| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From Anne Arundel Medical Center (LT, JEW, KLF, KTS), Annapolis, Maryland; and Pennsylvania State University College of Medicine (SJS, RTG), Hershey, Pennsylvania.
Correspondence: Address correspondence and reprint requests to: Lorraine Tafra, MD, The Breast Center, Anne Arundel Medical Center, 2002 Medical Parkway, Annapolis, MD 21401; E-mail: ltafra{at}aahs.org
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: Patients with ultrasonographically visible breast cancers that otherwise would have required wire localization participated. Before lumpectomy, a cryoprobe (Visica; Sanarus, Pleasanton, CA) was inserted through a 3-mm skin incision and directed by ultrasonography through the center of the tumor. An ice ball was created that enveloped the tumor plus an adjacent 5–10 mm of sonographically normal breast tissue.
Results: Twenty-four CAL procedures were performed and all lesions were successfully localized. Mean (±SD) tumor size was 1.2 ± .4 cm (range, .7–2.0 cm). Mean dimensions of the ice ball before excision were 3.9 ± .3 cm by 2.5 ± .5 cm, and the ice margin around the tumor was 8 ± 2 mm. The size of the ice ball was controlled to the millimeter, and the ice ball itself provided a precise template around which to dissect. The margin re-excision rate was 5.6% among patients with an ice margin greater than 6 mm.
Conclusions: CAL is a superior alternative to wire localization. Ultrasonographic visualization of the ice ball allows the size of the margin and tissue resected to be individually tailored and accurate within millimeters. The created template allows a precise lumpectomy, adding a dimension of control not previously realized with any other technology.
Key Words: Breast neoplasms • Breast-conserving surgery • Cryosurgery
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The management of nonpalpable breast cancer has involved a variety of imaging-based localization techniques, including needle placement and ultrasound.2–8 Needle localization is fraught with many pitfalls that have led to a re-excision rate of nearly 40% to 60%9–15 and, not surprisingly, a good deal of patient dissatisfaction. Needle localization also strains hospital resources, requires frequently difficult schedule coordination between departments, and is uncomfortable and inconvenient for the patient.
We hypothesized that the highly ultrasonographically visible ice ball created by cryosurgery could be used to advantage in the breast and possibly eliminate the need for needle-wire localization. The goals of the pilot trial were therefore to determine (1) the feasibility of intraoperative cryoprobe placement centrally in ultrasonographically visible breast cancers; (2) the feasibility of creating and controlling the size of the ice ball as it engulfs the tumor and the adjacent margin of normal tissue; (3) the advantage, if any, of creating a three-dimensional template around which to dissect the lumpectomy specimen; (4) the appropriate size of the ice ball beyond the ultrasonographically visible tumor that would most frequently result in negative margins; (5) associated morbidity; and (6) the affects of cryosurgery on pathological interpretation of the margins and primary tumor.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CAL Procedure
All procedures were performed with the Visica Cryoablation System (Sanarus Medical, Inc., Pleasanton, CA), involving the use of a 2.7-mm cryoprobe under ultrasonographic guidance (Fig. 1). The cryoprobe is highly insulated proximal to its distal 40-mm active zone. The lesion was localized with ultrasonography and its longest diameter was determined. Because the length of the ice ball along the cryoprobe is always its longest dimension, every effort was made to place the probe through the long axis of the cancer. The probe was inserted into the lesion through either the lumpectomy incision or a separate stab incision. Central placement, critical to the accuracy of the procedure, was confirmed by rotating the ultrasonography probe 90° to obtain a transverse image (cross-section) of the cancer, with the cryoprobe visualized on end in its center.
|
After resection the system was switched to the warm setting, triggering discontinuation of argon gas flow and initiation of helium gas flow, thus warming the probe and facilitating its removal. The frozen CAL specimen was thawed in warm sterile saline, and ultrasonography of the specimen was performed. If visual or ultrasonographic inspection or manual palpation raised concern that a margin might be involved, the surgeon could take more tissue from the breast at that margin.
Pathology
In the first eight cases, IHC was performed on the specimens. Several specimens demonstrated lack of ER-, PR-, or HER-2/Neu receptor concordance between the LCNB and CAL specimens. To avoid confusion after these first eight cases, IHC was no longer performed on specimens. A designated pathologist at each institution evaluated each of the CAL specimens and indicated on a form their quality with respect to type, pattern, nuclear grade, gland formation, mitotic rate, and lymphatic invasion.
Statistics
Fisher’s exact test was used to assess the relationship between the CAL specimens with positive margins and those without. All reported P values for Fisher’s exact test were two-tailed. Significance was determined at P < .05. A two-tailed unpaired t-test was used to evaluate the difference in mean ultrasonographic and pathologic tumor sizes.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Feasibility of Cryoprobe Localization
In all patients the lesion was visible intraoperatively and was successfully localized. All the lesions were centrally located in the lumpectomy specimen on gross examination. Figures 2 and 3 
demonstrate the easily visualized, centrally placed cryoprobe through the tumor.
|
|
|
|
Bleeding was minimal in all cases. Because the ice ball remained fixed to the probe during the dissection, leverage applied to the handle assisted in counter-traction, facilitating resection of the specimen. Unlike tissue removed during standard lumpectomy that is frequently fatty and difficult to handle, the CAL specimens became firm with freezing and were easy to manipulate. Similarly, dense, fibrous breast tissue, which frequently makes discrimination between tumor and normal tissue difficult, was well demarcated by the ice ball and allowed the distinction to be made without effort. Figure 5 shows the cryoprobe placement in the breast during excision.
|
Three of the four cases of positive margins occurred in the first half of the study, during which time the goal had been to create an ice ball that extended up to 5 mm beyond the tumor. This prompted an increase in the size of the ice ball to 5–10 mm beyond the tumor in the second half of the study.
After this adjustment in the ice ball size, one patient had a positive margin. Although this patient had 10 mm of tissue beyond the tumor border engulfed within the ice ball, a second primary tumor (of different histological type) was found in the CAL specimen at the margin. This second primary tumor, an invasive and in situ lobular carcinoma, had not been detected preoperatively or at the time of intraoperative ultrasonography. Re-excision of this patient’s CAL cavity revealed no evidence of residual tumor.
Two of the four patients who required re-excision in the study had ice balls that engulfed the tumor plus a 5-mm margin on ultrasonography. These tumors had ultrasonographic diameter measurements of 7 mm x 6 mm x 6 mm and 8 mm x 8 mm x 5 mm; both correlated fairly closely to the maximum tumor diameter found at histological examination. On initial lumpectomy both tumors had positive deep margins that abutted the pectoralis major. In the first case, the margin was involved with DCIS. In the other case, the margin was involved with invasive ductal carcinoma. Both re-excision specimens were negative for residual tumor. A third patient requiring re-excision had an ice ball that engulfed a 6-mm margin on ultrasonography. The ultrasonographic diameter measurements of the tumor were 9 mm x 5 mm x 5 mm. Maximum tumor diameter in the pathology specimen was reported as 5 mm, and the histologic finding was mixed ductal and lobular carcinoma; both invasive and in situ cancer was present at the margins. Re-excision still did not clear the in situ disease.
Ice ball and tumor dimensions in those patients with positive margins and those with clean margins were compared (Table 2). No statistical differences were found. However, for those patients in whom the ice ball encompassed no more than a 6-mm margin around the tumor (n = 6), the re-excision rate was 50.0%, whereas for those in whom the ice ball encompassed a margin larger than 6 mm (n = 18), the re-excision rate was 5.6%; 
2analysis demonstrated this difference to be significant (P < .04). The tumor size, as assessed by pathology, was larger than that determined by ultrasonography (14 mm ± 5 mm and 12 mm ± 4 mm, respectively), but this was not statistically significant.
|
Effects of Cryosurgery on Pathological Interpretation
The tumor size in the CAL specimens could be assessed without difficulty, and in no case was the ability to evaluate the margins hampered. Without the availability of prior LCNB specimens, however, the quality of the CAL specimens would not have been satisfactory for assessing the nuclear grade, mitotic rate, and DCIS pattern. IHC was also found to be unreliable on the CAL specimens.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cryosurgery technology has vastly improved over the past decade, progressing from devices almost as large as small cars to the current countertop model. Extensive experience with ultrasonography-guided cryoablation (tumor destruction by means of ultracold temperatures) has been described previously to be efficacious in the treatment of prostate and hepatic malignancies.16,17 The technology is now being used by breast surgeons to noninvasively treat benign fibroadenomas.18 Ultrasonography-guided cryoprobe-assisted hepatic resection of liver malignancies was the first organ procedure described in which the template formed by a created ice ball was used to advantage to clear margins that would otherwise be difficult or impossible to clear.19
Despite long-term results from randomized trials documenting success with breast-conserving therapy,20,21 the procedure remains inexact, particularly when the lesion is not palpable or has poorly defined borders, the surrounding tissue is excessively fatty or dense, or the breasts are very small and conserving normal breast tissue is thus a priority.
A number of new techniques have been described to replace needle localization breast-conserving therapy. These have included creating a small ultrasonographically visible autologous clot at the time of LCNB,22 use of intraoperative ultrasonography,23–26 and use of radionucleotide and gamma probe27 localization. The most promising is the use of intraoperative ultrasonography. A small, randomized trial of wire-guided resection versus ultrasonography-guided lumpectomy showed improvement in the rate of clean margins, from 55% to 89% (P = .007), respectively.28
Ultrasonography by itself, however, neither provides a dissection template that can be tailored intraoperatively to the size of the tumor nor creates a palpable mass from a nonpalpable tumor. By creating an ice ball that engulfs the tumor and a predetermined margin, CAL converts a nonpalpable lesion to a well-defined and palpable one. The expected advantages of performing such a procedure under ultrasonographic guidance are the ability to monitor the ice ball while it grows out from the center of the tumor and the ability to control the size of the adjacent margin of normal breast tissue. The ice ball then serves as a firm template around which the surgeon can dissect.
The precision of CAL was impressive; the ability to "dial up" the amount of tissue needing removal adds a dimension of control to breast surgery not previously realized with any other technology. Once the desired margin of ice was achieved beyond the tumor borders, reducing the system’s setting to low maintained the ice ball with little growth while lumpectomy was performed. Clear visualization of the ice ball engulfing the nonpalpable tumor plus a predetermined amount of normal-appearing surrounding tissue allowed rapid dissection with minimal concern for whether an adequate margin was being obtained. Furthermore, the firm ice ball was much easier to handle than the typically fatty breast tissue, and it was not difficult to distinguish between fibrous breast tissue and an area involved with tumor, as can occur with some frequency during standard lumpectomy.
Surgeons’ use of and comfort level with breast ultrasonography, particularly to guide minimally invasive procedures, are rising dramatically.29–32 In our experience, the learning curve for CAL was short for surgeons who perform ultrasound-guided LCNB regularly.
Of critical importance to the success of the procedure is placement of the cryoprobe through the center of the lesion. The tumors were approached by placing the long axis of the cryoprobe through the longest diameter of the cancer. This is because the forming ice ball is somewhat elliptical in shape. The cryoprobe vacuum insulation, in conjunction with continuous ultrasonographic monitoring and the ability to intermittently inject sterile saline between the skin and the ice ball, provided a high degree of control. There is less concern about protecting the pectoralis muscle, because of the cold sink provided by its substantial blood flow and the reports of previous experience with cryoablation of fibroadenomas.18
Typical positive margin rates for needle localization procedures are reported in the literature to range between 40% and 60%.9–15 Re-excision of an incompletely cleared breast cancer can be traumatic for patients. The overall positive margin rate for this study was 16.7%, but among patients who had more than 6 mm of ice created around their tumor before excision (n = 18), this rate fell to 5.6%, a statistically significant difference. A relatively short margin distance was chosen for the first half of the trial in an effort to minimize the time of the procedure and of tissue exposure to ultracold temperatures. As the study progressed, it became clear that creating an ice ball that engulfed the tumor plus a 5-mm margin was not sufficient to clear local disease; thus, the targeted margin size was increased to 5–10 mm in the second half of the study.
In this small series, factors contributing to the positive margin rate are difficult to determine, and it is therefore not surprising that no differences were statistically significant (Table 2). From these preliminary findings, however, the ice ball margin appears to be the most important factor. Further studies will be required to determine the specific ice ball margin that best serves specific tumor types, as well as other contributing factors that will lead to the lowest positive margin rates.
Others have considered improved techniques for reducing the positive margin rate at initial lumpectomy. Two series have been reported in which intra-operative ultrasonography was used to guide surgical excision of nonpalpable breast cancers.23,24 In one series, 17 of 20 malignancies (85%) were excised at first operation with a margin of at least 1 mm.23 Sixty-five cancers were excised with ultrasound guidance in the second series.24 Mean specimen excision size was 6.5 cm (± 1.5 cm) x 5.4 cm (± 1.3 cm) x 3.0 cm (± 1.1 cm). Mean maximum tumor diameter was 1.1 cm (range, .4–2.8 cm). Three patients (4.6%) required a second operation, two for DCIS-involved margins and one for a second ultrasound-occult tumor within 1 mm of the resected specimen. Although the positive margin rate was similar in our study, the ratio of mean specimen size to mean tumor size suggesting that less tissue was excised with the CAL procedure.
The cryosurgical procedure itself (perhaps in combination with specimen thawing) caused significant nuclear distortion on hematoxylin and eosin staining and inaccurate IHC results. Full characterization of the tumor histological features and receptor status by LCNB is required in advance of any CAL case. Most important, measuring tumor size or assessing whether the margins were clear was not an issue in any case. Intuitively, this makes sense, because the specimen resection line remains in unfrozen tissue adjacent to the ice ball.
The results of this study suggest the need for controlled trials comparing standard needle localization lumpectomy with intraoperative ultrasonography and CAL. Procedure time, cost-savings (including a decrease in re-excisions), specimen size, cosmetic result, and margin status could all be factors determining the usefulness of this new technology. It is clear, however, that both patients and surgeons are desirous of matching the extent of surgical resection with the extent of disease at the first operation, and CAL provides this precision.
|
ACKNOWLEDGMENTS |
|---|
The authors thank Lisa Tagliareni for manuscript preparation and Sanarus Medical, Inc. (especially Dr. Seth Stabinsky) for generous support of the project.
The acknowledgments are available online at www.annalssurgicaloncology.org.
|
FOOTNOTES |
|---|
Received for publication April 1, 2003. Accepted for publication July 16, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Koomen, E. D. Pisano, C. Kuzmiak, D. Pavic, and R. McLelland Future Directions in Breast Imaging J. Clin. Oncol., March 10, 2005; 23(8): 1674 - 1677. [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Simmons Freezing Breast Cancers to Enhance Complete Resection Ann. Surg. Oncol., November 1, 2003; 10(9): 999 - 999. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
