| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, University of Vienna, Austria, Waehringer Guertel 18-20, A-1090 Vienna, Austria.
Received May 3, 2001;
accepted after revision November 19, 2001.
P. Wunderbaldinger is supported by the Erwin
Schroedinger-Auslandsstipendium of the Austrian Science Fund (FWF), Vienna,
Austria.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Two hundred patients had large-core breast biopsy performed either in the prone (n = 100) or in the sitting (n = 100) position and subsequently underwent surgical resection. The histopathologic findings of large-core breast biopsy and surgery of all 200 patients were compared; sensitivity, specificity, and accuracy were calculated for both groups. Biopsy-associated complications were prospectively recorded for immediate and delayed events and for technical failures in both groups.
RESULTS. Sensitivity (96%), specificity (100%), and accuracy (98%) were the same for both groups with two false-negative findings in each group. The two false-negative results in the sitting group were caused by vasovagal reactions, whereas those in the prone group were caused by technical failure and uncomfortable biopsy position. More statistically significant complications (seven vs four, p < 0.001) and vasovagal reactions (seven vs two, p < 0.0001) were seen in the sitting group.
CONCLUSION. For performance of large-core breast biopsy, both the prone and sitting positions are reliable and accurate methods. However, vasovagal reactions that could potentially complicate biopsy were seen significantly more often in the sitting position.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The purpose this study was to compare the two different body positions for large-core breast biopsy with regard to sensitivity, specificity, and accuracy, as well as complication rate.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
During the study period, stereotactic large-core breast biopsy was offered as an alternative to surgical biopsy for the diagnosis of indeterminate or suspicious lesions that were evident mammographically, unless the lesion measured less than 5 mm in maximum diameter or the patient had a bleeding diathesis, was unable to cooperate with the procedure, had inadequate thickness of compressed breast parenchyma to accommodate the insertion of the needle, or presented with more than one lesion.
The included 200 surgically verified lesions were in 200 women who ranged from 33 to 87 years old (median age, 60 years old). All lesions were prospectively classified by the radiologist reviewing the mammograms as noncalcified mass (n = 118, 59%), mass with calcification (n = 24, 12%), and calcifications without mass (n = 58, 29%) (Table 1). These lesions were equally distributed in both biopsy groups (sitting vs prone); thus, the sample was suited for paired analysis. The range of mammographic lesion size in these 200 lesions was 0.5-5.6 cm (median size, 2.6 cm) (Table 1).
|
The biopsy protocol at the time of the study was to obtain a minimum of six tissue specimens per lesion (both masses and calcifications). The number of specimens actually obtained was prospectively recorded, and the reasons for obtaining less than six specimens were recorded for each lesion. Additional specimens were obtained at the discretion of the radiologist who performed the biopsy on the basis of a prospective review of stereotactic images, visual inspection of tissue specimens, and review of specimen radiographs in lesions evident as calcifications.
The pattern of obtaining core biopsy samples depended on the morphologic characteristics of the lesion. In masses, the initial core biopsy sample was obtained in the center, and accurate needle placement was documented by images obtained before and after needle insertion. Subsequent passes were made by moving the biopsy needle "around the clock" in 2-hr increments, starting at the 12-o'clock position, without obtaining additional images. The distance from each pass to the center of the lesion depended on the lesion size. For calcifications, there was more variability of needle placement and number of images. The intent was always to target the most suspicious area for needle biopsy; the first pass was always intended to target the center of a tight cluster of calcifications or the most suspicious area of calcifications. Subsequent passes were planned according to the geography of the calcifications.
For lesions evident as calcifications, specimens were placed on a piece of film and irrigated with saline, and a radiograph was obtained using standard imaging parameters [23]. Stereotactic images were obtained after biopsy in all 200 patients immediately after the stereotactic large-core breast biopsy. These images were prospectively reviewed to determine whether the mammographically evident lesion had been removed completely (no residual calcification or absence of the targeted mass) [24].
Reasons to perform surgery in these 200 lesions included the large-core breast biopsy histologic results, preferences of the patients or referring surgeons, and recommendations of the radiologist.
Histologic specimens were prospectively categorized by a pathologist for both the large-core breast biopsy and open surgical biopsy as either malignant (invasive carcinoma, ductal carcinoma in situ, or metastasis), high risk (atypical ductal hyperplasia, lobular carcinoma in situ, radial scar, phyllodes tumor), or benign (anything that is not malignant or high risk). Biopsy and surgical specimens were independently interpreted by a pathologist who was not aware of the histology of the large-core breast biopsy when he reviewed the surgical specimens. The final histologic diagnosis from needle biopsy and surgery in all 200 patients was then prospectively compared to evaluate agreement.
Complications in both groups were prospectively recorded for immediate events (vasovagal reactions and major bleeding requiring more than manual compression) and delayed events (infections requiring treatment or hematomas requiring more than manual compression). Data for delayed events were derived from a delayed review of patients' charts. In addition, technical failures and early termination of the needle biopsy caused by uncomfortable biopsy position were noted. A vasovagal reaction was defined as any form of nausea, vertigo, or collapse. Immediately after the biopsy, all the studied variables were recorded in the patient's file by the radiologist performing the biopsy. In addition, we compared the complication rates associated with large-core breast biopsy in our study with those reported in the literature.
All statistical calculations were made using SAS software (SAS Institute, Cary, NC). Results are expressed as mean (SD). Group comparisons were performed using the Student's t test or the Mann-Whitney U test as appropriate. Frequency data were analyzed with the chi-square test or Fisher's exact test for small samples. Sensitivity, specificity, and accuracy for both large-core breast biopsy groups were calculated and compared. A p value greater than 0.05 was considered significant, and all p values were corrected for ties, where applicable.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Reasons to perform surgery included histology that showed a malignant or high-risk lesion (n = 99, 49.5%), fibroadenomas larger than 1.5 cm in diameter (n = 23, 11.5%), personal preference of the patient (n = 45, 22.5%) or of the referring surgeon (n = 29, 14.5%) regardless of the biopsy result, and recommendation by the radiologist (n = 4, 2%).
Two false-negative findings in each group resulted in a sensitivity of 96% (48/50 for the sitting position and 46/48 for the prone position), a specificity of 100% (50 true-negative results and no false-positive results for the sitting position and 52 true-negative results and no false-positive results for the prone position), and an accuracy of 98% (98/100 for both positions). There was no statistically significant difference (p = 0.9) in performance between large-core breast biopsy using the sitting approach versus large-core breast biopsy using the prone approach.
In two patients in the sitting group, early biopsy termination was caused by vasovagal reactions (after two and three specimens), and in one patient in the prone group, by a technical failure (after four specimens). Open surgical biopsy had been recommended to all of these three patients because of the incompleteness of the large-core breast biopsy procedure. All three findings with acquisition of fewer than five specimens or early termination were false-negative. The other false-negative finding in the prone position group presented mammographically as a dense breast with spread out microcalcifications. We harvested 10 specimens in this patient, and in eight of them, specimen radiography showed the targeted calcifications. Because of lingering uncertainty, we intended to harvest more tissue, but the patient rejected this option because of her inability to remain in a prone position (caused by neck pain). Thus, despite a benign histologic diagnosis (fibrocystic changes with benign microcalcifications), open surgical biopsy was recommended by the radiologist and revealed ductal carcinoma in situ in this patient.
There were 24 other patients with five specimens per lesion.
Results for complications associated with large-core breast biopsy are displayed in Table 3.
|
Although the rate of early biopsy terminations (n = 2) was the same for both biopsy groups (p < 0.001), more statistically significant complications occurred overall in the sitting position (seven in the sitting vs four in the prone biopsy group).
Vasovagal reactions were seen in nine patients (4.5%) and occurred significantly more often (p < 0.0001) in the sitting position (seven in the sitting vs two in the prone biopsy group). In two patients (1%) in the sitting biopsy group, vasovagal reactions (severe nausea, collapse) led to early termination of the biopsy procedure, resulting in a false-negative diagnosis.
No hematoma or infection requiring treatment and no major bleeding were seen in either biopsy group.
None of the 200 lesions was completely removed by large-core breast biopsy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
However, more statistically significant vasovagal reactions occurred in the sitting biopsy group (seven vs two, p < 0.0001). Thus, our data support the general opinion that largecore breast biopsy performed with the patient in the prone position (using a dedicated system) is better tolerated by patients than largecore breast biopsy performed with the patient in the sitting position (using an add-on system) [1,2,3, 5, 9, 11,12,13,14,15,16,17]. Two major explanations have been proposed for the higher frequency of vasovagal reactions during the sitting position: first, the biopsy is performed in full view of the patient, which potentially leads to increased anxiety levels. Second, the sitting position might become uncomfortable with increased duration of the procedure [9, 17, 22]. On the basis of our data, we agree with the first statement because all seven patients in the sitting group with vasovagal reactions stated that these reactions were triggered by watching the biopsy procedure. Although the only case of patient discomfort (neck pain) leading to early biopsy termination was seen in the prone group, we agree with others that a prone table generally allows patients to lie more comfortably and, furthermore, might act as a psychologic barrier between patient and biopsy, reducing the risk of sudden movement and vasovagal reactions [1, 9, 14, 17, 22].
To validate our data, we have compared our complication rate with those reported in the literature [1, 3, 5, 12, 14,15,16,17,18, 21, 25,26,27,28,29,30] (Table 4). In accordance with previous studies, no delayed complications, such as hematoma requiring more than local compression or infection requiring treatment, were seen in either biopsy group [1, 3, 4, 14, 16, 18]. However, more vasovagal reactions occurred in our study. These seem to be explained by the definition of vasovagal reactions (nausea, vertigo, collapse) used in our study. However, if only the two cases that led to early termination of the biopsy procedure are taken into account, our vasovagal reaction rate is within the published range [2, 5, 6, 8,9,10, 12, 14, 17,18,19,20,21, 30,31,32,33].
|
General disadvantages of prone systems are their cost (approximately twice that of add-on systems), their considerable space requirements (approximately quadruple that of a regular unit), and their limited use (only to perform breast biopsies and thus sometimes underused) [1, 9, 13, 17, 31]. Because of these disadvantages, many breast imaging centers might have difficulty justifying the use of dedicated prone units, considering the small number of biopsies performed [17, 31]. Thus, for these institutions, an add-on system might improve overall equipment use and cost-effectiveness. In addition, patients unable to tolerate the prone position (e.g., older patients or patients with spine deformities) could undergo large-core breast biopsy on the add-on upright unit; furthermore, and no weight restrictions apply for add-on units, whereas the maximum weight for dedicated tables is approximately 300 lb (135 kg) [1, 15].
There are some limitations to our study. In 27 lesions, less than six specimens were obtained in two lesions because of vasovagal complications as already described and in one patient due to technical failure. In the remaining 24 patients, biopsy was completed after five specimens were obtained by the performing radiologist either because the gross appearance of the specimens was considered to be "probably adequate and diagnostic," as described by Berg et al. [25], or because the targeted calcifications were seen on the specimen radiographs [6, 20, 23, 25]. Although these findings might be helpful for fast assessment of harvested specimens, we agree with others that to establish an exact diagnosis, accurate sampling is absolutely necessary but cannot be assumed by visual inspection of the specimens nor by identification of calcifications on specimen radiographs [6, 20, 23, 25]. But to achieve sufficient diagnostic yield, a minimum of five specimens is required for large-core breast biopsy of masses, calcified masses, and calcifications with positive findings on specimen radiographs [6]. Currently, we harvest more than five specimens (large-core breast biopsy and directional vacuum-assisted breast biopsy) because we agree with the concept that more tissue clearly improves the diagnostic yield of breast biopsy [4, 21, 26, 27, 29, 33].
The inclusion in our study of the 24 patients with five specimens per lesion might be seen as controversial and incongruous with the study protocol. However, these patients were equally distributed in both biopsy groups, and there was agreement between large-core breast biopsy and open surgical biopsy in all patients. Thus, neither the results nor the statistical data of our study were changed.
The selection of the patient cohort (only surgically verified patients were included) might be seen as another limitation. Because we do not know whether any malignancy was missed in the non-surgically verified patients, we might have overestimated the sensitivity, specificity, and accuracy of both biopsy approaches. However, our false-negative rate of 2% is comparable to that of other published large-core breast biopsy studies with false-negative rates ranging from 1.5% to 2% [2, 5, 6, 8,9,10, 12, 14, 17,18,19,20,21, 30,31,32,33].
The high number of surgically verified benign lesions can be explained by a lack of acceptance as well as general doubts and prejudices of surgeons and patients toward imaging-guided biopsy methods in our country, regardless of the numerous publications that describe large-core breast biopsy as a reliable and efficient alternative to open surgical biopsy [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Furthermore, our surgeons prefer to surgically excise biopsy-proven fibroadenomas larger than 1.5 cm in diameter, even though we are unaware of increased risk of malignancy in such lesions. Thus, further studies with surgically verified cases or cases with adequate follow-up (>24 months) are necessary to increase acceptance of and trust in imaging-guided biopsy methods.
Directional vacuum-assisted breast biopsy has been advocated as the most accurate breast biopsy technique (especially in the evaluation of calcified lesions) [4, 9, 17, 21, 26,27, 29, 30] and would have been a valuable addition to the study, particularly because little data exist on directional vacuum-assisted breast biopsy performed with patients in the sitting position. Welle et al. [17] reported their initial experience with the use of directional vacuum-assisted breast biopsy in combination with a dedicated reclinable mammography chair for the decubitus biopsy position for add-on upright equipment. They and others encountered no complications and suggest that directional vacuum-assisted breast biopsy equipment seems to function equally well when mounted on add-on units [17].
In conclusion, large-core breast biopsy performed with patients in the prone and in sitting positions is a reliable and accurate method. Vasovagal reactions that could potentially complicate breast biopsy were seen significantly more often with patients in the sitting position. However, if large-core breast biopsy is performed only occasionally, systems with the patient in the sitting position seem to be more cost-effective.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. C. Riedl, G. Pfarl, M. Memarsadeghi, T. Wagner, F. Fitzal, M. Rudas, and T. H. Helbich Lesion Miss Rates and False-Negative Rates for 1115 Consecutive Cases of Stereotactically Guided Needle-localized Open Breast Biopsy with Long-term Follow-up Radiology, December 1, 2005; 237(3): 847 - 853. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Koskela, M. Sudah, M. H. Berg, V. J. Karja, P. K. Mustonen, V. Kataja, and R. S. Vanninen Add-on Device for Stereotactic Core-Needle Breast Biopsy: How Many Biopsy Specimens Are Needed for a Reliable Diagnosis? Radiology, September 1, 2005; 236(3): 801 - 809. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
