| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Brigham and Women's Hospital, 75 Francis St., Boston,
MA 02115.
2 Department of Radiology, Dana Farber Cancer Institute, Harvard Medical School,
44 Binney St., Boston, MA 02115.
3 Present address: Department of Radiology, University of Massachusetts Medical
Center, Worcester, MA.
4 Department of Pathology, Brigham and Women's Hospital, Boston, MA 02115.
5 Department of Surgery, Brigham and Women's Hospital and Dana Farber Cancer
Institute, Boston, MA 02115.
Received April 16, 2004;
accepted after revision June 30, 2004.
Address correspondence to E. vanSonnenberg
(ericvansonnenberg{at}yahoo.com).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Radiofrequency ablation of 22 thoracic sites was performed in vivo in three pigs with an array-style electrode. Tissue impedance and ablation duration were measured for each site. The intact lungs were excised for gross inspection and for imaging with CT and MRI. Representative lesions were evaluated histologically.
RESULTS. The mean intraprocedural tissue impedance was 93 
(range, 52–184 ). Six of 22 ablations exhibited a marked increase
in impedance after 5 min of treatment. On gross inspection, parenchymal
lesions were generally round and targetlike in appearance. CT showed sites of
ablation to be composed of a heterogeneous inner zone surrounded by a
high-density outer zone. On MRI, the inner zone was typically hyperintense on
T1-weighted fast spin-echo imaging, and the outer zone was hyperintense on
T2-weighted fast spin-echo imaging. At histology, the inner zone was
characterized by coagulation necrosis, and the outer zone by hyperemia and
edema. No acute lung-specific complications were seen. There was one extensive
skin burn and one cardiac-related death.
CONCLUSION. These results support current seminal clinical evidence that percutaneous radiofrequency ablation in the lung is feasible and can be applied safely. Radiofrequency-induced lesions in the normal porcine lung can be visualized with both CT and MRI; image appearance is concordant with histologic tissue changes.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Published reports on experimental ablation in lung tissues are few [18–23]. The purpose of our study was to assess the use and effects of in vivo percutaneous radiofrequency ablation in normal lung tissues in a large animal model using a U.S. Food and Drug Administration (FDA)– approved device set to clinical parameters. This was done in advance of our own planned clinical application of the technique in patients. Our assessment included intraprocedural observations on the use of the device, its safety, imaging data from both CT and MRI of the acute effects, and histology of the tissues for correlation with imaging.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
45 kg) under the approval of our institution's Animal
Use Committee. The committee is accredited by the American Association for the
Accreditation of Laboratory Animal Care. The experimental procedures were
performed with the animals under general anesthesia with isoflurane and
endotracheal intubation. The ablations were performed in advance of clinical
implementation, at a time when little had been published on the use of the
device in a large animal. Array-style 3.5-cm-diameter, 13-gauge LeVeen radiofrequency ablation electrodes (Boston Scientific-Oncology) were placed percutaneously into thoracic tissues under fluoroscopic visualization (Fig. 1A, 1B). Fluoroscopy was used for guidance, rather than CT, because of restrictions on the use of animals in our CT clinical facility. A description of the ablation device and its use in the liver has been published previously by others [24].
|
|
Separate, multiple placements of the electrode in each of the three animals targeted various sites in the pulmonary parenchyma and sites adjacent to pleura, diaphragm, and pulmonary hilum. The purpose was to create central and peripheral lesions in a nonoverlapping fashion. Ablation sites were separated by several centimeters. In the first animal, 0.35-inch guidewires were inserted as identifying markers for the ablation sites. Radiofrequency ablation was performed in 22 sites. The ablation protocol consisted of an initial power setting of 30 W with an increase of 5 W per minute (n = 19 sites; initial power was 20 W for n = 3 sites).
The end point for each ablation was the rapid increase of tissue impedance (i.e., electric resistance, measured in ohms) above the relatively constant baseline value observed during energy deposition while the electrode was active. This increase in impedance is colloquially referred to as "roll-off" and is characteristic of this device. Impedance is read from the display on the control panel of the radiofrequency generator (RF 3000, Boston Scientific-Oncology). The manufacturer's instructions for use of the device considers ablation to be complete if roll-off occurs after at least 5 min of radiofrequency ablation.
Animals were sacrificed without recovery from anesthesia by an IV overdose of potassium chloride. Intact lungs were excised immediately. The excised lungs underwent CT and MRI imaging. CT scans were obtained in a Somatom Plus 4 scanner (animals 1 and 2, Siemens Medical Solutions) and Sensation 16 (animal 3) scanner (Siemens Medical Solutions) with 2-mm collimation at 120–140 kV. Scans were reconstructed into 3-mm-thick coronal slices for appreciation of the anatomy and for side-by-side comparison with the corresponding coronal MR images. MRI was performed on a 0.5-T scanner (Signa SP, GE Health-care); 5-mm-thick coronal slices were acquired with fast spin-echo pulse sequences. Both T1-weighted fast spin-echo images (TR/TE, 400/18) and T2-weighted fast spin-echo images (3,000/88) were acquired with the following parameters: 30 x 30 cm field of view, 5-mm thickness, 256 x 128 inplane resolution, 2 excitations.
The lungs were observed by gross examination. Representative identifiable lesions were subsequently removed as tissue blocks and placed in formalin for fixation. Sections from the blocks were then cut and treated with an H and E stain. Sections were reviewed in consultation with a pathologist under light microscopy for comparison of the microscopic tissue effects with the CT and MRI appearances of the lesions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Overall, during energy deposition, the base-line tissue impedance varied
from its initial value by only a few ohms (until the onset of roll-off). The
mean baseline impedance for the 22 sites was 93 (range, 52–184
). Impedance roll-off was achieved in all sites with one exception. In
six sites, roll-off was achieved in a range between 5.5 and 19.5 min; in each
of these cases, the lung parenchyma, pleura, or diaphragm was targeted. In 15
sites, roll-off occurred in fewer than 5 min; in these cases, the targets
included the lung parenchyma, hilum, pleura, and diaphragm. Impedance values
at roll-off were observed to be between 100 and 900 (n = 21).
For the exception, roll-off was not achieved even after 30 min of
radiofrequency energy deposition (baseline impedance, 53 ; 200-W
maximal power applied); in this case, the hilum was targeted, and a vessel may
have prevented roll-off. Aside from this exception, the power settings at the
time of roll-off ranged between 20 and 90 W (n = 21).
Of the six sites that received treatment for durations ranging from 5.5 to 19.5 min, five were the subject of gross, CT, and MRI examinations. (The sixth lesion is described as one of the complications.) On gross observation at the time of necropsy, these five radiofrequency ablation lesions in the normal pig lung parenchyma were generally round and target-like in appearance. The dominant central portion of the lesions appeared to be a dense coagulated region, black-purple in color (inner zone), approximately 3.5 cm in diameter. This area was surrounded by a margin (approximately 0.5-cm thick) of bright red tissue (outer zone) (Fig. 2A, 2B).
|
|
On CT, lesions generally appeared target-like with two discernible zones: there was a not-so-dense, heterogeneous central zone that was surrounded by a broad highly dense region corresponding to the inner and outer zones (described previously), respectively (Fig. 3A). The inner zone was generally ovoid; the outer zone was markedly asymmetric (Fig. 3A, 3B, 3C).
|
|
|
|
The MRI appearance of the radiofrequency ablation lesions also was targetlike. Generally, the central zone was hyperintense on T1-weighted fast spin-echo images, and the outer margin appeared slightly hyperintense or nearly isointense with surrounding tissue (Fig. 3B). In contrast, this inner zone presented as hypo- to isointense on T2-weighted fast spin-echo images. The outer zone was markedly hyperintense on T2-weighted fast spin-echo images (Fig. 3C).
Results of histology from representative lesions showed that the inner zone was characterized by two effects: one was thermal coagulation with a loss of cellular structure (usually the centralmost region and presumably as a result of highest tissue temperatures). The other was thermal coagulation with some preservation of parenchymal structures and cellular architecture. Both types of effects were seen in the inner zone of any given lesion. Overall, no normal RBCs were present in the inner zone. The outer zone was composed of areas of marked hyperemia and edema and included normal RBCs (Fig. 4); this was true for all the lesions reviewed histologically.
Of the other 15 sites in which roll-off occurred in fewer than 5 min, several were identifiable on gross observation and ranged approximately 1–3 cm in diameter. While some coagulation was evident, these lesions were relatively small, and the zones as described previously were not as developed.
The following complications occurred during the procedures: one instance of
an extensive chest wall muscle coagulation with a second-degree burn to the
overlying skin that was due to a guidewire that had been left in place during
a more peripheral ablation and that had evidently made contact with the
electrode. (This site had a baseline impedance of 65 , and rolled-off
after 19.5 min at a final power setting of 90 W.) The second complication was
one death from irreversible ventricular fibrillation after the completion of 2
min of radiofrequency ablation at 20 W (premature ventricular and atrial
contractions were observed during ablation). This was coincident with
1–2 tines of the array having pierced the left ventricular wall, as
evident on gross postmortem inspection
(Fig. 5). An increased heart
rate was observed in another animal at 1.5 min of radiofrequency ablation at
30 W (reversible, postablation). In the same animal, another cardiac
abnormality was evident, with an inverted sinus wave (reversible,
postablation) at 1.75 min of radiofrequency ablation at 35 W. There were no
instances of pneumothorax or hemoptysis.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initial experiments of percutaneous radiofrequency ablation with customized needle electrodes in normal rabbit lung tissue and in implanted tumors in rabbit lungs have been described [18, 19]. For those reports, CT was used for both intraprocedural guidance and for postprocedural follow-up. Others have reported the use of MRI for follow-up of open surgical radiofrequency ablation in implanted tumors in rabbit lungs [20]. A separate study performed the ablation in open surgery on normal porcine lungs with a standard FDA-approved clinical electrode [21]. In the latter report, lesions were assessed by gross observation and histology, immediately and at 3, 7, and 28 days. More recently, work has been published on CT-guided percutaneous radiofrequency ablation with standard electrodes in a canine tumor model, and includes CT follow-up and pathologic–histologic observation that ranged from 8–48 hr [22].
The baseline impedance values observed in our experimental lung ablations
were relatively high. This is likely due to the insulating properties of air
in the surrounding lung parenchyma. The mean baseline impedance was 93
for the 22 ablations. In comparison and in support of the hypothesis on the
contribution of air, our unpublished experimental experience with the same
electrode in normal pig liver (more dense, continuous medium) showed an
average impedance of only 30 .. In our initial experience in patients
undergoing ablation of lung tumors, the impedance averaged 65
(n = 10 patients; range, 50–90 ). In patients with
malignancies, the impedance would not be expected to be as high as in a normal
lung because the tumor mass itself would provide added conductivity.
Of the 22 ablation sites in this study, only five parenchymal sites reached
roll-off after 5 min. These were considered complete ablations compared with
the cases of early roll-off (< 5 min) that were considered incomplete.
Clinically, in the event of early roll-off, the manufacturer's instructions
for use require that the location of the electrode be adjusted and power
reapplied. In these experiments, the average baseline impedance of the
complete ablations was 68 ; that of the incomplete ablations was 103
..
No lung-specific complications occurred in these experiments. The one instance of a burn to both skin and muscle was due to a guidewire that was left in situ that likely made contact with the electrode while it was active; the guidewire likely provided an electric return path to the skin. This serves as an obvious caution against leaving such conductors (i.e., needles) in proximity to the radiofrequency electrode, because this can be a path of least resistance for the current.
The death of one animal as a result of irreversible ventricular fibrillation after 2 min of radiofrequency ablation was coincident with one or two of the tines of the array having pierced the pericardium, evident on postmortem gross inspection. However, as noted in the Results section, there was also abnormal cardiac activity (increased heart rate, inverted sinus wave, and premature ventricular and atrial contractions) on three occasions in two animals. Notably, a recent report on radiofrequency ablation in the normal lungs of three sheep noted ventricular tachycardia in all ablations near the heart [25]. In that study, the condition reversed with the removal of the electrode, and no deaths occurred. Because the operating frequency of our electrode was 460 kHz, there was no expectation of electrical interference with the heart. However, a tine could have initiated a spark (perhaps on contact with blood) that might have generated frequencies outside the operating frequency. Further study is necessary on this issue. These cardiac effects and death imply the need for caution with juxtacardiac lesions and for careful monitoring during procedures.
Separately, there were no occasions in our experience in which the electrode generated problematic tissue charring on the tines of the array or in which the operator had trouble with the retraction of the tines for probe removal from the body. Such difficulties have been reported after the use of other array-type devices [25, 26].
Various histologic observations on the acute effects of direct thermal coagulation in the normal lung have been reported [18, 20, 21, 27]. Those reports are similar to ours, especially the zones of damage that were identified. A noteworthy observation is that although the tissue of the inner zone is heterogeneously thermally damaged, the cellular architecture remains generally intact after ablation. This overall normal appearance of cell structure belies the fact that the tissues are indeed irreversibly damaged and devitalized, evident with serial follow-up as reported in the literature [5, 27].
Under our experimental conditions, both CT and MRI provided noninvasive assessment of acute tissue effects—that is, coagulation, hemorrhage, and edema. The inner zone of coagulation was well suited for visualization with T1-weighted MRI; the T1-weighted bright signal has been reported previously for thermal ablations, such as with laser ablation in normal liver [28]. It has been proposed that the shortening of the T1 relaxation time is caused by increased access of water molecules to the paramagnetic centers in blood made available by the breakdown of erythrocytes from excessive heating [29]. RBC breakdown is evident by their absence in the inner zone histologically. The postablation dehydration of the tissue and the relative increase in the concentration of tissue proteins may contribute to the shortened T1 relaxation time. The outer zone was well visualized on both CT and T2-weighted MRI. The increased density on CT and the bright T2-weighted signal on MRI are due to the presence of fluid in the tissue in the outer zone as a result of hyperemia and edema.
Unenhanced CT and MRI were used to image and assess effects and possible complications of the radiofrequency ablations. The use of CT for guidance and follow-up has been reported in clinical radiofrequency ablation of lung tumors [11–17, 23]. Although MRI has been used intraprocedurally for imaging-guided radiofrequency ablation in the liver [30, 31], it has not been used to guide procedures in the lung. MRI does have excellent soft-tissue contrast, and an article reports its use to assess clinical radiofrequency ablation in the lung postprocedurally [23]. MRI may be considered for monitoring ablation in the lung when the focus is on a tumor mass and its surrounding pulmonary parenchyma. MRI may be advantageous in monitoring heated tissues in the lung because it has the proven capability to provide quantitative temperature mapping for other ablative techniques [32, 33].
In conclusion, these in vivo ablations in normal porcine lung tissue show the feasibility of radiofrequency ablation in the lung, and the potential for CT and MRI assessment of the acute lesion without injected contrast material.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. A. Durick, P. F. Laeseke, L. S. Broderick, F. T. Lee Jr, L. A. Sampson, T. M. Frey, T. F. Warner, J. P. Fine, D. W. van der Weide, and C. L. Brace Microwave Ablation with Triaxial Antennas Tuned for Lung: Results in an in Vivo Porcine Model Radiology, April 1, 2008; 247(1): 80 - 87. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. vanSonnenberg, S. Shankar, P. R. Morrison, R. T. Nair, S. G. Silverman, M. T. Jaklitsch, F. Liu, L. Cheung, K. Tuncali, A. T. Skarin, et al. Radiofrequency Ablation of Thoracic Lesions: Part 2, Initial Clinical Experience--Technical and Multidisciplinary Considerations in 30 Patients Am. J. Roentgenol., February 1, 2005; 184(2): 381 - 390. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
