| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Experimental Studies |
1 From the Institute of Diagnostic Radiology, University Hospital Zurich, Rämistrasse 100, CH-8091 Zurich, Switzerland (T.B.); Institute of Diagnostic and Interventional Radiology, Friedrich-Schiller-University Jena, Germany (A.M., J.R.R., M.F., W.A.K.); Department of Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Mass (S.N.G., P.H., M.R.); Research Laboratories, Schering, Berlin, Germany (P.H., M.R.); Berchtold, Tuttlingen, Germany (W.M.); and Department of Biostatistics, Institute of Social and Preventive Medicine, University of Zurich, Switzerland (B.S.). Received September 20, 2001; revision requested November 23; revision received February 15, 2002; accepted April 15. Address correspondence to T.B. (e-mail: thomas_boehm@gmx.net).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Twenty-eight VX2 tumors were implanted into the retroperitoneum of 14 rabbits. Tumors were examined with contrast-enhanced fundamental color Doppler US and harmonic power Doppler US before and 10 minutes after RF ablation. Saline-enhanced RF ablation (30 mL/h) was performed over 10 minutes with 28-W RF power. Follow-up included repeat US examinations. Necropsies and histopathologic assessment were performed after detection of residual untreated tumor at US or 3 weeks after ablation.
RESULTS: VX2 tumors reached a mean size of 21 mm ± 9 (SD) (size range, 6–43 mm) 25 days after implantation. All tumors larger than 31 mm showed signs of central necrosis at US. Before ablation, intense vascularity was detected in all tumors with both contrast-enhanced US modes. Histopathologic assessment at the end of the follow-up period revealed local relapses due to incomplete ablation in 14 (50%) of 28 cases. Detection of residual tumor was missed in all cases with contrast-enhanced color Doppler US. Contrast-enhanced harmonic power Doppler US depicted residual flow in 12 of the 14 cases (sensitivity, 86%) in which local relapses occurred. There was a significant (P < .005, McNemar test) improvement in detection of residual tumor when the harmonic power Doppler mode was used.
CONCLUSION: Contrast-enhanced harmonic power Doppler US has greater sensitivity than contrast-enhanced color Doppler US for detecting residual VX2 tumor following ablation. Therefore, contrast-enhanced harmonic power Doppler US may be a useful additional method for the detection of residual tumors after RF ablation.
© RSNA, 2002
Index terms: Animals • Radiofrequency (RF) ablation • Ultrasound (US), comparative studies • Ultrasound (US), experimental studies • Ultrasound (US), harmonic study
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Treatment of breast cancer has changed considerably during the past 15 years. Breast-conserving therapy (16,17) has become established, with convincing cosmetic and functional results (18). Although breast-conserving therapy is much less invasive than mastectomy, similar therapeutic outcomes have been observed for both. The next step toward reducing invasiveness in the treatment of breast cancer may be the use of local tumor ablation with visual control.
Various techniques based on the application of local heat have been used for local treatment of breast tumors (19–22). Harari et al (19) applied high-intensity focused ultrasound for palliative treatment of tumors of the chest wall and the breast in 14 patients, while other researchers have investigated the use of lasers (20,21), cryosurgery (23,24), and RF currents (22,25). The main problems associated with all of these methods are (a) precise localization of the tumor; (b) accurate imaging-guided, controlled induction of lesions; and (c) attainment of complete destruction of all tumor cells inside the thermal lesion.
Refining the imaging procedures used to determine complete treatment so that noninvasive, reliable definition of the necrosis zone after ablation is possible is another unresolved issue, which is a difficult problem in local treatment of all tumors in all locations. Rossi et al (26) used postinterventional arterial angiography to detect residual tumor perfusion in the liver; angiography results then formed the basis for the planning of additional ablation procedures. Garbagnati et al (27) demonstrated that angiography, contrast material–enhanced computed tomography (CT), and contrast-enhanced color Doppler ultrasonographic (US) techniques could help confirm complete devascularization of small hepatocellular carcinomas after RF treatment. Contrast-enhanced harmonic US imaging has been shown to be highly sensitive for detection of microbubble contrast media in vasculature surrounded by parenchyma (28).
The aims of the present study were (a) to assess contrast-enhanced harmonic power Doppler US imaging for the detection of residual viable tumor tissue after RF ablation of tumors embedded in fat and (b) to compare contrast-enhanced harmonic power Doppler US imaging with contrast-enhanced fundamental color Doppler US imaging for the detection of residual tumor tissue by using an appropriate pathologic standard of reference.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Animals were then shaved and depilated (M.R. and T.B.) with a depilation cream (Pilca Extra Mild; Hans Schwarzkopf, Hamburg, Germany) on both sides of the abdomen for an approximately 8 x 6-cm area. Diagnostic US was performed (T.B.) to assess the implantation site. Cryoconserved tumor material that had been implanted in the lower leg muscles of an additional animal and harvested after it reached a size of 1.5 cm was placed into saline solution and cut into cubes of 1 mm3 (P.H.). All 28 implantations were performed by the same individual investigator (T.B.), who implanted specimens of the same tumor into all rabbits to minimize interanimal variations in tumor growth rate.
The implantation method used was that described by Hauff et al (29). Only portions of tumor tissue that did not show any macroscopic signs of necrosis were used. A 16-gauge intravenous cannula was placed with US control into the retroperitoneal fat below the kidney, and the prepared cubes of tumor tissue were pushed through the cannula and placed in the preselected position. The same procedure was performed on both sides of each animal.
The origin of this tumor model, the follow-up strategy, and the validity of this model as a model for breast tumor ablation in humans are described in detail elsewhere (29,30). Tumors were grown for 25 days to achieve a minimum size of 6 mm and an average diameter of 2 cm. This timed approach was used to generate a range of tumor sizes, with the goal of documenting the utility of contrast-enhanced US imaging in a mixed population of completely and partially ablated tumors.
Saline-enhanced RF Ablation
RF ablation was performed 16 days after implantation in rabbits that had been anesthetized according to the procedure described above. During ablation, 0.3 mL per kilogram of body weight of metamizol natrium (Novaminsulfon Lichtenstein; Lichtenstein Pharmazeutica, Koblenz, Germany), an analgesic drug, was also administered. For saline-enhanced ablation, an RF generator with saline-perfused needle electrodes (Elektrotom 104 HF Thermo; Berchtold Medizinelektronik, Tuttlingen, Germany), equipped with an optional impedance measurement unit, was used in a way that has previously been described (25).
The electrode used for this saline-enhanced technique has double walls distally. The inner wall has small holes. The infused saline solution flows through the hollow shaft of the electrode and escapes through these holes into the space between the inner and outer walls of the needle tip. The outer wall is constructed with larger rectangular apertures that permit the even flow of saline solution into the tissues surrounding the needle tip. Electrodes with either 2.5-, 3.0-, or 4.0-cm exposed tips were used. The size of the exposed tip was chosen to be slightly larger than the tumor diameter. For tumors larger than 4 cm, a 4.0-cm exposed tip was used.
Ablation was performed with the anesthetized animal laid on one side. Diagnostic US was performed to help localize the tumor, measure maximum tumor diameter, and assess the presence or absence of necrotic areas. After diagnostic US but before ablation, two contrast-enhanced US imaging examinations were performed: one color Doppler US examination and one harmonic power Doppler US examination (please see next section for description of these examinations).
A small incision was made in the skin, and the ablation needle was inserted with US guidance into the center of the tumor. The time required for correct placement of the ablation electrode was measured. Interstitial perfusion with sterile saline solution (0.9% sodium chloride; Braun, Melsungen, Germany) was initiated immediately before RF ablation. The saline flows through the electrode and the holes in the wall of the electrode tip into the adjacent tissue and thus prevents overheating.
Ablation was performed for a 10-minute period with an RF power of 28 W and an interstitial saline perfusion rate of 30 mL/h. Electrical impedance was monitored continuously during ablation. In case of an increase in impedance, the perfusion rate was increased up to 50 mL/h. (Anesthesia was performed by T.B., P.H., or M.R.; tumor ablation by T.B., P.H., and W.M.). After the region of ablation had cooled (ie, 10 minutes after RF heating) repeat contrast-enhanced US examinations were performed (as described below) for detection of residual perfusion.
Contrast-enhanced US Examinations
Four contrast-enhanced US examinations were performed in each animal. Both contrast-enhanced color Doppler US and harmonic power Doppler US were performed before and after ablation. Immediately before each US examination, 0.4 mL/kg (300 mg/mL) of a microbubble US contrast medium (D-galactose, Levovist; Schering, Berlin, Germany) was injected into the posterior auricular vein. This was immediately followed by injection of 1.0 mL of a physiologic saline solution. Separate injections of contrast medium were administered for each of the two US examinations performed before and after ablation (ie, four injections per animal, with two injections before and two injections after ablation). According to our protocol, additional injections of contrast medium for reevaluation of an animal with the same US method were not allowed.
The ablation procedure was monitored with B mode only. Tumors were examined with fundamental Doppler color mode and harmonic imaging power mode with the same 10-MHz transducer and US unit (HDI-3000; ATL, Bothell, Wash) before and 10 minutes after ablation. The complete tumor volume and the adjacent region of fatty tissue were scanned during the first passage of contrast medium; the transducer was moved smoothly to avoid motion artifacts. After the entire volume of interest was scanned during the first pass, the investigator returned to regions in which equivocal findings were observed or regions in which flow was detected during the first pass to check for the presence of patent vessels.
Super VHS video documentation of the contrast-enhanced US examinations was performed to permit detailed evaluation of the examinations in consensus. The video sequences were subsequently digitized (T.B. and J.R.R.) at a frame rate of 25 images per second with a MiroVideo DC30 hardware compression board (Pinnacle Systems, Braunschweig, Germany). The digitized video sequences were analyzed on an image-by-image basis for signs of residual flow (ie, colored pixels) by two experienced sonologists (T.B. and A.M.) in consensus. Vascularity was considered to be present if (a) flow was detected on several consecutive images during the first pass with the US probe over the entire region of interest and (b) this perfusion was confirmed during the subsequent dedicated US examination. The results of the contrast-enhanced US examinations were later compared with the results of necropsies and histologic analyses performed 2–3 weeks after ablation.
Follow-up
During the follow-up period of 1–3 weeks, fundamental gray-scale US examinations were performed (T.B. or P.H. or M.R.) at regular weekly intervals for detection of local relapses. An animal underwent necropsy when a local relapse was clearly detected at US. When no relapse occurred, the animal was euthanized at a well-defined end point of 3 weeks after ablation. Necropsies and histopathologic examination were performed (T.B., A.M., and P.H.). Histopathologic examination included staining of tissue with hematoxylin-eosin and search for local relapses or islets of viable tumor in the ablated tissue at optical microscopy.
Statistical Assessment
Continuous data are presented as means ± SDs. The time required for proper needle positioning in small tumors was compared with that in large tumors by using the Mann-Whitney test. Sensitivities of US modes for detecting residual tumor were compared with the McNemar test. P values less than or equal to .05 were considered to represent a significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
Insertion of the ablation electrode into larger tumors (>20 mm) was technically easier due to the presence of necrosis. It was more difficult in smaller tumors due to the rougher consistency of the tissue and the relatively high degree of tumor mobility caused by the surrounding adipose tissue. In tumors smaller than 21 mm, correct needle positioning required from 1.5 to 4.5 minutes, while 1.0–2.5 minutes were required in tumors larger than 21 mm (P = .08). No differences among tumors in baseline impedance or other RF parameters were observed on the basis of tumor location.
Before ablation, all tumors showed a high degree of vascularization after the administration of contrast medium at both fundamental color Doppler US imaging and harmonic power Doppler US imaging. Monitoring the growth of the thermal lesion during ablation with fundamental B-mode US was not possible, because soon after the ablation was begun a large area of fatty tissue surrounding the tumor became hyperechoic and produced a shadow artifact. The typical well-defined cone-shaped hyperechoic lesion observed during ablation of parenchymal organs such as the liver was not observed.
Contrast-enhanced fundamental color Doppler US imaging performed 10 minutes after ablation revealed complete absence of perfusion in all 28 tumors (100%) after ablation. However, residual flow was detected in 12 (43%) of 28 tumors with the harmonic power Doppler mode (Fig 2).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Minimally invasive resection of the tumor before or after RF ablation could be used to enable histopathologic examination and proof of tumor margins (33). The growth of the thermal lesion in the liver during RF ablation could be monitored with US or temperature-coded MR imaging (34,35). Although the latter method is more expensive, it is also more accurate because it provides reliable information about heat distribution around the needle tip. In vitro experiments on breast tissue have shown that US is not useful for monitoring thermal lesions when the tissue adjacent to the area of ablation contains fat (25). On the basis of these results and in cooperation with members of the research laboratory staff of Schering (P.H. and M.R.), we previously developed an animal model for the evaluation of methods of breast tumor treatment (30); this animal model was also used in the present study.
Animal studies are an appropriate means of investigating the efficacy and efficiency of new therapeutic methods. The kind of experimental tumor used in the present study is almost ideal for evaluations of the quality of local tumor treatment techniques (30). The high malignancy of this tumor, when implanted in immune-competent rabbits, has yielded a response rate higher than 80% (36). In the present study, a 100% response rate was achieved.
Interventional techniques for tumor treatment must be applied to models that will allow for assessment of equipment to be used in patients. Consequently, very small animals like rats and mice are not appropriate models for these experiments. The VX2 tumor grows quickly, at a rate of 1 mm to more than 2–3 cm in about 2–3 weeks. This rapid growth rate makes it possible to check the completeness of tumor necrosis with serial follow-up imaging examinations during a relatively short time interval. This approach was preferred to histochemical methods of assessing for residual viable tumor because pathologic examination is an expensive and not always accurate alternative (37,38).
When the assumption is made that an induced thermal lesion is homogeneous, only a single part of the specimen is often sampled for pathologic evaluation (22). However, this assumption is not justified, especially in heterogeneous tumors that consist of several tissue compartments that have different tissue impedances and uneven thermal distribution. Even if one or a few viable tumor cells survive the ablation procedure, this may be sufficient to cause local relapses and, consequently, a failure of local tumor treatment.
The possibility of being able to perform better and easier quality control with follow-up examinations over a reasonable time period is an important argument in favor of the present tumor model. Additionally, the aggressiveness of the tumor used in this model and the use of large tumors facilitated our primary goal of determining the utility of contrast-enhanced US in evaluating for the presence of residual unablated tumor. Therefore, incomplete treatment in a certain percentage of tumors was one of the prerequisites of the study and another rationale for the use of larger tumors.
It has previously been shown that VX2 tumors grow even in locations with an initially poor blood supply, such as areas of retroperitoneal fat (30). This fact makes treatment of VX2 tumors a suitable model for local treatment of breast cancer, which is surrounded, at least in part, by adipose tissue. Adipose tissue is a rather complicated tissue in controlled thermal ablation due to the liquefaction of fat (25). Thus, RF ablation of VX2 tumors implanted in adipose tissue is a suitable model for RF ablation of highly malignant breast cancers because this model simulates those conditions in which achieving controlled and complete tumor treatment is the most difficult. VX2 tumors are highly vascularized due to induction of neovascularity in the host animal (39). The efficiency of local ablation techniques may consequently be reduced because of the heat-sink effects of flowing blood (40).
The problem of residual tumor detection after RF ablation has been explored previously by other researchers in liver tumor ablation (41,42). In ablation of liver tumors, contrast-enhanced spiral CT or MR imaging is routinely used, whereas the ablation procedure itself is most often monitored with B-mode US. However, detection of residual tumor with the same imaging modality would be very convenient for immediate guidance of additional RF ablation during the same session.
Solbiati et al (41) evaluated whether contrast-enhanced US could be used for residual tumor detection. They reported improved detection of residual tumor with color Doppler US performed after intravenous administration of D-galactose compared with nonenhanced US. However, even after administration of contrast material, the sensitivity of US was only 50% (specificity was 100%). Additionally, no histopathologic assessment was included in the design of that study, and results of CT imaging were used as the standard of reference. Assuming that contrast-enhanced CT also did not detect all foci of residual tumor, the sensitivity of contrast-enhanced US in that study may be even lower than 50%. Nevertheless, when residual perfusion was detected, additional information was gained inasmuch as the foci of enhancement served as a marker for real-time guidance for additional RF ablation.
Goldberg et al (42) reported results of an experimental study in which a polymeric microsphere contrast agent was used with US for the detection of residual tumor after RF ablation of rabbit VX2 tumors implanted in the liver. The diameter of histologically proved necrosis correlated well with the size of the nonperfused area seen on US images obtained after ablation. In two cases of histologically proved residual tumor, residual perfusion was detected on contrast-enhanced US images (sensitivity 100%; specificity 100%). Because that study was not designed to verify the possibility of residual tumor detection but rather to investigate whether the size of a thermal lesion correlates with the area of the perfusion deficit seen on US images after ablation, small tumors, which are more likely to be completely ablated, were used.
Because they used an alternative US contrast medium, it is difficult to compare the results observed by Goldberg et al (42) with those observed in the present study and with those observed by Solbiati et al (41). Nevertheless, the criterion used in the present study—that is, ablation was considered complete only if no local relapse was detected after 3 weeks of follow-up—is potentially stronger than the imaging criteria used by Solbiati et al (41) or the histopathologic criteria used by Goldberg et al (42).
The current study demonstrates that harmonic power Doppler US imaging is efficient in depicting residual tumor after ablation of VX2 tumors. The excellent sensitivity of contrast-enhanced fundamental color Doppler US reported by Goldberg et al (42) could not be confirmed in the present study. Differences in the contrast medium used and/or differences in the strategy of histopathologic correlation may explain the discrepant results. Despite the high sensitivity of contrast-enhanced harmonic power Doppler US, viable residual tumor may persist and cannot be safely ruled out even if no perfusion is detected in the area of ablation. Contrast-enhanced harmonic power Doppler US imaging, however, can be used as an additional diagnostic procedure performed immediately after RF ablation to enable additional RF treatments for the ablation of residual viable disease identified at the time of initial therapy. This can be followed with either contrast-enhanced CT or contrast-enhanced MR imaging.
The peritumoral adipose tissue in our animal model is hypovascular and serves as a good model for the fat that is commonly seen surrounding breast cancer in the clinical setting. Given minimal baseline enhancement, contrast-enhanced US is not capable of delineating the extent of the zone of necrosis as a postablation perfusion deficit in this fatty tissue in the way it can delineate necrosis in hypervascular tissues surrounding a liver tumor. Hence, the only information gained is limited to detection of the presence or absence of residual perfusion within the tumor only, and no assessment of small foci of tumor extension into the surrounding fat is possible. Thus, despite the relatively good performance of contrast-enhanced harmonic power Doppler US imaging in the detection of residual tumor, additional imaging may be required.
Furthermore, B-mode US, which is widely used to monitor the ablation procedure during treatment of liver tumors, was not useful in the presence of fat adjacent to the tumor. Use of contrast-enhanced MR imaging with simultaneous temperature mapping may potentially overcome these limitations in monitoring ablation and detecting residual tumor in RF ablation of breast tumors. However, monitoring of RF ablation with MR imaging is still at the experimental level. To our knowledge, only one vendor (Berchtold) offers a commercially available MR-compatible RF generator with titanium saline–perfused ablation electrodes, but this may not be the most ideal method for RF application in the breast (43).
Furthermore, most interventional breast coils for closed magnets are not compatible in size with the commercially available RF equipment, and, during application of RF current, no imaging is possible with an MR system. Low-field-strength open magnets, on the other hand, are not widely available and offer lower image quality compared with 1.0- or 1.5-T closed magnets. However, MR image quality has been shown to be sufficient for preoperative localization and biopsy (44). Whether image quality with low-field-strength MR imaging is sufficient for reliable detection of residual tumor immediately after breast tumor ablation has been performed in the low-field-strength magnet remains unclear. Therefore, given that its sensitivity in this study was 86%, contrast-enhanced harmonic power Doppler US imaging may be clinically useful as an additional modality for detecting residual tumor and guiding further intervention.
Practical application: In an experimental design that promoted a considerably high frequency of incomplete treatment after RF ablation, harmonic power Doppler US imaging with D-galactose had a sensitivity of 86% for depiction of residual tumor in initially highly vascularized tumors. This method is easy to use at the time of initial US-guided ablation, and, compared with MR imaging, is not expensive. Substantial information about areas of incomplete ablation may be gained and may immediately be used to guide further ablation. Further studies are necessary to determine the sensitivity of this US contrast method after RF ablation of breast tumors in the clinical setting. Defining an appropriate stepwise sequence of studies of RF ablation in the breast is also necessary. First, RF ablation should be assessed in a clinical trial for treatment of painful and growing fibroadenomas. Once success has been achieved in benign lesions, treatment of breast cancer in patients who do not wish to undergo more invasive treatment could be considered. Alternatively, RF could be studied as a neoadjuvant therapy for reducing the bulk of breast cancers prior to surgery. Success in ablating these tumors under these conditions and further improvement in defining tumor margins noninvasively with MR imaging may ultimately allow application of RF ablation in lieu of resection in the primary treatment of small breast cancers.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, T.B., A.M.; study concepts, T.B., A.M., J.R.R., S.N.G., M.F., W.A.K.; study design, T.B.; literature research, T.B., P.H., S.N.G., J.R.R.; experimental studies, T.B., P.H., M.R., A.M., W.M., M.F.; data acquisition, T.B., A.M., P.H., M.R.; data analysis/interpretation, T.B., A.M., P.H., J.R.R., B.S.; statistical analysis, T.B., B.S.; manuscript preparation, T.B.; manuscript definition of intellectual content, T.B., S.N.G.; manuscript editing, T.B.; manuscript revision/review, A.M., M.F., P.H., J.R.R., W.A.K.; manuscript final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  O. Seror, G. N'Kontchou, D. Haddar, M. Dordea, Y. Ajavon, N. Ganne, J. C. Trinchet, M. Beaugrand, and N. Sellier Large Infiltrative Hepatocellular Carcinomas: Treatment with Percutaneous Intraarterial Ethanol Injection Alone or in Combination with Conventional Percutaneous Ethanol Injection Radiology, January 1, 2005; 234(1): 299 - 309. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
