ORIGINAL ARTICLE
Magnetic Resonance-Based Visualization of Thermal Ablative Margins Around Hepatic Tumors by Means of Systemic Ferucarbotran Administration Before Radiofrequency Ablation Animal Study to Reveal the Connection Between Excess Iron Deposition and T2*-Weighted Hypointensity in Ablative Margins Michiko Nagai, MD,*† Masayuki Yamaguchi, MD, PhD,* Kensaku Mori, MD, PhD,*† Toshihiro Furuta, MD, PhD,*‡ Hiroki Ashino, MSc,§ Hiroyuki Kurosawa, DVM,§ Hiroyuki Kasahara, PhD,§ Manabu Minami, MD, PhD,† and Hirofumi Fujii, MD, PhD* Objective: The objective of this study was to demonstrate experimentally that radiofrequency ablation (RFA) of ferucarbotran-accumulated healthy liver tissues causes excess iron deposition in the ablated liver tissues on postablation days and produces sustained T2*-weighted low signals indicative of ablative margins surrounding hepatic tumors. Materials and Methods: We conducted 3 experiments using 30 rats. In experiment 1, we administered either ferucarbotran (n = 6) or saline (n = 4), acquired T2*weighted images (T2*-WIs) of the liver by using a 3-T magnetic resonance scanner, and subsequently performed RFA of healthy liver lobes. We acquired follow-up T2*WIs up to day 7 and histologically analyzed the liver specimens. In another 4 rats, we performed sham operation, instead of RFA, in ferucarbotran-accumulated liver lobes, followed by the same image acquisition and histological analysis. In experiment 2, we administered 59Fe-labeled ferucarbotran, subsequently performed either RFA (n = 4) or sham operation (n = 4) in the liver, and acquired autoradiograms of the liver specimens on day 7. In experiment 3, we conducted RFA treatment for 8 rats bearing orthotopic hepatic tumors after ferucarbotran administration and monitored tumor growth by using serial T2*-WIs. Results: On days 4 and 7 of the experiment 1, T2*-WIs of 6 rats with systemic ferucarbotran administration and subsequent hepatic RFA showed low-signal regions indicative of ablated liver tissues, whereas high-signal areas were seen in 4 saline-administered rats. Neither high nor low signal areas were detected in 4 shamoperated rats. Histologically, larger amounts of iron were observed in the RFAinduced necrotic liver tissues in the ferucarbotran-administered rats than in the saline-administered-rats. The 59Fe autoradiography of the rats in experiment 2 revealed accumulation of ferucarbotran-derived iron in necrotic liver tissues. Among 6 hepatic tumors grown in 6 rats of the experiment 3, a total of 4 tumors were stable in size, but the other 2 increased markedly on day 7. Retrospectively, T2*-WIs showed the former tumor sites surrounded completely by low-signal areas on day 4. Conclusions: The RFA of ferucarbotran-accumulated healthy liver tissues in the rats caused excess iron deposition in the ablated liver tissues and produced Received for publication September 17, 2014; and accepted for publication, after revision, November 26, 2014. From the *Division of Functional Imaging, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa, Chiba; †Department of Radiology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki; ‡Department of Radiology, Graduate School of Medicine, University of Tokyo, Tokyo; and §Fujifilm RI Pharma Co, Ltd, Sammu, Chiba, Japan. Conflicts of interest and source of funding: Authors Ashino, Kurosawa, and Kasahara are employees of Fujifilm RI Pharma Co, Ltd; supported by Health and Labor Science Research Grant for the Third-Term Comprehensive 10-Year Strategy for Cancer Control and a Grant-in-Aid for Cancer Research (21–5) from the Ministry of Health, Labor, and Welfare of Japan to author Fujii. Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.investigativeradiology.com) Reprints: Masayuki Yamaguchi, MD, PhD, Division of Functional Imaging, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwanoha 6-5-1, Kashiwa, Chiba 277-8577, Japan. E-mail: [email protected]. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0020-9996/15/5006–0376
376
www.investigativeradiology.com
sustained T2*-weighted hypointense regions. Similar hypointense regions surrounding hepatic tumors were indicative of ablative margins. Key Words: hepatic tumor, thermal ablation, ablative margins, magnetic resonance imaging, superparamagnetic iron oxides (Invest Radiol 2015;50: 376–383)
H
epatocellular carcinoma (HCC) is the second leading cause of cancerrelated deaths worldwide.1 Although liver resection is the criterion standard for patients with resectable HCCs, radiofrequency ablation (RFA) is a potentially curative treatment option for selected patients.2 To enhance curability and prevent local recurrence, a 5-mm ablative margin should be secured around the HCC nodule.3,4 Although contrastenhanced computed tomography and magnetic resonance (MR) imaging are currently available for the delineation of ablative margins,5–8 there are some pitfalls related to changes in hepatic tissue perfusion after RFA treatment. First, both ablated tumors and surrounding ablative margins are typically poorly enhanced owing to the lack of blood flow and cannot be clearly contrasted. Second, ablated liver tissues are typically surrounded by well-enhanced congestive or reactive inflammatory areas mimicking residual tumors, and this can sometimes lead to diagnostic errors. Mori et al9 reported a novel technique using ferucarbotranenhanced MR imaging to clearly display the ablative margin, regardless of post-RFA changes in hepatic perfusion. They administered ferucarbotran to patients with HCC several hours before RFA to tag Kupffer cells in the nontumor liver parenchyma by using superparamagnetic iron oxides (SPIOs). Three to 4 days after RFA treatment, they performed T2*-weighted MR imaging without additional contrast administration and visualized low-signal rims surrounding the high-signal tumor areas. Liver parenchyma outside the low-signal rim showed higher signal intensity because the negative enhancement effects by ferucarbotran in nontumor liver tissues had diminished. They also reported that the disruption of low-signal rims increased the risk for tumor recurrence; hence, the low-signal rims were presumably indicative of ablative margins. Because MR image-histology correlation studies regarding the hypointensity rim sign are very limited,10 we carried out RFA in ferucarbotran-accumulated rat liver lobes and performed serial MR imaging of the liver and histological analysis of liver specimens to test whether RFA of ferucarbotran-accumulated healthy liver tissues causes excess iron deposition in ablated liver tissues and produces sustained T2*weighted low signals (experiment 1). To demonstrate clearly the accumulation of ferucarbotran-derived iron particles in the ablated liver tissues in autoradiograms, we used 59Fe-labeled ferucarbotran in selected cases (experiment 2). Furthermore, we conducted RFA treatment in orthotropic hepatic tumor-bearing rats after systemic ferucarbotran administration to confirm whether sustained T2*-weighted low signals surrounding hepatic tumors are indicative of secured ablative margins (experiment 3). Investigative Radiology • Volume 50, Number 6, June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
MR Diagnosis of Ablative Margins Around Hepatomas
MATERIALS AND METHODS Contrast Materials We purchased ferucarbotran (Resovist) from Fujifilm RI Pharma Inc (Tokyo, Japan). In addition, one of the authors synthesized 59Fe-labeled, as well as nonlabeled, ferucarbotran as reported previously.11 In preliminary experiments, we confirmed the equivalent negative enhancement effects of synthesized ferucarbotran to Resovist (Fig. S1-1, Supplemental Digital Content 1, which indicates negative enhancement effect of the synthesized ferucarbotran to the rat liver, http://links.lww.com/RLI/A192).
next inserted the electrode into the subcapsular area and performed RFA with radiofrequency power at 3 W. We omitted coolant circulation through the electrode needle to localize thermal ablated areas.12 Ablation was completed when the temperature of the electrode tip reached 70°C. It took 60 ± 44 seconds (average ± standard deviation [SD]). The electrode was removed, and the abdominal incision was closed. For each sham-operated rat, we inserted the electrode into the liver but did not apply electric currents. After 3 minutes, the electrode was removed.
MR Imaging Animals Our institutional animal experimental committee approved the animal experimental protocol. We purchased 30 female Sprague-Dawley rats (aged 7–8 weeks; typical body weight, 200 g) from Japan SLC Inc (Hamamatsu, Japan) and acclimatized them for a week.
Experiment 1: Hepatic MR Imaging of Rats With Ferucarbotran Administration Before RFA To demonstrate the relationship between the alteration in T2*weighted signal intensities in the ablated liver tissues and ferucarbotran administration, we conducted this experiment using 14 rats divided into 3 groups. We intravenously administered ferucarbotran (Resovist; 20μmol iron/kg body weight [BW]) and subsequently performed hepatic RFA in the first group (n = 6, group 1). This group was subdivided into 2, and duplicated experiments were performed on separate days. We also administered saline (0.3 mL) and then performed RFA in the second group (n = 4, group 2). In the third group (n = 4, group 3), we performed sham operation instead of RFA after 20-μmol iron/kg BW ferucarbotran administration.
Radiofrequency Ablation We performed hepatic RFA using the Cool-tip RF Ablation System (Covidien, Boulder, CO) and an electrode, originally 10 mm in length (Cool-tip RFAblation Single Electrode Kit; Covidien) and deliberately shortened it to 5 mm to produce small ablative tissues in the rat liver. Three to 4 hours after either ferucarbotran or saline administration, anesthesia was induced by inhalation of 5% isoflurane in air and subsequently maintained with 2% isoflurane in air. Then, we made a midline abdominal incision and explored the left lateral lobe of the liver. We
We acquired MR images of the liver before, approximately 10 minutes after either ferucarbotran or saline administration, as well as on 3 or 4 and 7 days after RFA by using a 3.0-T scanner (Signa HDx; GE Healthcare, Milwaukee, WI) and dedicated multireceiver coil comprising 8 coils in the upper and 8 coils in the lower arrays.13 Anesthesia was induced by inhalation of 5% isoflurane in the gas mixture of oxygen and nitrous oxide (1:1) and subsequently maintained with 1% to 2% isoflurane in the gas mixture. We then placed 2 to 4 rats separately on acrylic animal beds in either the prone or supine position; inserted the rats to the coil with their abdominal surfaces close to the coils; and simultaneously acquired transverse T2-weighted (T2WIs), T2*-weighted (T2*WIs), and proton-density-weighted images of the upper abdomen with the parameters as shown in Table 1.
Histological Analysis Immediately after the completion of the MR imaging acquisition on the seventh day after RFA, we excised liver specimens and fixed them with 10% neutral-buffered formalin for a median of 21 days. We did not recognize any tissue damages on hematoxylin and eosin– (HE) as well as Prussian blue–stained liver specimens from the fixation. Then, we sliced the liver tissues into 4- to 5-mm transverse sections and embedded in paraffin wax. We acquired 2-μm sections of the specimens and stained them with HE and Prussian blue. On the Prussian bluestained specimens, we counted the number of blue dots in 5 randomly selected 0.1-mm2 areas in the ablated liver tissues and in the surrounding nonablated liver parenchyma by using a microscope (BZ-9000; Keyence, Osaka, Japan). In addition, we counted the number of blue dots in the fibrous capsule between the ablated and nonablated liver parenchyma in the same manner.
TABLE 1. MR Scan Parameters Parameters Slice orientation Repetition time, ms Effective echo time, ms Excitation flip angle, degrees Echo train length Receiver band width, kHz FOV, mm Frequency encoding steps Phase encoding steps Matrix size on final images Slice thickness, mm Slice gap, mm No. slices No. excitations Acquisition time
GRE T2*-Weighted Image
PROPELLER T2-Weighted Image
FSE Proton Density-Weighted Image
Transaxial 450 11 30 — 31.3 120 256 160 (phase FOV = 0.75) 512 512 2 0.4 15 4 3 min 35 s
Transaxial 4000 86 90 24 50.0 120 320 — 512 512 2 0.4 15 4.5 4 min 8 s
Transaxial 4000 5.8 90 8 31.3 120 128 128 (phase FOV = 0.8) 512 512 2 0.4 15 1 1 min 4 s
FOV indicates field of view; FSE, fast spin echo; GRE, gradient echo; MR, magnetic resonance; PROPELLER, periodically rotated overlapping parallel lines with enhanced reconstruction.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
377
Investigative Radiology • Volume 50, Number 6, June 2015
Nagai et al
MR Image Analysis After minimizing signal decline depending on the distance from the coil surface with proton density-weighted images–based coil sensitivity maps,14 we measured the areas of either high or low signal regions in the left liver lobes on representative T2*-WIs using ImageJ software (version 1.47, available from rsbweb.nih.gov/ij/) and compared them with the corresponding ablated areas on liver specimens. We also placed the maximum possible regions of interest (ROIs) in these high- or lowsignal areas of the liver lobes and paraspinal muscles, then measured signal intensities in each ROI, subsequently dividing signal intensities in these areas by those in the paraspinal muscle to determine relative signal intensities, and finally determined their temporal changes. Because we recognized neither high- nor low-signal areas in the rats before RFA or in the sham-operated rats, we alternatively measured the signal intensities in the left liver lobes placing equivalent spherical ROIs (typically 12 mm2).
Experiment 2: Hepatic Autoradiography of Rats with 59 Fe Ferucarbotran Administration Before RFA To demonstrate the accumulation of ferucarbotran-derived irons in the ablated liver parenchyma, we administered 59Fe-labeled ferucarbotran to 8 rats at a dose of 40 kBq/head (total iron load adjusted to 20-μmol iron/kg BW) and performed either RFA (n = 4) or sham operation (n = 4) in the left liver lobe of each rat 3 to 4 hours later as mentioned previously. After 7 days, we carried out perfusion fixation with 20% (w/v) sucrose dissolved in phosphate-buffered saline and then excised left liver lobes to obtain 5-mm transverse sections. After embedding and freezing them in optimum cutting temperature compound (Tissue-Tek O.C.T. compound 4583; Sakura Finetek USA, Inc,
Torrance, CA), we sliced the specimens into 20-μm sections by using a cryostat (Leica CM3050 S; Leica Microsystems GmbH, Wetzlar, Germany), subsequently placed the specimens on imaging plates (BAS-III IP; Fujifilm Co, Ltd, Tokyo, Japan) for 6 to 7 days, and finally acquired autoradiograms using an image reader (FLA-7000, Fujifilm).
Image Analysis Subsequently, we manually placed an ROI (2 mm2) in each ablated liver tissue and 3 ROIs of the same size in nonablated liver tissues on representative autoradiograms of the left liver lobes and assessed the signal strength in these ROIs by using densitometric analysis software (Multi Gauge version 3.0; Fujifilm). We also automatically counted signal strength in ring-shaped areas surrounding the ablated liver tissues (histologically corresponding to the fibrous capsule). Ratios of 59Fe accumulation in the ablated liver tissue areas and in the fibrous capsule areas were calculated by using the following equation: 59 Fe accumulation ratio = signal strength in either ablated area or fibrous capsule area/average of signal strengths in nonablated liver tissue areas
Experiment 3: Serial MR Imaging Examination of Hepatic Tumor-Bearing Rat Liver With Systemic Ferucarbotran Administration Followed by RFA Treatment To demonstrate the relationship between the treatment efficacy of RFA to hepatic tumors and sustained T2*-weighted low signals surrounding the tumor, we performed experiments using an orthotopic hepatic tumor model. We inoculated N1-S1 hepatoma cells (CRL-1604; American Type Culture Collection, Manassas, VA; 4 106 cells per
FIGURE 1. Representative series of T2*-weighted MR images of the liver from rats receiving ferucarbotran administration as well as subsequent RFA (A–D, group 1), saline administration and subsequent RFA (E–H, group 2), or ferucarbotran administration and subsequent sham operation (I–L, group 3). The first and second columns show the images acquired before (A, E, and I) and after (B, F, and J) either ferucarbotran or saline administration. The third and fourth columns show the images acquired 3 to 4 days (C, G, and K) and 7 days (D, H, and L) after RFA or sham operation. Note the well-demarcated low-signal areas (arrows in C and D) and ill-demarcated slightly high-signal areas (arrowheads in G and H) in the ventral parts of the left liver lobes indicative of ablated liver tissues. The former low signal areas contain small high signals indicative of electrode needle paths. Bar represents 10 mm.
378
www.investigativeradiology.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
MR Diagnosis of Ablative Margins Around Hepatomas
FIGURE 2. Graphs showing the changes in relative T2*-weighted signal intensities of the ablated, sham-operated and nonablated areas of the liver as a function of time. A, After ferucarbotran administration, relative signal intensities in the liver decrease significantly. In the ablated area (solid line), relative signal intensities remain low at days 4 and 7 after RFA, whereas those in the nonablated area (dotted line) are restored toward the baseline levels. B, After normal saline (NS) administration, hepatic signal intensities show no changes when compared with the baseline. On days 4 and 7 after RFA, relative signal intensities in the ablated area (solid line) are slightly higher than those in the nonablated area (dotted line). C, In sham-operated rats, hepatic signal intensities are significantly reduced after ferucarbotran administration and signal intensities in the sham-operated (solid line) and nonoperated areas (dotted line) are comparably restored to the baseline levels during the experimental period. Average values in relative signal intensities are indicated in arbitrary units. Error bars represent standard deviation. Asterisks indicate statistical significance (P < 0.001, Tukey test). AU indicates arbitrary unit; NS, normal saline.
head) into the left lateral lobes of the liver in 8 rats by following the technique described in the literature.13,15,16 Seven days later, we acquired ferucarbotran-enhanced T2*-WIs of the liver (20-μmol iron /kg BW) and subsequently performed RFA in the hepatic tumor lesions by following the same technique as mentioned previously except for the length of electrode of 5 to 10 mm depending on tumor size. In addition, we carried out multi-RFA sessions for larger tumors. On postablation days 4 and 7, we acquired T2*-WIs of the liver and evaluated temporal changes in tumor size by measuring their maximum perpendicular diameters (MPDs). One of the authors retrospectively analyzed the relationship between the MR imaging appearance and tumor growth. The duplicate experiments were performed on separate days using 4 rats each.
Statistical Tests We assessed the differences in relative signal intensities between ablated and nonablated liver tissues at each time point by using the Tukey test provided by commercially available software (SPSS version 20; IBM Inc, Tokyo, Japan). We also evaluated the differences in the number of blue dots in the ablated, nonablated liver tissues, as well as fibrous capsule between groups 1 and 2 by using the Welch t test. Furthermore, we assessed the differences in 59Fe accumulation ratio between ablated liver tissue and fibrous capsule areas using the same test. Finally, we assessed the difference between high-signal and low-signal
areas indicative of ablated tumor and tumor margins, respectively, on T2*-weighted MR images on day 7 using paired t test. P < 0.01 was considered to be statistically significant.
RESULTS Laparotomy and RFA procedures were completed in all rats without serious complications.
Experiment 1 In group 1 (Fig. 1, A–D), signal intensities in healthy liver were similar to those of the paraspinal muscle on baseline scans, but they decreased significantly after ferucarbotran administration and returned to baseline levels toward day 7. Four days after RFA, approximately 5 mm in diameter, well-demarcated, spheroid hypointense areas appeared in the ventral parts of the left liver lobes where RFA procedures were performed (arrow in Fig. 1C). Seven days after RFA, these areas were still visible (arrow in Fig. 1D). In contrast, poorly demarcated, slightly high-signal areas were seen in the ventral parts of the left liver lobes on days 4 and 7 in group 2 (arrowheads in Fig. 1, G and H). In other liver lobes, signal intensities were comparable with those of the paraspinal muscle throughout the experimental period. In group 3 (Fig. 1, I–L), no abnormal signals were seen in sham-operated liver lobes. Figure 2 summarizes the temporal changes in relative signal
FIGURE 3. Representative specimen of the left liver lobe in a rat with ferucarbotran administration followed by RFA (original magnification 12.5 in a, and 40 in b). Spheroid area of coagulative necrosis is seen in the subcapsular region where RFA was performed (ablated area; Ab). The round defect in the center is the electrode needle path. A thin fibrous capsule (arrow heads) is present between the ablated area and nonablated liver parenchyma.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
379
Investigative Radiology • Volume 50, Number 6, June 2015
Nagai et al
intensities of the ablated and nonablated liver tissues. T2-weighted MR imaging results are summarized in digital content (Fig. S2-1, Supplemental Digital Content 2, which shows T2-weighted MR image series of the liver from rats receiving ferucarbotran administration and subsequent RFA, http://links.lww.com/RLI/A193). Histologically, 12.1 ± 2.4 mm2, well-demarcated, spheroid coagulation necrotic regions were produced in the ventral parts of the left lateral lobes by RFA in group 1 (Fig. 3A). They were well correlated with 14.0 ± 0.6 mm2 and low-signal regions in the corresponding liver lobes seen on T2*-weighted MR images. In group 2, the necrotic regions were 12.4 ± 1.2 mm2 and correlated with the 17.7 ± 7.4 mm2 as well as high-signal areas on T2*-weighted images. Surrounding these coagulative necrotic regions were thin fibrogranulation tissues, known as fibrous capsules (arrowheads in Fig. 3).17,18 These findings
were common in rats in groups 1 and 2. In the sham-operated rats (group 3), approximately 3-mm pale tissues were recognized under the liver capsule. On Prussian blue-stained specimens, blue dots indicative of iron ions were scattered in the dilated sinusoids in RFA-induced necrotic tissues (Fig. 4, A and B). In the fibrous capsule, numerous blue dots were seen except in the liver surface region in groups 1 and 2 (Fig. 4, C and D). In nonablated liver tissues, blue dots were sporadically observed along the sinusoids. The number of blue dots in the ablated area (RFA-induced necrotic tissue) was significantly larger in the ferucarbotran-administered group 1 rats than in the saline-administered group 2 rats (31.3 ± 14.0 vs 5.0 ± 2.0 /mm2; Fig. 5). In the fibrous capsule, the numbers were 579.7 ± 190.2 in group 1 and 264.5 ± 120.5 in group 2. The difference was marginally significant (P = 0.012). In the
FIGURE 4. Representative HE-stained and Prussian blue-stained specimens from the same liver specimens shown in Figure 3. In the ablated area (A, B; original magnification, 400), necrotic and shrunken hepatocytes with small, homogeneous, slightly basophilic nuclei and homogeneous eosinophilic cytoplasm are arranged like the hepatic cord (A). In the slightly dilated sinusoids, several blue dots are present adjacent to Kupffer cells' nuclei (arrow heads, B). C, The fibrous capsule shows a reactive inflammatory change containing numerous inflammatory cells, fibroblasts, and newly developed capillaries. Multinuclear giant cells are occasionally observed, as reported previously.26 D, Prussian blue-stained specimen shows a large number of blue dots in the fibrous capsule, representing iron ion particles. E, Nonablated liver parenchyma shows normal architecture composed of regular hepatic cords and sinusoids. F, Blue dots are scattered in the sinusoid, indicating iron deposition in Kupffer cells. Figure 4 can be viewed online in color at www.investigativeradiology.com.
380
www.investigativeradiology.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
MR Diagnosis of Ablative Margins Around Hepatomas
Experiment 3 On T2*-WIs, 8 high-signal nodules were detected in the left liver in 8 rats on 7 days after tumor cell inoculation, but 2 high-signal regions in 2 rats were excluded from further analysis because of their small sizes (5.5 and 5.6 mm2). The other 6 nodules showed MPDs of 35.4 ± 14.9 mm2 (mean ± SD), ranging from 22.2 to 63.6 mm2. T2*WIs on postablation day 4 demonstrated 4 nodules (MPDs, 34.6 ± 7.7 mm2) surrounded by deep low-signal rims (Fig. 7) and the other 2 nodules (36 and 109 mm2) without the low-signal rim. On postablation day 7, the former 4 nodules were stable in size (MPDs: 25.9 ± 6.5 mm2; growth: 90% ± 19%), whereas the latter 2 lesions had increased in size (85.4 and 149.2 mm2; 235% and 268%). The former 4 lesions were still surrounded by low-signal areas in which MPDs were 74.1 ± 32.6 mm2 larger than high-signal areas indicative of ablated tumor lesions (25.9 ± 6.5 mm2 in MPD). The difference in these areas was marginally significant (P =0.050 in paired t test). In histological specimens, 51.2 ± 11.6 -mm2 ablated liver parenchyma surrounded ablated tumor cell areas (15.4 ± 9.0 mm2) in 3 rats (Fig. 8). In the remaining 1 rat, no tumor cells but instead necrotic cells were observed to be surrounded by the ablated liver parenchyma. FIGURE 5. Bar chart showing the number of blue dots per square millimeter on Prussian-blue stained specimens in group 1 (rats with ferucarbotran administration followed by hepatic RFA) and group 2 (rats with saline administration followed by hepatic RFA). Group 1 shows more blue dots than group 2 does. The number of blue dots is beyond 200 times higher in the fibrous capsule than in the ablated and nonablated tissues in both groups, but the difference is marginally significant (P = 0.012, Welch t test). Blue dot count in ablated tissues in group 1 is significantly higher than that in group 2 (asterisk, P = 0.005). Also, blue dot count in nonablated tissues in group 1 is higher than that in group 2, but the difference is not statistically significant (P = 0.074).
nonablated area, they were 30.7 ± 24.6 in group 1 and 8.0 ± 5.2 in group 2 (not significant, P = 0.074).
Experiment 2 Ring-shaped, intense accumulations of 59Fe, histologically corresponding to the fibrous capsule, were seen in the ventral parts of liver specimens in 4 RFA-treated rats (Fig. 6A). Thermal-ablated, necrotic liver tissues were present inside the ring-shaped 59Fe accumulations (Fig. 6C, HE-stained specimen). These findings were not observed in the 4 sham-operated rats. Accumulation of 59Fe in the necrotic liver tissues was comparable with nonablated healthy liver parenchyma. The 59Fe accumulation ratios were 0.94 ± 0.11 and 4.4 ± 2.5, respectively, in the thermal-ablated liver tissue area and in the ring-shaped area (P < 0.001).
DISCUSSION Our results demonstrated that the sustained T2*-weighted low signals in the ablated liver tissues were closely related to ferucarbotran administration and subsequent thermal ablation. Histologically, the lowsignal intensity areas corresponded with the areas of RFA-induced coagulative necrosis. Prussian blue-stained specimens revealed 6-fold increase in iron ion deposits in the necrotic areas (ablated areas) in the ferucarbotran-administered rats than in the saline-administered rats (Fig. 5). The 59Fe autoradiography confirmed that ferucarbotran-derived iron particles were weakly deposited in necrotic regions. Collectively, the sustained T2*-weighted hypointensity in the ablated liver tissues strongly suggested persistent accumulation of ferucarbotran-derived iron particles in the RFA-induced coagulative necrotic regions. We also confirmed that sustained T2*-weighted low signals around hepatic tumors were associated with the therapeutic effects of RFA. The ferucarbotran-derived iron particles accumulating in the liver are typically either SPIO crystals or paramagnetic iron ions. We deduced that the presence of SPIO crystals was related to the production of sustained T2*-weighted hypointensity in ablated liver tissues because these crystals are well known to markedly shorten T2* relaxation times in the liver than iron ions.19,20 While we did not directly demonstrate the presence of SPIO crystals in the ablated liver tissues because Prussian blue staining does not react with iron-oxide crystals, our ex vivo data support this notion on the basis of T2* shortening in the ferucarbotran-accumulated and ablated liver tissues (read the second
FIGURE 6. A, Representative autoradiogram of the left liver lobe of an RFA-treated rat. Magnified autoradiographic views of ablated liver tissue (B) and microscopic image of corresponding liver area (C, HE staining) are shown for comparison. A and B, Ring-shaped, intense accumulation of 59Fe is seen beneath the ventral surface of the left liver. Liver tissues inside and outside the ring show comparable radioactivity. The HE-stained specimen shows fibrous capsule surrounding the thermal-ablated, coagulative necrotic liver tissue.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
381
Investigative Radiology • Volume 50, Number 6, June 2015
Nagai et al
FIGURE 7. Representative T2*-weighted MR images (T2*-WIs) of the liver from hepatic tumor-bearing rats. The T2*-WIs acquired after systemic ferucarbotran administration but prior to radiofrequency ablation (RFA, A), post-RFA day 4 (B), and post-RFA day 7 (C). The size of the hepatic tumor is not significantly altered during 7 days. Bar represents 10 mm.
subsection titled T2 and T2* relaxometry of the excised liver samples in Supplemental Digital Content 1, http://links.lww.com/RLI/A192). Other sources that promote T2*-relaxation,10 including hemosiderin and extravascular red blood cells, were seen around the ablated liver tissues comparably in the ferucarbotran-administered rats and saline-administered rats; hence, they were not attributed to the difference in T2*-weighted signals between both rat groups. The reduction in water proton density, which can theoretically drop the signal intensities in T2*-weighted imaging, were also unlikely to cause the T2* hypointensity only seen in the ablated liver tissues in the ferucarbotranadministered rats because vaporization of water molecules strongly depends on tissue heating during RFA, and this was the same in the ferucarbotran-administered and saline-administered rats.
It is well known that transverse relaxation times (T2 and T2*) in the rat liver return to the reference range within 7 days after SPIO administration at comparable doses to this study as Kupffer cells convert SPIO crystals to nonsuperparamagnetic iron ions.21,22 Hence, SPIO crystals rarely remain in the normal liver tissue even after 7 days in the condition under which Kupffer cells function well. Therefore, we deduce that SPIO degradation in the ablated liver tissue may be very different from that in the normal liver. Possibly, the elimination of Kupffer cell function by thermal ablation may be associated with the difference.23 As shown in Figure 6, autoradiography demonstrated the intense 59 Fe accumulation in the fibrous capsule than ablated liver tissues (the inner core). Prussian blue-stained specimen demonstrated the large number
FIGURE 8. T2*-weighted MR images (T2*-WIs) of the liver in another hepatoma-bearing rat 7 days after RFA (A), macroscopic (B), and microscopic views (C, D; magnified views of the left part [arrowhead in B] and the right part [arrow in B], respectively) of HE-stained liver section of the same rat. On the basis of the orientation of the rat, note that a T2*-weighted high-intensity tumor is surrounded by a low-intensity area, which is wide in the right side but narrow in the dorsal (bottom) side (A). Likewise, a necrotic tumor area (T) is surrounded by ablated liver tissues (Ab), which seem also wide in the right side (arrowhead in B) but narrow in the dorsal (bottom) side (arrow) of the specimen. Magnified views clearly demonstrate ablated liver tissues (Ab), or tumor margins, thicker in C than in D. Fibrous capsule is seen beneath the ablated liver tissue. Bars represent 10 mm in A and B as well as 500 μm in D. Figure 8 can be viewed online in color at www.investigativeradiology.com.
382
www.investigativeradiology.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
of iron ions in the fibrous capsule. Macrophages and giant cells trapped these irons. On the other hand, T2*-weighted MR images failed to visualize ring-shaped low signals corresponding to the fibrous capsule, even on high-resolution three-dimensional images (Fig. S1-2 in Supplemental Digital Content 1, which shows high-resolution T2*-weighted images of the excised liver specimens from a rat received ferucarbotran administration and subsequent RFA, http://links.lww.com/RLI/A192). Taken together, 59Fe-labeled irons in the fibrous capsule were predominantly nonsuperparamagnetic, iron ions, which were degradation products of SPIO crystals. In addition, these degraded products were trapped by the phagocytic cells there. Fibrous capsule are gradually formed during 7 days after RFA in response to inflammation against thermal-induced necrotic tissues. During this period, 59Fe, originally engulfed by Kupffer cells in the form of SPIO crystals and accumulated in the sinusoid at the time of RFA, were broken down to iron ions, intralesionally transported and eventually trapped by the fibrous capsule. We speculated that irons were possibly transported to the fibrous capsule via similar mechanisms observed with a low-molecular-weight compound18; however, further study is required to clarify this. Despite the lack of currently available histological techniques for direct visualization of iron-oxide particles in tissue samples, the results of this animal experiment provided insight into SPIO degradation products in ferucarbotran-accumulated and thermal-ablated liver tissues. This contributes to the understanding of diagnostic MR imaging of ablative margins in patients with HCC by using systemic ferucarbotran administration before RFA.9,24 Our data warrant further biochemical studies to investigate the ferucarbotran degradation process in thermalablated liver tissues. Recently, Mori et al25 have proposed a practical technique to predict the RFA treatment efficacy to human HCC by using apparent diffusion coefficient (ADC) mapping. They demonstrated that lower pretreatment ADC levels in HCC nodules were correlated with the high risk for recurrence, attributing the low ADC values to the presence of poorly differentiated HCC cells because such HCC cells are highly invasive and have higher metastatic potential. Although the comparison in the diagnostic accuracy for the prediction of post-RFA recurrence between their technique and ours is beyond the scope of this animal study, animal models created using HCC cells with different invasive and metastatic potential may be required to perform such study in future. In conclusion, we experimentally demonstrated that RFA to ferucarbotran-accumulated healthy liver tissues in rats caused excess iron deposition in ablated liver tissues and produced sustained T2*weighted low signals indicative of ablative margins surrounding hepatic tumors. Our results will assist to clarify the mechanisms related to MR visualization of the ablative margins in patients with HCC after systemic administration of ferucarbotran followed by RFA. ACKNOWLEDGMENTS The authors thank Mr Chun-Jen Chen, Mr Keisuke Uchida, and Mr Yuto Nakahara of Fujifilm RI Pharma Co Ltd, as well as Mr Tomoaki Adachi of Century Medical Inc for their technical support regarding 59Fe autoradiography and RFA experiments. They are also grateful to Mr Akira Nabetani and Mr Atsushi Nozaki of GE Healthcare Japan for constructing the multiarray coil as well as Mr Kenji Yamazaki of the Association for Nuclear Technology in Medicine for constructing the acrylic animal beds used in this study. The authors also thank Dr Shusei Hamamichi for English proofreading. REFERENCES 1. Maluccio M, Covey A. Recent progress in understanding, diagnosing, and treating hepatocellular carcinoma. CA Cancer J Clin. 2012;62:394–399. 2. Gillams AR. The use of radiofrequency in cancer. Br J Cancer. 2005;92: 1825–1829.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
MR Diagnosis of Ablative Margins Around Hepatomas
3. Minami Y, Kudo M. Radiofrequency ablation of hepatocellular carcinoma: a literature review. Int J Hepatol. 2011;2011:104685. 4. Sakakibara M, Ohkawa K, Katayama K, et al. Three-dimensional registration of images obtained before and after radiofrequency ablation of hepatocellular carcinoma to assess treatment adequacy. AJR Am J Roentgenol. 2014;202: W487–W495. 5. Limanond P, Zimmerman P, Raman SS, et al. Interpretation of CT and MRI after radiofrequency ablation of hepatic malignancies. AJR Am J Roentgenol. 2003; 181:1635–1640. 6. Morimoto M, Sugimori K, Shirato K, et al. Treatment of hepatocellular carcinoma with radiofrequency ablation: radiologic-histologic correlation during follow-up periods. Hepatology. 2002;35:1467–1475. 7. Ni Y, Mulier S, Miao Y, et al. A review of the general aspects of radiofrequency ablation. Abdom Imaging. 2005;30:381–400. 8. Lee SH, Lee JM, Kim KW, et al. Dual-energy computed tomography to assess tumor response to hepatic radiofrequency ablation: potential diagnostic value of virtual noncontrast images and iodine maps. Invest Radiol. 2011;46:77–84. 9. Mori K, Fukuda K, Asaoka H, et al. Radiofrequency ablation of the liver: determination of ablative margin at MR imaging with impaired clearance of ferucarbotran—feasibility study. Radiology. 2009;251:557–565. 10. Kakite S, Fujii S, Nakamatsu S, et al. Usefulness of administration of SPIO prior to RF ablation for evaluation of the therapeutic effect: an experimental study using miniature pigs. Eur J Radiol. 2011;78:282–286. 11. Lawaczeck R, Bauer H, Frenzel T, et al. Magnetic iron oxide particles coated with carboxydextran for parenteral administration and liver contrasting. Pre-clinical profile of SH U555A. Acta Radiol. 1997;38:584–597. 12. Miyazaki M, Aoki J, Miyazaki A, et al. Percutaneous radiofrequency ablation of osteoid osteoma using cool-tip electrodes without the cooling system. Jpn J Radiol. 2011;29:138–143. 13. Yamaguchi M, Mitsuda M, Ezawa K, et al. Artifact-reduced simultaneous MRI of multiple rats with liver cancer using PROPELLER. J Magn Reson Imaging. 2013; 38:225–230. 14. Mitsuda M, Yamaguchi M, Nakagami R, et al. Intensity correction method customized for multi-animal abdominal MR imaging with 3 T clinical scanner and multi-array coil. Magn Reson Med Sci. 2013;12:95–103. 15. Guo Y, Zhang Y, Klein R, et al. Irreversible electroporation therapy in the liver: longitudinal efficacy studies in a rat model of hepatocellular carcinoma. Cancer Res. 2010;70:1555–1563. 16. Guo Y, Jin N, Klein R, et al. Gas challenge-blood oxygen level-dependent (GCBOLD) MRI in the rat Novikoff hepatoma model. Magn Reson Imaging. 2012; 30:133–138. 17. Kasper HU, Bangard C, Gossmann A, et al. Pathomorphological changes after radiofrequency ablation in the liver. Pathol Int. 2010;60:149–155. 18. Blanco E, Qian F, Weinberg B, et al. Effect of fibrous capsule formation on doxorubicin distribution in radiofrequency ablated rat livers. J Biomed Mater Res A. 2004;69:398–406. 19. Wang YX, Hussain SM, Krestin GP. Superparamagnetic iron oxide contrast agents: physicochemical characteristics and applications in MR imaging. Eur Radiol. 2001;11:2319–2331. 20. Reimer P, Balzer T. Ferucarbotran (Resovist): a new clinically approved RESspecific contrast agent for contrast-enhanced MRI of the liver: properties, clinical development, and applications. Eur Radiol. 2003;13:1266–1276. 21. Weissleder R, Stark DD, Engelstad BL, et al. Superparamagnetic iron oxide: pharmacokinetics and toxicity. AJR Am J Roentgenol. 1989;152:167–173. 22. Furuta T, Yamaguchi M, Nakagami R, et al. Delayed hepatic signal recovery on ferucarbotran-enhanced magnetic resonance images in a rat model with regional liver irradiation. In: MAGMA. 2014;27:501–508. 23. Furuta T, Yamaguchi M, Nakagami R, et al. Delayed hepatic signal recovery on ferucarbotran-enhanced magnetic resonance images: an experimental study in rat livers with gadolinium chloride-induced Kupffer cell damage. MAGMA. 2013;26:313–324. 24. Koda M, Tokunaga S, Miyoshi K, et al. Ablative margin states by magnetic resonance imaging with ferucarbotran in radiofrequency ablation for hepatocellular carcinoma can predict local tumor progression. J Gastroenterol. 2013;48: 1283–1292. 25. Mori Y, Tamai H, Shingaki N, et al. Signal intensity of small hepatocellular carcinoma on apparent diffusion coefficient mapping and outcome after radiofrequency ablation. In: Hepatol Res. 2015;45:75–87. 26. Bangard C, Stippel DL, Berg F, et al. Conspicuity of zones of ablation after radiofrequency ablation in porcine livers: comparison of an extracellular and an SPIO contrast agent. J Magn Reson Imaging. 2008;28:263–270.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
383
Magnetic Resonance-Based Visualization of Thermal Ablative Margins Around Hepatic Tumors by Means of Systemic Ferucarbotran Administration Before Radiofrequency Ablation Animal Study to Reveal the Connection Between Excess Iron Deposition and T2*-Weighted Hypointensity in Ablative Margins Michiko Nagai, MD,*† Masayuki Yamaguchi, MD, PhD,* Kensaku Mori, MD, PhD,*† Toshihiro Furuta, MD, PhD,*‡ Hiroki Ashino, MSc,§ Hiroyuki Kurosawa, DVM,§ Hiroyuki Kasahara, PhD,§ Manabu Minami, MD, PhD,† and Hirofumi Fujii, MD, PhD* Objective: The objective of this study was to demonstrate experimentally that radiofrequency ablation (RFA) of ferucarbotran-accumulated healthy liver tissues causes excess iron deposition in the ablated liver tissues on postablation days and produces sustained T2*-weighted low signals indicative of ablative margins surrounding hepatic tumors. Materials and Methods: We conducted 3 experiments using 30 rats. In experiment 1, we administered either ferucarbotran (n = 6) or saline (n = 4), acquired T2*weighted images (T2*-WIs) of the liver by using a 3-T magnetic resonance scanner, and subsequently performed RFA of healthy liver lobes. We acquired follow-up T2*WIs up to day 7 and histologically analyzed the liver specimens. In another 4 rats, we performed sham operation, instead of RFA, in ferucarbotran-accumulated liver lobes, followed by the same image acquisition and histological analysis. In experiment 2, we administered 59Fe-labeled ferucarbotran, subsequently performed either RFA (n = 4) or sham operation (n = 4) in the liver, and acquired autoradiograms of the liver specimens on day 7. In experiment 3, we conducted RFA treatment for 8 rats bearing orthotopic hepatic tumors after ferucarbotran administration and monitored tumor growth by using serial T2*-WIs. Results: On days 4 and 7 of the experiment 1, T2*-WIs of 6 rats with systemic ferucarbotran administration and subsequent hepatic RFA showed low-signal regions indicative of ablated liver tissues, whereas high-signal areas were seen in 4 saline-administered rats. Neither high nor low signal areas were detected in 4 shamoperated rats. Histologically, larger amounts of iron were observed in the RFAinduced necrotic liver tissues in the ferucarbotran-administered rats than in the saline-administered-rats. The 59Fe autoradiography of the rats in experiment 2 revealed accumulation of ferucarbotran-derived iron in necrotic liver tissues. Among 6 hepatic tumors grown in 6 rats of the experiment 3, a total of 4 tumors were stable in size, but the other 2 increased markedly on day 7. Retrospectively, T2*-WIs showed the former tumor sites surrounded completely by low-signal areas on day 4. Conclusions: The RFA of ferucarbotran-accumulated healthy liver tissues in the rats caused excess iron deposition in the ablated liver tissues and produced Received for publication September 17, 2014; and accepted for publication, after revision, November 26, 2014. From the *Division of Functional Imaging, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa, Chiba; †Department of Radiology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki; ‡Department of Radiology, Graduate School of Medicine, University of Tokyo, Tokyo; and §Fujifilm RI Pharma Co, Ltd, Sammu, Chiba, Japan. Conflicts of interest and source of funding: Authors Ashino, Kurosawa, and Kasahara are employees of Fujifilm RI Pharma Co, Ltd; supported by Health and Labor Science Research Grant for the Third-Term Comprehensive 10-Year Strategy for Cancer Control and a Grant-in-Aid for Cancer Research (21–5) from the Ministry of Health, Labor, and Welfare of Japan to author Fujii. Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.investigativeradiology.com) Reprints: Masayuki Yamaguchi, MD, PhD, Division of Functional Imaging, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwanoha 6-5-1, Kashiwa, Chiba 277-8577, Japan. E-mail: [email protected]. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0020-9996/15/5006–0376
376
www.investigativeradiology.com
sustained T2*-weighted hypointense regions. Similar hypointense regions surrounding hepatic tumors were indicative of ablative margins. Key Words: hepatic tumor, thermal ablation, ablative margins, magnetic resonance imaging, superparamagnetic iron oxides (Invest Radiol 2015;50: 376–383)
H
epatocellular carcinoma (HCC) is the second leading cause of cancerrelated deaths worldwide.1 Although liver resection is the criterion standard for patients with resectable HCCs, radiofrequency ablation (RFA) is a potentially curative treatment option for selected patients.2 To enhance curability and prevent local recurrence, a 5-mm ablative margin should be secured around the HCC nodule.3,4 Although contrastenhanced computed tomography and magnetic resonance (MR) imaging are currently available for the delineation of ablative margins,5–8 there are some pitfalls related to changes in hepatic tissue perfusion after RFA treatment. First, both ablated tumors and surrounding ablative margins are typically poorly enhanced owing to the lack of blood flow and cannot be clearly contrasted. Second, ablated liver tissues are typically surrounded by well-enhanced congestive or reactive inflammatory areas mimicking residual tumors, and this can sometimes lead to diagnostic errors. Mori et al9 reported a novel technique using ferucarbotranenhanced MR imaging to clearly display the ablative margin, regardless of post-RFA changes in hepatic perfusion. They administered ferucarbotran to patients with HCC several hours before RFA to tag Kupffer cells in the nontumor liver parenchyma by using superparamagnetic iron oxides (SPIOs). Three to 4 days after RFA treatment, they performed T2*-weighted MR imaging without additional contrast administration and visualized low-signal rims surrounding the high-signal tumor areas. Liver parenchyma outside the low-signal rim showed higher signal intensity because the negative enhancement effects by ferucarbotran in nontumor liver tissues had diminished. They also reported that the disruption of low-signal rims increased the risk for tumor recurrence; hence, the low-signal rims were presumably indicative of ablative margins. Because MR image-histology correlation studies regarding the hypointensity rim sign are very limited,10 we carried out RFA in ferucarbotran-accumulated rat liver lobes and performed serial MR imaging of the liver and histological analysis of liver specimens to test whether RFA of ferucarbotran-accumulated healthy liver tissues causes excess iron deposition in ablated liver tissues and produces sustained T2*weighted low signals (experiment 1). To demonstrate clearly the accumulation of ferucarbotran-derived iron particles in the ablated liver tissues in autoradiograms, we used 59Fe-labeled ferucarbotran in selected cases (experiment 2). Furthermore, we conducted RFA treatment in orthotropic hepatic tumor-bearing rats after systemic ferucarbotran administration to confirm whether sustained T2*-weighted low signals surrounding hepatic tumors are indicative of secured ablative margins (experiment 3). Investigative Radiology • Volume 50, Number 6, June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
MR Diagnosis of Ablative Margins Around Hepatomas
MATERIALS AND METHODS Contrast Materials We purchased ferucarbotran (Resovist) from Fujifilm RI Pharma Inc (Tokyo, Japan). In addition, one of the authors synthesized 59Fe-labeled, as well as nonlabeled, ferucarbotran as reported previously.11 In preliminary experiments, we confirmed the equivalent negative enhancement effects of synthesized ferucarbotran to Resovist (Fig. S1-1, Supplemental Digital Content 1, which indicates negative enhancement effect of the synthesized ferucarbotran to the rat liver, http://links.lww.com/RLI/A192).
next inserted the electrode into the subcapsular area and performed RFA with radiofrequency power at 3 W. We omitted coolant circulation through the electrode needle to localize thermal ablated areas.12 Ablation was completed when the temperature of the electrode tip reached 70°C. It took 60 ± 44 seconds (average ± standard deviation [SD]). The electrode was removed, and the abdominal incision was closed. For each sham-operated rat, we inserted the electrode into the liver but did not apply electric currents. After 3 minutes, the electrode was removed.
MR Imaging Animals Our institutional animal experimental committee approved the animal experimental protocol. We purchased 30 female Sprague-Dawley rats (aged 7–8 weeks; typical body weight, 200 g) from Japan SLC Inc (Hamamatsu, Japan) and acclimatized them for a week.
Experiment 1: Hepatic MR Imaging of Rats With Ferucarbotran Administration Before RFA To demonstrate the relationship between the alteration in T2*weighted signal intensities in the ablated liver tissues and ferucarbotran administration, we conducted this experiment using 14 rats divided into 3 groups. We intravenously administered ferucarbotran (Resovist; 20μmol iron/kg body weight [BW]) and subsequently performed hepatic RFA in the first group (n = 6, group 1). This group was subdivided into 2, and duplicated experiments were performed on separate days. We also administered saline (0.3 mL) and then performed RFA in the second group (n = 4, group 2). In the third group (n = 4, group 3), we performed sham operation instead of RFA after 20-μmol iron/kg BW ferucarbotran administration.
Radiofrequency Ablation We performed hepatic RFA using the Cool-tip RF Ablation System (Covidien, Boulder, CO) and an electrode, originally 10 mm in length (Cool-tip RFAblation Single Electrode Kit; Covidien) and deliberately shortened it to 5 mm to produce small ablative tissues in the rat liver. Three to 4 hours after either ferucarbotran or saline administration, anesthesia was induced by inhalation of 5% isoflurane in air and subsequently maintained with 2% isoflurane in air. Then, we made a midline abdominal incision and explored the left lateral lobe of the liver. We
We acquired MR images of the liver before, approximately 10 minutes after either ferucarbotran or saline administration, as well as on 3 or 4 and 7 days after RFA by using a 3.0-T scanner (Signa HDx; GE Healthcare, Milwaukee, WI) and dedicated multireceiver coil comprising 8 coils in the upper and 8 coils in the lower arrays.13 Anesthesia was induced by inhalation of 5% isoflurane in the gas mixture of oxygen and nitrous oxide (1:1) and subsequently maintained with 1% to 2% isoflurane in the gas mixture. We then placed 2 to 4 rats separately on acrylic animal beds in either the prone or supine position; inserted the rats to the coil with their abdominal surfaces close to the coils; and simultaneously acquired transverse T2-weighted (T2WIs), T2*-weighted (T2*WIs), and proton-density-weighted images of the upper abdomen with the parameters as shown in Table 1.
Histological Analysis Immediately after the completion of the MR imaging acquisition on the seventh day after RFA, we excised liver specimens and fixed them with 10% neutral-buffered formalin for a median of 21 days. We did not recognize any tissue damages on hematoxylin and eosin– (HE) as well as Prussian blue–stained liver specimens from the fixation. Then, we sliced the liver tissues into 4- to 5-mm transverse sections and embedded in paraffin wax. We acquired 2-μm sections of the specimens and stained them with HE and Prussian blue. On the Prussian bluestained specimens, we counted the number of blue dots in 5 randomly selected 0.1-mm2 areas in the ablated liver tissues and in the surrounding nonablated liver parenchyma by using a microscope (BZ-9000; Keyence, Osaka, Japan). In addition, we counted the number of blue dots in the fibrous capsule between the ablated and nonablated liver parenchyma in the same manner.
TABLE 1. MR Scan Parameters Parameters Slice orientation Repetition time, ms Effective echo time, ms Excitation flip angle, degrees Echo train length Receiver band width, kHz FOV, mm Frequency encoding steps Phase encoding steps Matrix size on final images Slice thickness, mm Slice gap, mm No. slices No. excitations Acquisition time
GRE T2*-Weighted Image
PROPELLER T2-Weighted Image
FSE Proton Density-Weighted Image
Transaxial 450 11 30 — 31.3 120 256 160 (phase FOV = 0.75) 512 512 2 0.4 15 4 3 min 35 s
Transaxial 4000 86 90 24 50.0 120 320 — 512 512 2 0.4 15 4.5 4 min 8 s
Transaxial 4000 5.8 90 8 31.3 120 128 128 (phase FOV = 0.8) 512 512 2 0.4 15 1 1 min 4 s
FOV indicates field of view; FSE, fast spin echo; GRE, gradient echo; MR, magnetic resonance; PROPELLER, periodically rotated overlapping parallel lines with enhanced reconstruction.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
377
Investigative Radiology • Volume 50, Number 6, June 2015
Nagai et al
MR Image Analysis After minimizing signal decline depending on the distance from the coil surface with proton density-weighted images–based coil sensitivity maps,14 we measured the areas of either high or low signal regions in the left liver lobes on representative T2*-WIs using ImageJ software (version 1.47, available from rsbweb.nih.gov/ij/) and compared them with the corresponding ablated areas on liver specimens. We also placed the maximum possible regions of interest (ROIs) in these high- or lowsignal areas of the liver lobes and paraspinal muscles, then measured signal intensities in each ROI, subsequently dividing signal intensities in these areas by those in the paraspinal muscle to determine relative signal intensities, and finally determined their temporal changes. Because we recognized neither high- nor low-signal areas in the rats before RFA or in the sham-operated rats, we alternatively measured the signal intensities in the left liver lobes placing equivalent spherical ROIs (typically 12 mm2).
Experiment 2: Hepatic Autoradiography of Rats with 59 Fe Ferucarbotran Administration Before RFA To demonstrate the accumulation of ferucarbotran-derived irons in the ablated liver parenchyma, we administered 59Fe-labeled ferucarbotran to 8 rats at a dose of 40 kBq/head (total iron load adjusted to 20-μmol iron/kg BW) and performed either RFA (n = 4) or sham operation (n = 4) in the left liver lobe of each rat 3 to 4 hours later as mentioned previously. After 7 days, we carried out perfusion fixation with 20% (w/v) sucrose dissolved in phosphate-buffered saline and then excised left liver lobes to obtain 5-mm transverse sections. After embedding and freezing them in optimum cutting temperature compound (Tissue-Tek O.C.T. compound 4583; Sakura Finetek USA, Inc,
Torrance, CA), we sliced the specimens into 20-μm sections by using a cryostat (Leica CM3050 S; Leica Microsystems GmbH, Wetzlar, Germany), subsequently placed the specimens on imaging plates (BAS-III IP; Fujifilm Co, Ltd, Tokyo, Japan) for 6 to 7 days, and finally acquired autoradiograms using an image reader (FLA-7000, Fujifilm).
Image Analysis Subsequently, we manually placed an ROI (2 mm2) in each ablated liver tissue and 3 ROIs of the same size in nonablated liver tissues on representative autoradiograms of the left liver lobes and assessed the signal strength in these ROIs by using densitometric analysis software (Multi Gauge version 3.0; Fujifilm). We also automatically counted signal strength in ring-shaped areas surrounding the ablated liver tissues (histologically corresponding to the fibrous capsule). Ratios of 59Fe accumulation in the ablated liver tissue areas and in the fibrous capsule areas were calculated by using the following equation: 59 Fe accumulation ratio = signal strength in either ablated area or fibrous capsule area/average of signal strengths in nonablated liver tissue areas
Experiment 3: Serial MR Imaging Examination of Hepatic Tumor-Bearing Rat Liver With Systemic Ferucarbotran Administration Followed by RFA Treatment To demonstrate the relationship between the treatment efficacy of RFA to hepatic tumors and sustained T2*-weighted low signals surrounding the tumor, we performed experiments using an orthotopic hepatic tumor model. We inoculated N1-S1 hepatoma cells (CRL-1604; American Type Culture Collection, Manassas, VA; 4 106 cells per
FIGURE 1. Representative series of T2*-weighted MR images of the liver from rats receiving ferucarbotran administration as well as subsequent RFA (A–D, group 1), saline administration and subsequent RFA (E–H, group 2), or ferucarbotran administration and subsequent sham operation (I–L, group 3). The first and second columns show the images acquired before (A, E, and I) and after (B, F, and J) either ferucarbotran or saline administration. The third and fourth columns show the images acquired 3 to 4 days (C, G, and K) and 7 days (D, H, and L) after RFA or sham operation. Note the well-demarcated low-signal areas (arrows in C and D) and ill-demarcated slightly high-signal areas (arrowheads in G and H) in the ventral parts of the left liver lobes indicative of ablated liver tissues. The former low signal areas contain small high signals indicative of electrode needle paths. Bar represents 10 mm.
378
www.investigativeradiology.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
MR Diagnosis of Ablative Margins Around Hepatomas
FIGURE 2. Graphs showing the changes in relative T2*-weighted signal intensities of the ablated, sham-operated and nonablated areas of the liver as a function of time. A, After ferucarbotran administration, relative signal intensities in the liver decrease significantly. In the ablated area (solid line), relative signal intensities remain low at days 4 and 7 after RFA, whereas those in the nonablated area (dotted line) are restored toward the baseline levels. B, After normal saline (NS) administration, hepatic signal intensities show no changes when compared with the baseline. On days 4 and 7 after RFA, relative signal intensities in the ablated area (solid line) are slightly higher than those in the nonablated area (dotted line). C, In sham-operated rats, hepatic signal intensities are significantly reduced after ferucarbotran administration and signal intensities in the sham-operated (solid line) and nonoperated areas (dotted line) are comparably restored to the baseline levels during the experimental period. Average values in relative signal intensities are indicated in arbitrary units. Error bars represent standard deviation. Asterisks indicate statistical significance (P < 0.001, Tukey test). AU indicates arbitrary unit; NS, normal saline.
head) into the left lateral lobes of the liver in 8 rats by following the technique described in the literature.13,15,16 Seven days later, we acquired ferucarbotran-enhanced T2*-WIs of the liver (20-μmol iron /kg BW) and subsequently performed RFA in the hepatic tumor lesions by following the same technique as mentioned previously except for the length of electrode of 5 to 10 mm depending on tumor size. In addition, we carried out multi-RFA sessions for larger tumors. On postablation days 4 and 7, we acquired T2*-WIs of the liver and evaluated temporal changes in tumor size by measuring their maximum perpendicular diameters (MPDs). One of the authors retrospectively analyzed the relationship between the MR imaging appearance and tumor growth. The duplicate experiments were performed on separate days using 4 rats each.
Statistical Tests We assessed the differences in relative signal intensities between ablated and nonablated liver tissues at each time point by using the Tukey test provided by commercially available software (SPSS version 20; IBM Inc, Tokyo, Japan). We also evaluated the differences in the number of blue dots in the ablated, nonablated liver tissues, as well as fibrous capsule between groups 1 and 2 by using the Welch t test. Furthermore, we assessed the differences in 59Fe accumulation ratio between ablated liver tissue and fibrous capsule areas using the same test. Finally, we assessed the difference between high-signal and low-signal
areas indicative of ablated tumor and tumor margins, respectively, on T2*-weighted MR images on day 7 using paired t test. P < 0.01 was considered to be statistically significant.
RESULTS Laparotomy and RFA procedures were completed in all rats without serious complications.
Experiment 1 In group 1 (Fig. 1, A–D), signal intensities in healthy liver were similar to those of the paraspinal muscle on baseline scans, but they decreased significantly after ferucarbotran administration and returned to baseline levels toward day 7. Four days after RFA, approximately 5 mm in diameter, well-demarcated, spheroid hypointense areas appeared in the ventral parts of the left liver lobes where RFA procedures were performed (arrow in Fig. 1C). Seven days after RFA, these areas were still visible (arrow in Fig. 1D). In contrast, poorly demarcated, slightly high-signal areas were seen in the ventral parts of the left liver lobes on days 4 and 7 in group 2 (arrowheads in Fig. 1, G and H). In other liver lobes, signal intensities were comparable with those of the paraspinal muscle throughout the experimental period. In group 3 (Fig. 1, I–L), no abnormal signals were seen in sham-operated liver lobes. Figure 2 summarizes the temporal changes in relative signal
FIGURE 3. Representative specimen of the left liver lobe in a rat with ferucarbotran administration followed by RFA (original magnification 12.5 in a, and 40 in b). Spheroid area of coagulative necrosis is seen in the subcapsular region where RFA was performed (ablated area; Ab). The round defect in the center is the electrode needle path. A thin fibrous capsule (arrow heads) is present between the ablated area and nonablated liver parenchyma.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
379
Investigative Radiology • Volume 50, Number 6, June 2015
Nagai et al
intensities of the ablated and nonablated liver tissues. T2-weighted MR imaging results are summarized in digital content (Fig. S2-1, Supplemental Digital Content 2, which shows T2-weighted MR image series of the liver from rats receiving ferucarbotran administration and subsequent RFA, http://links.lww.com/RLI/A193). Histologically, 12.1 ± 2.4 mm2, well-demarcated, spheroid coagulation necrotic regions were produced in the ventral parts of the left lateral lobes by RFA in group 1 (Fig. 3A). They were well correlated with 14.0 ± 0.6 mm2 and low-signal regions in the corresponding liver lobes seen on T2*-weighted MR images. In group 2, the necrotic regions were 12.4 ± 1.2 mm2 and correlated with the 17.7 ± 7.4 mm2 as well as high-signal areas on T2*-weighted images. Surrounding these coagulative necrotic regions were thin fibrogranulation tissues, known as fibrous capsules (arrowheads in Fig. 3).17,18 These findings
were common in rats in groups 1 and 2. In the sham-operated rats (group 3), approximately 3-mm pale tissues were recognized under the liver capsule. On Prussian blue-stained specimens, blue dots indicative of iron ions were scattered in the dilated sinusoids in RFA-induced necrotic tissues (Fig. 4, A and B). In the fibrous capsule, numerous blue dots were seen except in the liver surface region in groups 1 and 2 (Fig. 4, C and D). In nonablated liver tissues, blue dots were sporadically observed along the sinusoids. The number of blue dots in the ablated area (RFA-induced necrotic tissue) was significantly larger in the ferucarbotran-administered group 1 rats than in the saline-administered group 2 rats (31.3 ± 14.0 vs 5.0 ± 2.0 /mm2; Fig. 5). In the fibrous capsule, the numbers were 579.7 ± 190.2 in group 1 and 264.5 ± 120.5 in group 2. The difference was marginally significant (P = 0.012). In the
FIGURE 4. Representative HE-stained and Prussian blue-stained specimens from the same liver specimens shown in Figure 3. In the ablated area (A, B; original magnification, 400), necrotic and shrunken hepatocytes with small, homogeneous, slightly basophilic nuclei and homogeneous eosinophilic cytoplasm are arranged like the hepatic cord (A). In the slightly dilated sinusoids, several blue dots are present adjacent to Kupffer cells' nuclei (arrow heads, B). C, The fibrous capsule shows a reactive inflammatory change containing numerous inflammatory cells, fibroblasts, and newly developed capillaries. Multinuclear giant cells are occasionally observed, as reported previously.26 D, Prussian blue-stained specimen shows a large number of blue dots in the fibrous capsule, representing iron ion particles. E, Nonablated liver parenchyma shows normal architecture composed of regular hepatic cords and sinusoids. F, Blue dots are scattered in the sinusoid, indicating iron deposition in Kupffer cells. Figure 4 can be viewed online in color at www.investigativeradiology.com.
380
www.investigativeradiology.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
MR Diagnosis of Ablative Margins Around Hepatomas
Experiment 3 On T2*-WIs, 8 high-signal nodules were detected in the left liver in 8 rats on 7 days after tumor cell inoculation, but 2 high-signal regions in 2 rats were excluded from further analysis because of their small sizes (5.5 and 5.6 mm2). The other 6 nodules showed MPDs of 35.4 ± 14.9 mm2 (mean ± SD), ranging from 22.2 to 63.6 mm2. T2*WIs on postablation day 4 demonstrated 4 nodules (MPDs, 34.6 ± 7.7 mm2) surrounded by deep low-signal rims (Fig. 7) and the other 2 nodules (36 and 109 mm2) without the low-signal rim. On postablation day 7, the former 4 nodules were stable in size (MPDs: 25.9 ± 6.5 mm2; growth: 90% ± 19%), whereas the latter 2 lesions had increased in size (85.4 and 149.2 mm2; 235% and 268%). The former 4 lesions were still surrounded by low-signal areas in which MPDs were 74.1 ± 32.6 mm2 larger than high-signal areas indicative of ablated tumor lesions (25.9 ± 6.5 mm2 in MPD). The difference in these areas was marginally significant (P =0.050 in paired t test). In histological specimens, 51.2 ± 11.6 -mm2 ablated liver parenchyma surrounded ablated tumor cell areas (15.4 ± 9.0 mm2) in 3 rats (Fig. 8). In the remaining 1 rat, no tumor cells but instead necrotic cells were observed to be surrounded by the ablated liver parenchyma. FIGURE 5. Bar chart showing the number of blue dots per square millimeter on Prussian-blue stained specimens in group 1 (rats with ferucarbotran administration followed by hepatic RFA) and group 2 (rats with saline administration followed by hepatic RFA). Group 1 shows more blue dots than group 2 does. The number of blue dots is beyond 200 times higher in the fibrous capsule than in the ablated and nonablated tissues in both groups, but the difference is marginally significant (P = 0.012, Welch t test). Blue dot count in ablated tissues in group 1 is significantly higher than that in group 2 (asterisk, P = 0.005). Also, blue dot count in nonablated tissues in group 1 is higher than that in group 2, but the difference is not statistically significant (P = 0.074).
nonablated area, they were 30.7 ± 24.6 in group 1 and 8.0 ± 5.2 in group 2 (not significant, P = 0.074).
Experiment 2 Ring-shaped, intense accumulations of 59Fe, histologically corresponding to the fibrous capsule, were seen in the ventral parts of liver specimens in 4 RFA-treated rats (Fig. 6A). Thermal-ablated, necrotic liver tissues were present inside the ring-shaped 59Fe accumulations (Fig. 6C, HE-stained specimen). These findings were not observed in the 4 sham-operated rats. Accumulation of 59Fe in the necrotic liver tissues was comparable with nonablated healthy liver parenchyma. The 59Fe accumulation ratios were 0.94 ± 0.11 and 4.4 ± 2.5, respectively, in the thermal-ablated liver tissue area and in the ring-shaped area (P < 0.001).
DISCUSSION Our results demonstrated that the sustained T2*-weighted low signals in the ablated liver tissues were closely related to ferucarbotran administration and subsequent thermal ablation. Histologically, the lowsignal intensity areas corresponded with the areas of RFA-induced coagulative necrosis. Prussian blue-stained specimens revealed 6-fold increase in iron ion deposits in the necrotic areas (ablated areas) in the ferucarbotran-administered rats than in the saline-administered rats (Fig. 5). The 59Fe autoradiography confirmed that ferucarbotran-derived iron particles were weakly deposited in necrotic regions. Collectively, the sustained T2*-weighted hypointensity in the ablated liver tissues strongly suggested persistent accumulation of ferucarbotran-derived iron particles in the RFA-induced coagulative necrotic regions. We also confirmed that sustained T2*-weighted low signals around hepatic tumors were associated with the therapeutic effects of RFA. The ferucarbotran-derived iron particles accumulating in the liver are typically either SPIO crystals or paramagnetic iron ions. We deduced that the presence of SPIO crystals was related to the production of sustained T2*-weighted hypointensity in ablated liver tissues because these crystals are well known to markedly shorten T2* relaxation times in the liver than iron ions.19,20 While we did not directly demonstrate the presence of SPIO crystals in the ablated liver tissues because Prussian blue staining does not react with iron-oxide crystals, our ex vivo data support this notion on the basis of T2* shortening in the ferucarbotran-accumulated and ablated liver tissues (read the second
FIGURE 6. A, Representative autoradiogram of the left liver lobe of an RFA-treated rat. Magnified autoradiographic views of ablated liver tissue (B) and microscopic image of corresponding liver area (C, HE staining) are shown for comparison. A and B, Ring-shaped, intense accumulation of 59Fe is seen beneath the ventral surface of the left liver. Liver tissues inside and outside the ring show comparable radioactivity. The HE-stained specimen shows fibrous capsule surrounding the thermal-ablated, coagulative necrotic liver tissue.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
381
Investigative Radiology • Volume 50, Number 6, June 2015
Nagai et al
FIGURE 7. Representative T2*-weighted MR images (T2*-WIs) of the liver from hepatic tumor-bearing rats. The T2*-WIs acquired after systemic ferucarbotran administration but prior to radiofrequency ablation (RFA, A), post-RFA day 4 (B), and post-RFA day 7 (C). The size of the hepatic tumor is not significantly altered during 7 days. Bar represents 10 mm.
subsection titled T2 and T2* relaxometry of the excised liver samples in Supplemental Digital Content 1, http://links.lww.com/RLI/A192). Other sources that promote T2*-relaxation,10 including hemosiderin and extravascular red blood cells, were seen around the ablated liver tissues comparably in the ferucarbotran-administered rats and saline-administered rats; hence, they were not attributed to the difference in T2*-weighted signals between both rat groups. The reduction in water proton density, which can theoretically drop the signal intensities in T2*-weighted imaging, were also unlikely to cause the T2* hypointensity only seen in the ablated liver tissues in the ferucarbotranadministered rats because vaporization of water molecules strongly depends on tissue heating during RFA, and this was the same in the ferucarbotran-administered and saline-administered rats.
It is well known that transverse relaxation times (T2 and T2*) in the rat liver return to the reference range within 7 days after SPIO administration at comparable doses to this study as Kupffer cells convert SPIO crystals to nonsuperparamagnetic iron ions.21,22 Hence, SPIO crystals rarely remain in the normal liver tissue even after 7 days in the condition under which Kupffer cells function well. Therefore, we deduce that SPIO degradation in the ablated liver tissue may be very different from that in the normal liver. Possibly, the elimination of Kupffer cell function by thermal ablation may be associated with the difference.23 As shown in Figure 6, autoradiography demonstrated the intense 59 Fe accumulation in the fibrous capsule than ablated liver tissues (the inner core). Prussian blue-stained specimen demonstrated the large number
FIGURE 8. T2*-weighted MR images (T2*-WIs) of the liver in another hepatoma-bearing rat 7 days after RFA (A), macroscopic (B), and microscopic views (C, D; magnified views of the left part [arrowhead in B] and the right part [arrow in B], respectively) of HE-stained liver section of the same rat. On the basis of the orientation of the rat, note that a T2*-weighted high-intensity tumor is surrounded by a low-intensity area, which is wide in the right side but narrow in the dorsal (bottom) side (A). Likewise, a necrotic tumor area (T) is surrounded by ablated liver tissues (Ab), which seem also wide in the right side (arrowhead in B) but narrow in the dorsal (bottom) side (arrow) of the specimen. Magnified views clearly demonstrate ablated liver tissues (Ab), or tumor margins, thicker in C than in D. Fibrous capsule is seen beneath the ablated liver tissue. Bars represent 10 mm in A and B as well as 500 μm in D. Figure 8 can be viewed online in color at www.investigativeradiology.com.
382
www.investigativeradiology.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 6, June 2015
of iron ions in the fibrous capsule. Macrophages and giant cells trapped these irons. On the other hand, T2*-weighted MR images failed to visualize ring-shaped low signals corresponding to the fibrous capsule, even on high-resolution three-dimensional images (Fig. S1-2 in Supplemental Digital Content 1, which shows high-resolution T2*-weighted images of the excised liver specimens from a rat received ferucarbotran administration and subsequent RFA, http://links.lww.com/RLI/A192). Taken together, 59Fe-labeled irons in the fibrous capsule were predominantly nonsuperparamagnetic, iron ions, which were degradation products of SPIO crystals. In addition, these degraded products were trapped by the phagocytic cells there. Fibrous capsule are gradually formed during 7 days after RFA in response to inflammation against thermal-induced necrotic tissues. During this period, 59Fe, originally engulfed by Kupffer cells in the form of SPIO crystals and accumulated in the sinusoid at the time of RFA, were broken down to iron ions, intralesionally transported and eventually trapped by the fibrous capsule. We speculated that irons were possibly transported to the fibrous capsule via similar mechanisms observed with a low-molecular-weight compound18; however, further study is required to clarify this. Despite the lack of currently available histological techniques for direct visualization of iron-oxide particles in tissue samples, the results of this animal experiment provided insight into SPIO degradation products in ferucarbotran-accumulated and thermal-ablated liver tissues. This contributes to the understanding of diagnostic MR imaging of ablative margins in patients with HCC by using systemic ferucarbotran administration before RFA.9,24 Our data warrant further biochemical studies to investigate the ferucarbotran degradation process in thermalablated liver tissues. Recently, Mori et al25 have proposed a practical technique to predict the RFA treatment efficacy to human HCC by using apparent diffusion coefficient (ADC) mapping. They demonstrated that lower pretreatment ADC levels in HCC nodules were correlated with the high risk for recurrence, attributing the low ADC values to the presence of poorly differentiated HCC cells because such HCC cells are highly invasive and have higher metastatic potential. Although the comparison in the diagnostic accuracy for the prediction of post-RFA recurrence between their technique and ours is beyond the scope of this animal study, animal models created using HCC cells with different invasive and metastatic potential may be required to perform such study in future. In conclusion, we experimentally demonstrated that RFA to ferucarbotran-accumulated healthy liver tissues in rats caused excess iron deposition in ablated liver tissues and produced sustained T2*weighted low signals indicative of ablative margins surrounding hepatic tumors. Our results will assist to clarify the mechanisms related to MR visualization of the ablative margins in patients with HCC after systemic administration of ferucarbotran followed by RFA. ACKNOWLEDGMENTS The authors thank Mr Chun-Jen Chen, Mr Keisuke Uchida, and Mr Yuto Nakahara of Fujifilm RI Pharma Co Ltd, as well as Mr Tomoaki Adachi of Century Medical Inc for their technical support regarding 59Fe autoradiography and RFA experiments. They are also grateful to Mr Akira Nabetani and Mr Atsushi Nozaki of GE Healthcare Japan for constructing the multiarray coil as well as Mr Kenji Yamazaki of the Association for Nuclear Technology in Medicine for constructing the acrylic animal beds used in this study. The authors also thank Dr Shusei Hamamichi for English proofreading. REFERENCES 1. Maluccio M, Covey A. Recent progress in understanding, diagnosing, and treating hepatocellular carcinoma. CA Cancer J Clin. 2012;62:394–399. 2. Gillams AR. The use of radiofrequency in cancer. Br J Cancer. 2005;92: 1825–1829.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
MR Diagnosis of Ablative Margins Around Hepatomas
3. Minami Y, Kudo M. Radiofrequency ablation of hepatocellular carcinoma: a literature review. Int J Hepatol. 2011;2011:104685. 4. Sakakibara M, Ohkawa K, Katayama K, et al. Three-dimensional registration of images obtained before and after radiofrequency ablation of hepatocellular carcinoma to assess treatment adequacy. AJR Am J Roentgenol. 2014;202: W487–W495. 5. Limanond P, Zimmerman P, Raman SS, et al. Interpretation of CT and MRI after radiofrequency ablation of hepatic malignancies. AJR Am J Roentgenol. 2003; 181:1635–1640. 6. Morimoto M, Sugimori K, Shirato K, et al. Treatment of hepatocellular carcinoma with radiofrequency ablation: radiologic-histologic correlation during follow-up periods. Hepatology. 2002;35:1467–1475. 7. Ni Y, Mulier S, Miao Y, et al. A review of the general aspects of radiofrequency ablation. Abdom Imaging. 2005;30:381–400. 8. Lee SH, Lee JM, Kim KW, et al. Dual-energy computed tomography to assess tumor response to hepatic radiofrequency ablation: potential diagnostic value of virtual noncontrast images and iodine maps. Invest Radiol. 2011;46:77–84. 9. Mori K, Fukuda K, Asaoka H, et al. Radiofrequency ablation of the liver: determination of ablative margin at MR imaging with impaired clearance of ferucarbotran—feasibility study. Radiology. 2009;251:557–565. 10. Kakite S, Fujii S, Nakamatsu S, et al. Usefulness of administration of SPIO prior to RF ablation for evaluation of the therapeutic effect: an experimental study using miniature pigs. Eur J Radiol. 2011;78:282–286. 11. Lawaczeck R, Bauer H, Frenzel T, et al. Magnetic iron oxide particles coated with carboxydextran for parenteral administration and liver contrasting. Pre-clinical profile of SH U555A. Acta Radiol. 1997;38:584–597. 12. Miyazaki M, Aoki J, Miyazaki A, et al. Percutaneous radiofrequency ablation of osteoid osteoma using cool-tip electrodes without the cooling system. Jpn J Radiol. 2011;29:138–143. 13. Yamaguchi M, Mitsuda M, Ezawa K, et al. Artifact-reduced simultaneous MRI of multiple rats with liver cancer using PROPELLER. J Magn Reson Imaging. 2013; 38:225–230. 14. Mitsuda M, Yamaguchi M, Nakagami R, et al. Intensity correction method customized for multi-animal abdominal MR imaging with 3 T clinical scanner and multi-array coil. Magn Reson Med Sci. 2013;12:95–103. 15. Guo Y, Zhang Y, Klein R, et al. Irreversible electroporation therapy in the liver: longitudinal efficacy studies in a rat model of hepatocellular carcinoma. Cancer Res. 2010;70:1555–1563. 16. Guo Y, Jin N, Klein R, et al. Gas challenge-blood oxygen level-dependent (GCBOLD) MRI in the rat Novikoff hepatoma model. Magn Reson Imaging. 2012; 30:133–138. 17. Kasper HU, Bangard C, Gossmann A, et al. Pathomorphological changes after radiofrequency ablation in the liver. Pathol Int. 2010;60:149–155. 18. Blanco E, Qian F, Weinberg B, et al. Effect of fibrous capsule formation on doxorubicin distribution in radiofrequency ablated rat livers. J Biomed Mater Res A. 2004;69:398–406. 19. Wang YX, Hussain SM, Krestin GP. Superparamagnetic iron oxide contrast agents: physicochemical characteristics and applications in MR imaging. Eur Radiol. 2001;11:2319–2331. 20. Reimer P, Balzer T. Ferucarbotran (Resovist): a new clinically approved RESspecific contrast agent for contrast-enhanced MRI of the liver: properties, clinical development, and applications. Eur Radiol. 2003;13:1266–1276. 21. Weissleder R, Stark DD, Engelstad BL, et al. Superparamagnetic iron oxide: pharmacokinetics and toxicity. AJR Am J Roentgenol. 1989;152:167–173. 22. Furuta T, Yamaguchi M, Nakagami R, et al. Delayed hepatic signal recovery on ferucarbotran-enhanced magnetic resonance images in a rat model with regional liver irradiation. In: MAGMA. 2014;27:501–508. 23. Furuta T, Yamaguchi M, Nakagami R, et al. Delayed hepatic signal recovery on ferucarbotran-enhanced magnetic resonance images: an experimental study in rat livers with gadolinium chloride-induced Kupffer cell damage. MAGMA. 2013;26:313–324. 24. Koda M, Tokunaga S, Miyoshi K, et al. Ablative margin states by magnetic resonance imaging with ferucarbotran in radiofrequency ablation for hepatocellular carcinoma can predict local tumor progression. J Gastroenterol. 2013;48: 1283–1292. 25. Mori Y, Tamai H, Shingaki N, et al. Signal intensity of small hepatocellular carcinoma on apparent diffusion coefficient mapping and outcome after radiofrequency ablation. In: Hepatol Res. 2015;45:75–87. 26. Bangard C, Stippel DL, Berg F, et al. Conspicuity of zones of ablation after radiofrequency ablation in porcine livers: comparison of an extracellular and an SPIO contrast agent. J Magn Reson Imaging. 2008;28:263–270.
www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
383
