPP is negligible during this phase. The HPP was calculated by dividing the peak gradient of the liver time-attenuation curve after the peak splenic attenuation by the peak portal trunk CT number increase. The TLP was calculated by adding the HAP and HPP values, and
•
1843
the HAPI equals HAP divided by TLP. All ROIs contained at least 20 pixels (256 ⫻ 256 matrix) to minimize the effects of noise. The quantitative maps of CT perfusion parameters were automatically created by the perfusion software (Fig 2). The quality of the perfusion acquisition was estimated from the attenuation curves over the time obtained in the different regions of interest. If reliable measurements could be performed at the selected slice, the quality was rated as good, and if no reliable measurements could be obtained, the quality was rated as poor. To minimize the calculating bias, all perfusion parameters were calculated three times, and the average was calculated for every patient. The lipiodol deposition in the tumor after chemoembolization was depicted on the unenhanced CT images. Two radiologists with 8 and 6 years of abdominal CT experience, respectively, independently reviewed the CT images to observe the lipiodol deposition in the tumor after chemoembolization. The radiologists performing the lipiodol area measurements were blinded to the findings of the prechemoembolization perfusion imaging. Any discrepancy between the two rates was resolved by consensus. The lipiodol accumulation percentage was calculated from the following measurements: (a) the area of tumor with lipiodol accumulation and (b) the area of the tumor as a whole. The lipiodol deposition in the tumor after chemoembolization was divided into three grades (12–14). Grade I was defined as the iodized oil remaining and dispersing throughout the tumor or the area of iodized oil deposition over 60% of the tumor. Grade II was defined as the area of iodized oil deposition over 20%– 60% of the tumor. Grade III was defined as no accumulation or only a vague accumulation of iodized oil in the tumor or an area of iodized oil deposition less than 20% of the tumor (Fig 3). Statistical Analysis SPSS 12.0 (SPSS. Inc., Chicago, Illinois) was applied to conduct statistical analysis. The CT perfusion parameters were expressed as the mean ⫾ standard deviation (SD). The correlation between the CT perfusion parameters of HCC before chemoembolization
1844
•
Correlation between Tumor Perfusion and Lipiodol Deposition
Figure 2. The axial perfusion images before transarterial chemoembolization were created by the maximum slope method. The tumor shows an increased hepatic arterial perfusion and decreased hepatic portal perfusion compared with the normal parenchyma. (a) HAP image, (b) HPP image, (c) TLP image, and (d) HAPI image.
and the grades of lipiodol deposition in the tumor after chemoembolization were assessed with Spearman correlation analysis. The P value cutoff for statistical significance was set at .05 (two-tailed level of significance).
RESULTS All of the analyzed tumors have the typical arterial enhancement and portal venous washout characteristics on multiphase CT imaging. Before chemoembolization, the CT perfusion parameters of the tumors were calculated (Table 2), and quantitative maps were created in all 24 patients. All patients had good ratings of perfusion acquisition. Four weeks after chemoembolization, the diameter of the tumors on the CT images was 9.0 ⫾ 5.4 cm, which did not significantly change when compared with that before chemoembolization (P ⫽ .795). The lipiodol deposition in the tumor after chemoembolization was depicted on the axial CT images. There was no correlation between tumor size
and quality of lipiodol deposition. According to the criterion explained in the materials and methods section, 12, six, and six cases, respectively, were rated as grade I, grade II, and grade III iodized oil deposition. The HAP and TLP values of the tumors before chemoembolization showed correlation with the grades of the lipiodol deposition in the tumors after chemoembolization (r ⫽ 0.768, P ⬍ .0001 and r ⫽ 0.616, P ⫽ .001, respectively), whereas the HPP and HAPI values did not demonstrate a correlation with the grades of lipiodol deposition in the tumor (r ⫽ 0.227, P ⫽ .286 and r ⫽ 0.111, P ⫽ .607, respectively) (Table 2).
DISCUSSION CT perfusion imaging is a new application for the quantification of tissue perfusion using dynamic CT scanning, which is performed after intravenous bolus administration of an iodinated contrast agent. This technique can quantify perfusion in abso-
December 2010
JVIR
lute units and at high spatial resolution (7,8,15). In this study, we studied liver tumor perfusion with CT perfusion imaging, and we investigated the correlation between tumor perfusion and lipiodol deposition in the tumor tissue after chemoembolization. We found that the values of HAP and TLP of the tumors correlated well with the grades of lipiodol deposition in the tumors after chemoembolization, whereas the HPP and HAPI values of the tumors were not related to the grades of lipiodol deposition in tumor. Although the data in our study need further validation in larger cohorts, they indicate the potential role of MDCT perfusion imaging in HCC management. According to our data, MDCT perfusion imaging may be performed in HCC patients who are being considered for chemoembolization. Increased HAP on MDCT correlated well with lipiodol deposition after chemoembolization and may have a role in selecting more appropriate patients for transarterial locoregional therapies. The correlation of the hypervascularity of HCC on angiography has been documented (5,6). Kanematsu et al (5) studied 200 patients who had HCC and found that HCC in the form of hypervascular masses as seen on the routine angiography showed a good accumulation of lipiodol when compared with hypovascular HCC. In 79 of the 200 patients, CT examination was carried out before and 1 month after chemoembolization. Among those 79 patients, 70 showed a good accumulation of lipiodol in the tumor, and nine showed poor deposition. A decrease in the size of the tumor was evident in 30 of the 70 patients with good deposition, and the tumor did not decrease in size in the nine patients with poor deposition. Kenji et al (6) studied the clinical and pathologic characteristics of 14 cases of HCC that achieved total tumor necrosis in response to transcatheter arterial embolization. On lipiodol CT, lipiodol was densely and homogeneously retained within the whole tumor. Histologic examination found a trabecular pattern with broad blood spaces in which lipiodol was positive with Sudan III staining. They found that deposition of lipiodol throughout the tumor was essential, and the cases showed arterial hypervascularity. These studies in-
Volume 21
Number 12
Yang et al
•
1845
Figure 3. The grades of lipiodol deposition in the tumor after chemoembolization on unenhanced CT images. Grade I was defined as the lipiodol depositing very completely in the tumor (a). Grade II was defined as the lipiodol depositing in some area of the tumor (b). Grade III was defined as only a vague accumulation of lipiodol in the tumor (c).
Table 2 Correlation between CT Perfusion Parameters and Lipiodol Deposition in HCC after Chemoembolization Parameters
Grade I
Grade II
Grade III
R square
r
P Value
Hepatic arterial perfusion (mL/min/mL) Hepatic portal perfusion (mL/min/mL) Total liver perfusion (mL/min/mL) Hepatic arterial perfusion index (%)
0.561 ⫾ 0.213 0.425 ⫾ 0.356 0.985 ⫾ 0.390 61.2 ⫾ 21.4
0.329 ⫾ 0.126 0.242 ⫾ 0.157 0.572 ⫾ 220 59.9 ⫾ 18.6
0.211 ⫾ 0.079 0.278 ⫾ 0.235 0.488 ⫾ 0.276 51.1 ⫾ 20.9
0.462 0.059 0.322 0.039
0.768 0.227 0.616 0.111
⬍ .0001 .286 .001 .607
dicated that hypervascular HCC showed good lipiodol deposition after chemoembolization. In recent years, CT perfusion imaging has been developed to monitor liver tumor interventional therapy. Tsushima et al (16) studied the hemodynamic parameters of 36 malignant liver tumors in 28 patients. In four patients who underwent chemoembolization, CT perfusion imaging was performed before and after chemoembolization. They found that liver tumors were seen as hypervascular lesions on arterial perfusion CT, and arterial perfusion of the liver tumors was obviously decreased after successful chemoembolization. In our study, we focused on the correlation of tumor perfusion and lipiodol deposition in HCC after transarterial chemoembolization. We found that the HAP and TLP of the tumors before chemoembolization had significant correlation with the grades of lipiodol deposition in the tumors after chemoembolization. The ethiodol-retention grading scale was reported in the literature (12–14), but it has not been unified. Matsuo et al (12) classified the lipiodol CT findings into four types: type I, homogeneous accumulation; type II, partial
defect; type III, sporadic accumulation; and type IV, punctate or no accumulation. According to the studies of Catalano et al (13,14), ethiodol uptake was classified as: 0, absent; I, lower than 10% of the tumor volume; II, lower than 50%; III, higher than 50%; and IV, homogeneous. In our study, because the group was limited, the ethiodol uptake was classified into three types. This study has several limitations. First, the disadvantage of CT perfusion imaging with a single-slice CT scanner is the limited sample volume, which makes the choice of location for the principal investigation such that only a limited section of the organ of interest can be studied (16). Using an MDCT scanner, more tumor tissue can be examined, which may help overcome these limitations (9,17,18). Second, perfusion CT study will increase radiation exposure. Further research is needed to reduce the total radiation dose. Third, patient motion out of the image plane caused by breathing will lead to loss of data for perfusion CT. Respiratory gating and careful instruction of the patient may ameliorate some of the motion problems. Fourth, the clearance of lipiodol may occur
with time. To study this parameter, the timeframe between chemoembolization and the measurement of lipiodol deposition would need to be considered. Finally, there are a few factors that effect lipiodol uptake (eg, catheter position, adequate antegrade flow, and chemoembolization technique). Selectivity of chemoembolization mainly depends on the identification and catheterization of the tumor-feeding arteries. Postchemoembolization CT may demonstrate poor lipiodol uptake regardless of the tumor vascular characteristics if there is poor antegrade flow to the tumor or if the tumor-feeding vessels are excluded from the treatment field secondary to catheter position. In this study, the tumor-feeding artery and appropriate catheter position were evaluated only by angiography. Recently, it has been reported that C-arm CT is superior to angiography for identifying tumorfeeding arteries during superselective chemoembolization for HCC (19). It will be used to evaluate if tumors are covered completely by local treatment and to change the catheter position if necessary (20). This study focuses on MDCT perfusion imaging and its ability to predict postchemoembolization
1846
•
Correlation between Tumor Perfusion and Lipiodol Deposition
lipiodol deposition in the tumor. Clinical outcomes, such as imaging response and survival, have not been studied and should be the focus of future analyses. Our results indicate that tumor perfusion correlated with the grades of lipiodol deposition in HCC after chemoembolization. MDCT perfusion imaging has the potential to help select more appropriate patients with HCC for chemoembolization. References 1. Lau WY, Lai EC. Hepatocellular carcinoma: current management and recent advances. Hepatobiliary Pancreat Dis Int 2008; 7:237–257. 2. Marelli L, Stigliano R, Triantos C, et al. Transarterial therapy for hepatocellular carcinoma: which technique is more effective? A systematic review of cohort and randomized studies. Cardiovasc Intervent Radiol 2007; 30:6 –25. 3. Lo CM, Ngan H, Tso WK, et al. Randomized controlled trial of transarterial lipiodol chemoembolization for unresectable hepatocellular carcinoma. Hepatology 2002; 35:1164 –1171. 4. 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; 359:1734 – 1739. 5. Kanematsu T, Furuta T, Takenaka K, et al. A 5-year experience of lipiodolization. Hepatology 1989; 10:98 –102.
6. Kenji J, Hyodo I, Tanimizu M, et al. Total necrosis of hepatocellular carcinoma with a combination therapy of arterial infusion of chemotherapeutic lipiodol and transcatheter arterial embolization: report of 14 cases. Semin Oncol 1997; 24:S671–S680. 7. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993; 188:405– 411. 8. Fournier LS, Cuenod CA, de Bazelaire C, et al. Early modifications of hepatic perfusion measured by functional CT in a rat model of hepatocellular carcinoma using a blood pool contrast agent. Eur Radiol 2004; 14:2125–2133. 9. Kobayashi T, Hayashi T, Funabasama S, et al. Three-dimensional perfusion imaging of hepatocellular carcinoma using 256-slice multidetector-row computed tomography. Radiat Med 2008; 26:557–561. 10. Youn SW, Kim JH, Weon YC, et al. Perfusion CT of the brain using 40mm-wide detector and toggling table technique for initial imaging of acute stroke. AJR Am J Roentgenol 2008; 191: w120 –w126. 11. Ohishi H, Uchida H, Yoshimura H, et al. Hepatocellular carcinoma detected by iodized oil. Radiology 1985; 154:19 –24. 12. Matsuo N, Uchida H, Sakaguchi H, et al. Optimal lipiodol volume in transcatheter arterial chemoembolotherapy for hepatocellular carcinoma: study based on lipiodol accumulation patterns and histopathologic findings. Semin Oncol 1997; 24:S661–S670.
December 2010
JVIR
13. Catalano O, Esposito M, Sandomenico F, et al. Multiphasic helical computerized tomography of hepatocarcinoma. Assessment after chemoembolization. Radiol Med 2000; 99:456 – 460. 14. Catalano O, Cusati B, Esposito M, et al. The correlation between Doppler echography with a contrast medium and CT in the study of a hepatocarcinoma submitted to chemoembolization. Radiol Med 1998; 95:608 – 613. 15. Pandharipande PV, Krinsky GA, Rusinek H, et al. Perfusion imaging of the liver: current challenges and future goals. Radiology 2005; 234:661– 673. 16. Tsushima Y, Funabasama S, Aoki J, et al. Quantitative perfusion map of malignant liver tumors, created from dynamic computed tomography data. Acad Radiol 2004; 11:215–223. 17. Ippolito D, Sironi S, Pozzi M, et al. Perfusion CT in cirrhotic patients with early stage hepatocellular carcinoma: assessment of tumor-related vascularization. Eur J Radiol 2008; 73:148 –152. 18. Laghi A. Multidetector CT (64 Slices) of the liver: examination techniques. Eur Radiol 2007; 17:675– 683. 19. Iwazawa J, Ohue S, Mitani T, et al. Identifying feeding arteries during TACE of hepatic tumors: comparison of C-arm CT and digital subtraction angiography. AJR Am J Roentgenol 2009; 192:1057–1063. 20. Huppert PE, Firlbeck G, Meissner OA, et al. C-arm CT for chemo-embolization of liver tumors. Radiologe 2009; 49:830 – 836.