| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Interventional Radiology |
1 Department of Radiology, Case Western Reserve University, University Hospitals
of Cleveland, 11100 Euclid Ave., Cleveland, OH 44106.
2 Present address: Department of Radiology,
Charité-Universitätsmedizin Berlin, Campus Benjamin Franklin,
Hindenburgdamm 30, Berlin 12203, Germany. Address
3 Department of Pathology, Case Western Reserve University, University Hospitals
of Cleveland, Cleveland, OH 44106.
4 Department of Radiology, Johns Hopkins School of Medicine, The Johns Hopkins
Hospital, 601 N Caroline St., Rm. 4210, Baltimore, MD 21287.
Received October 23, 2003; accepted after revision April 20, 2004.
Address correspondence to F. K. Wacker
(wackerfrank{at}web.de).
Supported in part by Siemens Medical Solutions and by National Cancer
Institute grants R33CA88144 and R01CA81431.
OBJECTIVE. The purpose of this study was to assess the feasibility of MRI to guide and monitor radiofrequency ablation of normal pulmonary tissue in a rabbit model.
MATERIALS AND METHODS. Percutaneous puncture and lung radiofrequency ablation were performed in six New Zealand white rabbits under MRI control using a 0.2-T open MRI scanner. Technical feasibility and complication detection were evaluated. The ablation zone appearance and size were assessed using MRI, CT, and gross pathology. Interclass correlation coefficients (ICCs) of the maximum short-axis diameters of the lesions on gross pathology and the corresponding diameters as measured on each MRI pulse sequence and on CT scans were calculated.
RESULTS. MRI guidance of percutaneous puncture and radiofrequency ablation of pulmonary tissue is feasible. A pneumothorax was detected and treated using MRI. In the specimen, the mean coagulation necrosis diameter was 9.8 mm. The T1-weighted spoiled gradient-echo fast low-angle shot images showed the highest ICC (0.81) for the thermal lesion diameter.
CONCLUSION. Our results indicate that MRI guidance is feasible and useful for radiofrequency ablation of normal pulmonary tissue.
According to the American Cancer Society, lung cancer remained the primary cause of cancer deaths in the United States in 2002, accounting for 28% of all American cancer deaths. To date, there has been limited progress in the management of this devastating disease [1]. Most patients with lung malignancies are not good surgical candidates because of poor pulmonary function or severe medical comorbidity [2]. Systemic chemotherapy and radiation therapy have not greatly improved the outcome in these patients whose overall 10-year relative survival rate is as low as 7%. Consequently, research continues for better therapeutic regimens [1].
Since 1990, minimally invasive techniques have been introduced as potential alternatives to surgical resection. The imaging-guided interventional and video-assisted thoracoscopic approaches have become alternatives to open thoracic surgery. If endoscopic, bronchoscopic, or percutaneous treatments can provide similar disease-free survival rates with lower perioperative mortality and morbidity rates and lower cost, they could become the preferred therapies [3]. Imaging-guided thermal ablation techniques represent relatively recent additions to these new therapeutic approaches. Laser and, more commonly, radiofrequency ablation have been used for lung tumors [2, 4–9]. These purely local treatments minimize damage to the lung parenchyma without significant adverse systemic effects to a patient's general health; this advantage is important, especially in patients with limited pulmonary reserve. It is generally accepted that MRI is capable of measuring parameters such as T1-time (longitudinal relaxation time), phase shift, and diffusion that are directly dependent on the temperature of the respective tissue and that CT and sonography do not allow determination of tissue temperature [10–21], but MRI guidance has not been well evaluated for lung ablation. The purpose of our study was to assess the feasibility of MRI at 0.2 T to guide and monitor radiofrequency ablation of pulmonary tissue in a rabbit model.
Materials and Methods
Following an institutional animal care and use committee–approved protocol, we performed radiofrequency ablation in six New Zealand white rabbits (2.8–3.5 kg), using a custom-made, 18-gauge, internally cooled, nonferromagnetic titanium electrode with a 1-cm active tip. A conventional monopolar radiofrequency technique was implemented.
Induction of anesthesia was achieved via an intramuscular injection of 2 mL of a cocktail (0.6 mL/kg)— a combination of ketamine (0.26 mL/kg), xylazine (0.26 mL/kg), and acepromazine (0.08 mL/kg). Anesthesia was maintained with intramuscular injections every 20–40 min that alternated between 0.5 mL of ketamine (0.15 mL/kg) and 1 mL of the cocktail (0.3 mL/kg) mentioned previously.
Animals were positioned prone on the MRI scanner after placement of an 8 x 12 cm wire mesh grounding pad coated with conductive gel (Aquasonic 100, Parker Laboratories) on the rabbit's shaved abdomen. Puncture and thermal ablation were performed under near real-time MRI control in an interventional MRI suite using a 0.2-T open C-arm MRI (Magnetom Open, Siemens Medical Solutions), which allowed the radiologist to operate the scanner and simultaneously view images at the scanner side through an in-room, high-resolution, radiofrequency-shielded liquid crystal monitor controlled by an MRI-compatible mouse and foot pedal. A semiopen head coil was used to allow room for radiofrequency electrode manipulation. For percutaneous insertion of the electrode, rapid gradient-echo sequences were used (true fast imaging with steady-state free precession [true FISP]: 12.5/5.9; flip angle, 90°; temporal resolution, 2 frames per second; fast low-angle shot (FLASH): 108/11.8; flip angle, 25°; slice thickness, 6 mm; temporal resolution, 3 slices in 18 sec). Six ablations were performed.
After acquiring baseline images of the whole lung, we selected an ablation site with enough lung parenchyma to create a 10-mm-diameter intraparenchymal lesion that was at least 10 mm distant from any major vessels. After the electrode was positioned, the radiofrequency ablation was performed. The electrode was simultaneously cooled to 10–15°C by circulating ice water through its internal channels using a pump. The ablation time was set to 5 min to achieve a thermal lesion measuring approximately 10 mm in diameter. In two ablations, the impedance-based self-regulating feedback system implemented on the 200-W radiofrequency generator (RFG-3C, Radionics) stopped the ablation after 189 and 235 sec because of an impedance increase. Probe-tip temperature, tissue impedance, and radiofrequency current were recorded from the generator display at the baseline and at 1-min intervals throughout the entire ablation cycle.
We obtained immediate postablation MR images (reversed FISP [PSIF] sequence: 15.4/7.4; flip angle, 45°; slice thickness, 6 mm; STIR: 2,700/48; inversion time, 110 msec; slice thickness, 6; true FISP: 12.5/59; flip angle, 90°; slice thickness, 6; and T1-weighted spoiled gradient-echo FLASH: 108/11.8; flip angle, 25°; slice thickness, 6 mm). Immediately after MRI was performed, all animals were transported to the CT suite under anesthesia and underwent CT on an MX 8000 IDT scanner (Philips Medical Systems). Scanning was performed in a craniocaudal direction covering the whole lung with a 16 x 0.75 mm collimation and a pitch of 0.4. Images were reconstructed in 1-mm slices with a 0.5-mm increment. The animals were in a prone position, and no contrast medium was administered.
Four animals were euthanized immediately after CT, 2, 3, 5, and 6 hr after ablation with IV administration of 2–3 mL of pentobarbital sodium (Euthasol [390 mg/mL], Diamond Animal Health). Two animals were euthanized after a waiting period of 48 hr to evaluate whether a better defined thermal lesion could be seen; all lungs were then harvested and immediately sliced centrally while attempting to cut exactly through the plane of the radiofrequency electrode track. Tissue faces were then digitally photographed and underwent formalin fixation. The maximum thermal lesion diameter perpendicular to the radiofrequency electrode track as measured on gross pathology specimens before formalin fixation was compared with the diameter measured on each of the postablation MR images and CT scans.
The ability of MRI to detect short-term complications such as bleeding and pneumothorax was evaluated on the basis of the presence or absence of intra- or immediately postprocedural CT findings compared with MRI findings.
To evaluate the performance of the various implemented pulse sequences to predict the actual thermal lesion size, we calculated the unbiased interclass correlation coefficients (ICCs) of the maximum short-axis diameters of the lesions on gross pathology and the corresponding diameters as measured on each pulse sequence on the MR images and CT scans, respectively. The ICC [22] is an index of absolute agreement of continuous measures of the same set of objects based on two or more approaches. It ranges from infinity to plus one. ICC equaling one corresponds to perfect agreement. Positive values are meaningfully interpretable as measures of the level of agreement. When ICC is positive, 1 – ICC may be interpreted as the proportion of variance in the sample due to disagreement. The null hypothesis is ICC equals zero or less. Therefore, we only concluded the existence of any level of agreement when the ICC was statistically significantly greater than zero. In this study, we used the small sample bias correction formula to calculate ICC [22].
Results
Successful MRI-guided radiofrequency electrode positioning into the desired portion of the lung was achieved in all procedures. The radiofrequency electrodes could be visualized because of their dark susceptibility artifact compared with the slightly higher signal intensity of the surrounding lung tissue on both the FLASH and the true FISP sequences. The image quality was sufficient to allow safe and confident advancement of the radiofrequency electrode away from the main pulmonary vessels. A pneumothorax occurred in one animal during the MRI-guided puncture. It was rapidly detected on MRI and was successfully treated by suction under MRI control (Figs. 1A, 1B, 1C, and 1D).
|
|
|
|
On gross pathologic examination, the mean coagulation necrosis diameter was 9.8 mm (SD, 0.78 mm) (Table 1). Findings of histopathology showed a central pale area representing thermal necrosis surrounded by a narrow dark rim corresponding to edema, congestion, and very small zones of hemorrhage. This was similar in all specimens and was not dependent on how long after the ablation the tissue was collected.
|
On MRI, oval areas of increased signal intensity compared with normal lung tissue were found on both T1-weighted and T2-weighted sequences (Figs. 2A, 2B, 2C, 2D, 2E, 2F, and 2G). These areas consisted of a central zone that was hyperintense on T1-weighted images and hypointense on T2-weighted images and a peripheral zone that was hyperintense on both T1- and T2-weighted images. STIR images showed a much higher signal intensity in the peripheral zone. Because of the excellent contrast, the peripheral zone could be better appreciated on STIR than on the T1-weighted FLASH images. A comparison of the pathologic lesion diameter with the various imaging techniques (Table 1) showed the highest ICC of 0.81 for the FLASH sequence, whereas PSIF and CT had relatively low ICCs (0.23 and 0.1). STIR allowed good delineation of the edema surrounding the thermal lesion, but STIR and PSIF imaging tended to underestimate the thermal lesion size when only the central hypointense zone was assumed to be the thermal lesion.
|
|
|
|
|
|
|
Discussion
Percutaneous thermal tumor ablation has become an alternate therapy for solid tumors in selected organs such as the liver and kidneys. The most challenging aspects of the successful use of thermal ablation are its potential to reproducibly kill all tissue in a targeted zone and its ability to precisely direct the application of thermal energy. As ablation technology improves and ablation zones become larger, monitoring thermal lesions becomes even more important to ensure both adequate coverage of the tumor and avoidance of damage to important adjacent normal structures. In parenchymal organs such as the liver, the most reliable techniques with which to guide thermal ablations are MRI [10–21] and contrast-enhanced sonography [23]. Because sonography is not feasible for lung tumors, CT has evolved as the method of choice for guiding lung ablation. However, thermal monitoring capabilities of CT are poor.
In contrast to CT, MRI facilitates the immediate monitoring of thermal tissue damage [10–21]. As shown in this study on lung tissue, MRI can be used to monitor the lung tissue response immediately after therapy with higher precision than that achieved using CT. This monitoring is important to ensure complete tumor ablation and to create a sufficient safety margin around a tumor. The latter might be difficult to achieve with radiofrequency energy in lung tissue because of the high air content in the normal lung. On one hand, the high air content leads to an oven effect as the surrounding normal lung tissue isolates the tumor, which can thus be completely destroyed. On the other hand, the high impedance of normal lung makes it extremely difficult to achieve a sufficient safety margin around the solid tumor, which is essential for locally curative treatment. Therefore, the potential of MRI to actually visualize a sufficient safety margin can contribute substantially to improving the efficacy of lung tumor ablation.
If MRI is to be used as the sole imaging technique for percutaneous tumor treatment, two other conditions must be fulfilled. Applicator insertion should be possible in a safe and rapid fashion using MRI guidance. In the past, CT was primarily used for percutaneous needle interventions in the lung. Our study shows that MRI can be used to safely direct a radiofrequency electrode in the thorax. Because no tumor was involved in this study, further evaluation is necessary. Complications such as pneumothorax [24] or bleeding must be detected during the course of treatment. The hemorrhage that was visible in the histopathologic specimen consisted of very small zones of bleeding at the border between necrosis and normal tissue that were not visible on MRI or CT. A pneumothorax occurred during ablation in one animal. It was rapidly detected and was successfully treated under MRI control. Therefore, all three prerequisites for the safe performance of imaging-guided ablation therapy were met in this study.
One shortcoming of this study was that we did not investigate the ability of MRI to determine successful lung tumor ablation. Although the potential of MRI to assess thermal necrosis in a tumor and to detect the remaining viable tumor is well known from MRI-guided tumor ablation studies involving other organs [10, 14–16, 21], this potential might be different for malignant lung tumors. One reason for not using a rabbit tumor model was that many of the available models have a different tumor biology from that of human lung cancer [25]. Also tumor implantation often leads to rapid tumor cell migration through blood and lymph circulation and tumor seeding through the needle track to the pleura, all of which prohibit local tumor treatment and therapy-related survival assessment, which are the gold standards for cancer therapies.
As in solid organ ablation, various histologic and imaging changes in a radiofrequency lesion were observed in the lung. The central pale area represents thermal necrosis, which was hyperintense on FLASH and PSIF and hypointense on STIR images. The periphery was hyperintense on both FLASH and STIR but had much higher signal intensity on STIR images. The tendency of the FLASH sequence to slightly overestimate the lesion size can be explained by the difficulty in differentiating the two hyperintense zones. STIR, on the other hand, was very sensitive to the edema surrounding the lesion and tended to underestimate the lesion size. Underestimation might be due to the fact that the edema was masking the periphery of the ablation zone (which shows more variable histologic changes) with congestion and hemorrhage close to its center and edema in its outer regions.
MRI studies of the lung are problematic because of the strong susceptibility effects originating from the tissue–air borders in the lung. In 1.5-T MRI scanners, short TEs must be used [26] to overcome this problem. Measuring at a low field strength also reduces the susceptibility artifacts because they are proportional to the external magnetic field (B0) [27]. The use of an open low-field scanner in our study not only facilitated easy access to the animal in the scanner but also reduced the susceptibility artifacts [27].
On the basis of our current results showing a technical success rate of 100% for pulmonary tissue ablation under MRI guidance and an excellent correlation between the MRI appearance of thermal lung lesions and the pathology results, we conclude that this technique is feasible. Because the entire ablation procedure can be performed and adequately monitored using MRI as the sole imaging technique, the use of MRI certainly has the potential to further improve the efficacy and clinical outcome of the ablation procedure.
Acknowledgments
We thank Bonnie Hami for her editorial assistance, Les Ciancibello for his technical assistance, and Jack A. Jesberger for his help with the statistical evaluation.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Okuma, T. Matsuoka, T. Okamura, Y. Wada, A. Yamamoto, Y. Oyama, K. Koyama, K. Nakamura, Y. Watanabe, and Y. Inoue 18F-FDG Small-Animal PET for Monitoring the Therapeutic Effect of CT-Guided Radiofrequency Ablation on Implanted VX2 Lung Tumors in Rabbits J. Nucl. Med., August 1, 2006; 47(8): 1351 - 1358. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
