| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Departments of Radiology (S.A.S., L.A.S., T.C.W., F.T.L.), Surgery (K.M., A.S.W., D.M.M.), Pathology and Laboratory Medicine (T.F.W.), and Biostatistics (J.P.F.), University of Wisconsin Hospitals and Clinics, E3/311 Clinical Science Center, 600 Highland Ave, Madison, WI 53792. Received October 18, 2002; revision requested January 7, 2003; final revision received May 13; accepted August 14. Supported in part by Vivant Medical, Mountain View, Calif. Address correspondence to F.T.L. (e-mail: ftlee@wisc.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Microwave energy was applied for 7 minutes at 60 W in six porcine livers (mean weight, 68.2 kg) by using single (n = 7) or dual 2.7-cm loop microwave probes in parallel (n = 9) or orthogonal (n = 9) configurations. Volume, diameter, shape, and temperature of the zone of necrosis and the presence of viable tissue inside the loop were determined and compared by means of factorial analysis of variance.
RESULTS: Mean lesion volume and maximum diameter, respectively, were 32.2 cm3 ± 14.4 (SD) and 4.6 cm ± 1.4 for lesions ablated with parallel probes (parallel lesions), 29.5 cm3 ± 8.1 and 4.3 cm ± 0.6 for lesions ablated with orthogonal probes (orthogonal lesions), and 6.4 cm3 ± 1.9 and 3.4 cm ± 0.62 for lesions ablated with single probes (single lesions) (P < .05, single vs parallel and orthogonal lesions). Mean minimum diameter was greatest for orthogonal lesions (3.5 cm ± 0.53; P = .017, parallel vs orthogonal lesions). Orthogonal lesions had the highest mean internal temperature (97.2°C) versus parallel (91.9°C) and single (60.0°C) lesions. All orthogonal lesions heated to 60°C in comparison to eight of nine parallel and four of seven single lesions. The mean time to reach 60°C was shortest for orthogonal lesions (93.3 seconds) versus parallel (123.8 seconds) and single (263.0 seconds) lesions. Orthogonal lesions were the most spherical. Viable tissue was present in the center of five of seven single, six of nine parallel, and zero of nine orthogonal lesions.
CONCLUSION: Loop microwave antennas allow precise control and effective ablation of targeted tissue, particularly in the orthogonal configuration.
© RSNA, 2004
Index terms: Animals • Liver neoplasms, therapy • Microwaves • Radiofrequency (RF) ablation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Microwave ablation offers many of the advantages of RF ablation but has several other theoretic advantages that may increase its effectiveness in the treatment of tumors. The zone of active tissue heating in RF ablation is limited to a few millimeters surrounding the probe (8). Heat then radiates from this zone into surrounding tissue by means of thermal conduction. Unfortunately, temperature decreases rapidly with increasing distances from the RF probe (1/r4, where r is radius) (8). Compared with RF ablation, microwave ablation has a wider zone of active heating up to 2.0 cm surrounding the antenna (9). This may allow more uniform tumor killing both within a targeted area and next to vessels. Unlike RF energy, microwave energy does not appear to be limited by charring and tissue desiccation; thus, lesion temperature may become considerably higher with microwave systems than it does with RF systems (10).
One limitation of microwave ablation therapy is the inability to treat large tumors without numerous overlapping ablation procedures; in one study (11), a mean of 46 ablation procedures were required. Recent engineering advances have resulted in a new design of microwave antennas (Vivant Medical, Mountain View, Calif) that is tuned to the dielectric properties of the liver, and this design reduces feedback and increases the amount of energy deposited (12). With this system, as many as eight probes can be operated simultaneously (10). Clinical trials with this device are currently underway in the United States.
To date, only straight antennas have been used for microwave ablation. These antennas have resulted in elliptic zones of necrosis that may necessitate overlapping ablation procedures to cover a round tumor (13). In addition, use of straight microwave antennas and RF electrodes necessitates puncture of the tumor, with an associated risk of tract seeding. Recently, a prototype loop-shaped microwave antenna was developed. Theoretic advantages of this configuration include the ability to encircle a tumor, deliver large amounts of precisely targeted microwave energy to the tumor with minimal collateral damage to normal structures, and decrease the risk of tumor seeding because the tumor is not punctured. We hypothesized that simultaneous use of more than one loop antenna to create a "cage" around the tumor may lead to very high intratumoral temperatures as a result of the simultaneous delivery of microwave energy from multiple sources and the shielding effect of multiple probes on vessel-mediated cooling. Thus, the purpose of this study was to determine the effectiveness of tissue ablation with a loop microwave antenna in various configurations in porcine liver tissue.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
After the liver was exposed, lesions were created with single (n = 7), parallel (n = 9), or orthogonal (n = 9) loop microwave antenna configurations (hereafter, single, parallel, and orthogonal lesions, respectively). Each probe was placed in an area of the liver lobe where the antennas would be entirely within the substance of the liver parenchyma when they were deployed (Fig 1). Twenty-five microwave lesions were produced in six pigs: One lesion was produced in each of the four lobes of the livers in the six pigs, and a second lesion was produced in the exceptionally large median lobe of one of the six pigs. The potential for bias was reduced by using an incomplete block design where each configuration was used at least once and at most twice in each liver.
|
|
|
Single microwave probes (n = 7) (Fig 1a) inserted into the liver to measure temperature were deployed as described earlier. A 21.5-gauge fiberoptic thermocouple (Probe 790; Luxtron, Santa Clara, Calif) was then placed into the center of the deployed loop antenna with intraoperative ultrasonographic (US) guidance.
Dual-loop–antenna thermal lesions were formed in one of two configurations: parallel or orthogonal (Fig 1b, 1c). Each of these configurations was created by fixing two single probes in a handle either 1.6 cm (parallel) or 1.7 cm (orthogonal) apart. The parallel configuration (n = 9) consists of two loop antennas deployed parallel to each other. In contrast, the orthogonal configuration (n = 9) also consists of two loop antennas, but the probes are canted toward each other at a 45° angle. When deployed, these loop antennas can encircle a lesion placed between them. For this study, both types of dual lesions were created by placing the probes into pig liver with direct inspection. The loop antennas were each deployed with 60–70 W of power from the electrocautery generator. A thermocouple was then placed directly into the center of the two loop antennas by passing an 18-gauge introducer needle through a hole that had been drilled into the probe handle. A thermocouple was then passed through the needle, which was withdrawn from the liver tissue. All lesions were created with 60 W of power applied for 7 minutes.
Sacrifice of Animals
The animals were sacrificed by using an intravenous injection of pentobarbital sodium and phenytoin sodium (Beuthanasia-D [0.045 mL/kg]; Schering-Plough, Kenilworth, NJ). The livers were removed en bloc and sliced at 3-mm intervals. Three slices from each lesion (one from each distal end and one from the center of the lesion) were stained for viability with 2,3,5-triphenyltetrazolium chloride (TTC), which is reduced by enzymes of functioning mitochondria to yield a deep blue-purple color (14). Slices were placed directly on an optical scanner (Perfection 2450 Photograph, model G860A; Epson, Long Beach, Calif), and images were saved as electronic files. Select portions of each type of lesion were mounted in paraffin blocks, sliced at a thickness of 10 µm, and stained with hematoxylin-eosin. Samples were chosen for histopathologic examination from tissue adjacent to vessels and in peripheral and central portions of thermal lesions for each loop antenna configuration. At histopathologic examination, tissue was examined for the following features: preservation of lobular architecture; appearance of hepatocytes including cell membranes, cytoplasm, and nucleus; appearance of sinusoids; and presence of intact erythrocytes and Kupffer cells.
Lesion Volume and Diameter Determination
All lesion volumes and diameters were determined by consensus of two authors (S.A.S., L.A.S.). The zone of necrosis was measured by using software (NIH Image; National Institutes of Health, available at: www.rsb.info.nih.gov). The total volume per lesion (V) was calculated by multiplying the area for each lesion by the section thickness and then summing each individual volume: Vtotal = (A1 x T) + (A2 x T) . . . (An x T), where A is the area of each individual slice, T is the section thickness, and n is the total number of lesions (15).
Shape of Zone of Necrosis
Lesion shape was estimated by one author (S.A.S.) by calculating the mean isoperimetric ratio for each lesion type in a representative slice. This ratio provides an estimation of "roundness" in two dimensions (15,16). The closer the value is to unity, the more round (R) the lesion. This value was computed with the following formula: R = 4
A/L2, where A is area and L is perimeter.
Statistical Analysis
The sample sizes of seven single, nine orthogonal, and nine parallel lesions are comparable to those in previous animal studies of microwave ablation (10,13). We expected that there would be large differences between the single lesions and the parallel and orthogonal lesions, and it was these differences that were of interest. A post hoc power analysis was undertaken with the volume data from our study by means of pairwise t tests with Bonferroni correction to maintain overall type I error of .05.
For this study, interactions between thermal lesions and antenna configurations were believed to be minimal, if present. The pig liver is divided into four anatomically distinct lobes that are separated by deep fissures. Each lobe has a separate blood supply, and the lobes were shielded with surgical pads during ablation. Thermal lesions were also created consecutively and not simultaneously, and this method minimized the possibility of electrical interference.
Lesion volume, maximum and minimum diameters, maximum internal temperature, time to heat to 60°C, and isoperimetric ratio for each group were compared by means of factorial analysis of variance (S.A.S., F.T.L.). The P values were calculated with the Fisher probable least significant difference test. Differences with a P value of .05 were considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Size of Zone of Necrosis
The size of the zone of necrosis, which was measured on the basis of volume, maximum diameter, and minimum diameter, is summarized in Table 1. Mean lesion volume was greatest with the parallel loop configuration followed by the orthogonal and single loop configurations (P < .05 for single vs parallel and orthogonal, P = .58 for parallel vs orthogonal; analysis of variance). Mean maximum lesion diameter with the parallel configuration was also greater than that with either the orthogonal or single configurations (P < .05 for single vs parallel and orthogonal configurations, P = .57 for parallel vs orthogonal configuration; analysis of variance). Mean lesion minimum diameter was greatest with the orthogonal configuration followed by the parallel and single configurations (P < .05, analysis of variance).
|
|
|
|
|
|
|
|
|
|
Zone 1.—Peripheral hemorrhagic border of lesions treated with microwave ablation consisted of intact liver cell plates, visible cell borders, normal punctate cytoplasmic basophilia (RNA), and normal nuclear morphology. The sinusoids were congested with intact erythrocytes, but the sinusoidal lining cells (Kupffer cells) were separated from hepatocytes (space of Disse). The lobules were normal in outline, with normal septal fibrocollagenous tissue. In this zone, some viable-appearing hepatocytes were present (Fig 6a).
|
|
|
Zone 3.—This is the most central zone of the microwave lesion, and liver damage appeared the most severe. In this area, liver cell plates were compressed and distorted, especially by cystic spaces. Liver cells were separated, and sinusoids were filled with amorphous basophilic or eosinophilic material. This material was also present in some cystic spaces; others were empty, probably gas-filled spaces. Connective tissue spaces, triads, and septae were edematous, and collagenous fibers were basophilic (Fig 6c).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
A second advantage of loop microwave antennas is the ability to precisely control the area of ablation. In this study, use of an orthogonal configuration resulted in the destruction of all tissue encircled by the loop, even next to blood vessels. In contrast, these areas are relatively protected from RF ablation as a result of vessel-mediated cooling (17). This has the potential to be a major advantage for loop microwave ablation compared with RF ablation. With the advent of more powerful RF devices, the precise destruction of targeted tissue without collateral damage has become more tenuous, particularly when the devices are combined with infusion of saline or other conductors (18,19). The issue of precise control of the area of ablation is particularly important because of the lack of a highly accurate and widely accepted method to monitor any type of thermal ablation produced by heating while in progress (20).
A major question addressed in this study was how to determine the best geometric configuration to deploy multiple loop microwave antennas. It is clear that a single loop used in isolation is not sufficient for clinical use. We hypothesized that multiple loop antennas encircling the tumor in a cage configuration would have thermal advantages, but the exact configuration was unknown. The most important factor in the design of loop microwave antennas is assurance of a consistent and symmetric (preferably spherical) thermal lesion. Spherical thermal lesions would be most likely to result in the best tumor coverage. The dual loop antenna systems, parallel and orthogonal, in this study employed the minimum or maximum amount of overlap, respectively, while they still produced a symmetric lesion.
On the basis of results of this study, the orthogonal configuration appears to have substantial advantages compared with a parallel configuration in terms of intralesional temperature and completeness of ablation in perivascular areas encircled by the loop. Despite the larger volume of tissue ablated with the parallel configuration, orthogonal lesions had greater mean minimum diameters and were more spherical in shape; both are important limiting factors in determination of the size of tumor that can be ablated with a specific instrument. It is possible that other array shapes may perform even better than the current orthogonal design, but that question was beyond the scope of this study. Furthermore, different applied powers, microwave frequencies, and application times and larger loop microwave antennas were not studied.
There are several disadvantages to the use of loop microwave antennas delivered through a single needle compared with use of straight microwave antennas or RF arrays. The most obvious disadvantage is the more complicated targeting of tumors. The ability to precisely place the loop antennas will be greatly increased if or when US-based biopsy guidance software becomes available to demonstrate the predicted path of the loop antenna prior to deployment. Accurate placement of the system used in this study will require excellent US procedural skills, including the ability to precisely place and follow the needle–loop antenna complex in three dimensions. This may be very difficult to do in a percutaneous environment because of the limited area through the body wall to place a second probe at a fixed distance of 1.7 cm from the first probe. Furthermore, physicians who use this system must be very confident that the loop antenna entirely covers the targeted tumor, because the amount of ablation outside the loop is limited to approximately 0.4–0.8 cm (mean minimum or maximum orthogonal lesion radius minus the radius of the loop). However, this is not different from ablation with current RF arrays, where coverage of the tumor by the thermal lesion is dependent on precise probe placement, which can be difficult to confirm with imaging.
An additional difference between placement of loop microwave antennas and conventional RF electrodes or microwave antennas is the use of a Bovie electrocautery unit during deployment of the loop antenna to assist "cutting" through the target tissue. This has the effect of easing deployment of the loop with minimal deflection and distortion. However, malpositioning of the loop antenna may cause more severe tissue damage than that expected during probe placement with systems that do not use RF energy during prong deployment. Placement of an energized loop antenna through bowel, especially the colon, may create an enterotomy that could require surgical repair. Injury to other adjacent structures, such as the gallbladder or central bile ducts, is also possible. This underscores the importance of precise and accurate imaging guidance during probe placement and loop antenna deployment.
A limitation of this study and similar studies with normal pigs is the lack of a tumor model to determine the exact degree of tumor cell death with this system. In particular, the multitude of vessels of various sizes that course between the deployed loops would not likely be encountered when a metastatic or primary tumor is encircled. While some large-animal tumor models exist, they are expensive and complicated and may not be applicable to the study of human colorectal metastases or primary hepatic tumors that are encountered in clinical practice.
Practical application: Loop microwave antennas can be used in a thermal ablation method that creates high intratumoral temperatures that reproducibly and precisely ablate all targeted tissue when deployed in an orthogonal configuration. The major practical impediment to use of loop antennas, particularly percutaneously, appears to be the creation of an accurate, reproducible, and simple method for the targeting of tumors to be encircled by the loop.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, F.T.L.; study concepts, F.T.L.; study design, F.T.L., T.F.W., D.M.M., J.P.F.; literature research, F.T.L., S.A.S.; experimental studies, S.A.S., F.T.L., K.M., A.S.W., T.F.W., L.A.S.; data acquisition, S.A.S., L.A.S., F.T.L., K.M., A.S.W., T.F.W.; data analysis/interpretation, S.A.S., F.T.L., T.F.W., T.C.W., J.P.F.; statistical analysis, S.A.S., F.T.L., L.A.S., J.P.F.; manuscript preparation, S.A.S., F.T.L., T.C.W., J.P.F.; manuscript definition of intellectual content, S.A.S., F.T.L., J.P.F.; manuscript editing, S.A.S., F.T.L., A.S.W., T.F.W., T.C.W.; manuscript revision/review, all authors; manuscript final version approval, F.T.L., T.C.W.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. J. Wolf, D. J. Grand, J. T. Machan, T. A. DiPetrillo, W. W. Mayo-Smith, and D. E. Dupuy Microwave Ablation of Lung Malignancies: Effectiveness, CT Findings, and Safety in 50 Patients Radiology, June 1, 2008; 247(3): 871 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. E. Clark, R. D. Woodruff, R. J. Zagoria, and M. C. Hall Microwave Ablation of Renal Parenchymal Tumors Before Nephrectomy: Phase I Study Am. J. Roentgenol., May 1, 2007; 188(5): 1212 - 1214. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kuang, M. D. Lu, X. Y. Xie, H. X. Xu, L. Q. Mo, G. J. Liu, Z. F. Xu, Y. L. Zheng, and J. Y. Liang Liver Cancer: Increased Microwave Delivery to Ablation Zone with Cooled-Shaft Antenna--Experimental and Clinical Studies Radiology, March 1, 2007; 242(3): 914 - 924. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Simon, D. E. Dupuy, D. A. Iannitti, D. S. K. Lu, N. C. Yu, B. I. Aswad, R. W. Busuttil, and C. Lassman Intraoperative Triple Antenna Hepatic Microwave Ablation Am. J. Roentgenol., October 1, 2006; 187(4): W333 - W340. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. U. Hines-Peralta, N. Pirani, P. Clegg, N. Cronin, T. P. Ryan, Z. Liu, and S. N. Goldberg Microwave Ablation: Results with a 2.45-GHz Applicator in ex Vivo Bovine and in Vivo Porcine Liver Radiology, April 1, 2006; 239(1): 94 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. C. Yu, D. S. K. Lu, S. S. Raman, D. E. Dupuy, C. J. Simon, C. Lassman, B. I. Aswad, D. Ianniti, and R. W. Busuttil Hepatocellular Carcinoma: Microwave Ablation with Multiple Straight and Loop Antenna Clusters--Pilot Comparison with Pathologic Findings Radiology, April 1, 2006; 239(1): 269 - 275. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Simon, D. E. Dupuy, and W. W. Mayo-Smith Microwave Ablation: Principles and Applications RadioGraphics, October 1, 2005; 25(suppl_1): S69 - S83. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) caused by formation of noncontiguous lesion. (b) Orthogonal lesion with tissue necrosis throughout. Note large vessel (arrow) that creates an indentation on lesion periphery.
