| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Indiana University School of Medicine, Indiana
University Hospital, 550 N. University Blvd., Rm. 0279, Indianapolis, IN
46202.
2
Department of Surgery, Indiana University School of Medicine, Indianapolis, IN
46202.
Received June 2, 2000;
accepted after revision July 11, 2000.
Address correspondence to H. E. Reynolds.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Retrospective review revealed 119 women with breast cancer who underwent lymphoscintigraphy and sentinel node biopsy at our institution during the study period. A planned axillary dissection was performed in 13 of the first 16 patients; otherwise, axillary dissection was only performed if tumor was present in sentinel nodes or if the sentinel node biopsy was unsuccessful.
RESULTS. Sentinel node biopsy was successful in 96% of patients, and sentinel node metastases were found in 20%. In 78% of patients, a negative sentinel node biopsy obviated axillary dissection. Prior excisional biopsy was not associated with a failed sentinel node biopsy (p = 0.750) but was associated with failed lymphoscintigraphy (p = 0.01). Successful lymphoscintigraphy was associated with successful sentinel node biopsy (p < 0.0001). No association was found between the histology or size of the tumor and a failed sentinel node biopsy (p = 0.46 and p = 0.1, respectively) or failed lymphoscintigraphy (p = 0.36 and p = 0.47, respectively).
CONCLUSION. Sentinel node biopsy guided by lymphoscintigraphy, intraoperative gamma probe, and isosulfan blue dye is an effective alternative to axillary dissection in patients with breast cancer. Lymphoscintigraphy improved the success rate of sentinel node biopsy. Large tumor size or prior excisional biopsy should not prevent patients from having sentinel node biopsy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The purpose of this study was to determine the technical success rate of sentinel node biopsy with lymphoscintigraphic mapping in women with clinically node-negative breast cancer and to determine the frequency with which sentinel node biopsy obviated axillary dissection. Particular focus was on the association of prior surgical biopsy, tumor histology or tumor size, and failed sentinel node biopsy or failed presurgical lymphoscintigraphic mapping. The association between positive nodal disease and the size and histology of the tumor was also evaluated to better establish patient selection guidelines for sentinel node biopsy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In 13 of the first 16 patients, a conventional axillary dissection was performed and correlated with the findings in the sentinel node as a training set to validate the sentinel node biopsy technique. In the subsequent 103 women, the sentinel nodes were evaluated by intraoperative frozen section. If frozen-section analysis revealed axillary metastases, the patient underwent conventional axillary dissection. Axillary dissection was also performed in those patients in whom identification of the sentinel node was unsuccessful on surgery.
The 119 patients were injected with filtered 99mTc sulfur colloid. A dose of 1.0 mCi was divided into four syringes and injected into the breast parenchyma around the tumor site. Lymphoscintigraphy was performed with a large field of view dual-head gamma camera (Apex SP6; Elscint, Haifa, Israel) with a low-energy high-resolution collimator. Images were obtained in the anterior and lateral projections immediately, 15 min, and 30 min after injection. If the sentinel node was not identified on these initial images, then an additional image was obtained at 60 min. Lymphoscintigraphic mapping was considered successful if we saw a focal area of radiotracer uptake separate from the injection site on the lymphoscintigraphic images. Immediately after the final lymphoscintigraphic image was obtained, the patient was transported to the operating room.
An intraoperative gamma probe (neo2000; Neoprobe, Dublin, OH) was used to obtain background counts, counts from the liver, counts from the breast injection site, and counts from the axilla. The region in the axilla with the highest counts was marked on the patient's skin. Intraoperatively, 5 mL of isosulfan blue dye (Lymphazurin; Zenith Parenterals, Rosemont, IL) was injected around the tumor site. At 5 min, an axillary incision was made, and a sentinel node biopsy was performed. Both the gamma probe and the appearance of isosulfan blue dye in lymphatic channels and nodes guided the biopsy. A node was considered a sentinel node if it exhibited blue staining at surgery or if its removal resulted in a decrease in axillary counts, or both. All sentinel nodes were excised.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Histologic findings in the 120 cancers were infiltrating ductal carcinoma in 88 (73%), which included four patients with ductal carcinoma in situ with microinvasion; infiltrating lobular carcinoma in 13 (11%); mixed infiltrating ductal and infiltrating lobular carcinoma in four (3%); mucinous in one (1%); medullary in one (1%); tubular in two (2%); tubulolobular in two (2%); and ductal carcinoma in situ in nine (8%). The nine patients with ductal carcinoma in situ were included primarily in the latter half of our study to further evaluate the efficacy of sentinel node biopsy in patients with ductal carcinoma in situ. There were no specific criteria for selecting the women with ductal carcinoma in situ who would undergo sentinel node biopsy, but findings in the nine patients selected were either high-grade comedo-type, multifocal, extensive ductal carcinoma in situ, or a combination of these diseases. The procedure of sentinel lymph node biopsy in patients with ductal carcinoma in situ was abandoned after these nine patients.
Biopsy of Sentinel Nodes
In 13 of the initial 16 attempted sentinel node biopsies performed, a
planned conventional axillary dissection was also performed. Pathologic
specimens were examined with a single H and E section, and findings in the
sentinel nodes were correlated with the remaining axillary nodes. Surgical
identification of sentinel nodes was successful in all 16 patients. A total of
43 sentinel nodes were found, and a total of 174 nonsentinel nodes were
removed. In five (31%) of the 16 women, the sentinel nodes contained
metastases. In these five patients, tumor was found in nine (60%) of 15
sentinel nodes removed. Axillary dissection performed in four women with tumor
in sentinel nodes revealed additional metastases in three women (18 positive
findings on 52 nodes removed) and no additional metastases in one woman. In
one patient with endstage renal disease, two of five sentinel nodes removed
contained tumor, but because of the patient's comorbid disease, an axillary
dissection was not performed. Sentinel nodes were free of tumor in 11 (69%) of
the 16 women. Axillary dissection was performed in nine of these 11 women and
revealed no additional metastases. Accuracy of 100% (13/13), specificity of
100%, sensitivity of 100%, and a false-negative rate of 0% were calculated
with only the 13 patients in whom a conventional axillary dissection was
planned before surgery.
Ninety-nine (95%) of the subsequent 104 attempted sentinel node biopsies were technically successful with sentinel nodes retrieved at surgery. A total of 199 sentinel nodes were found, and a total of 510 nonsentinel nodes were removed. Pathologic specimens were examined with a single H and E section. Immunohistochemistry was performed on 26 sentinel nodes retrieved in a subset of 13 of the 99 successful sentinel node biopsies. Findings of immunohistochemistry were negative in all 13 biopsies, and nonsentinel nodes were excluded from immunohistochemistry examination.
Sentinel nodes were free of tumor in 81 (82%) of 99 patients. In 81 (78%) of the 104 patients, a negative sentinel node biopsy obviated axillary dissection. In the 18 patients with sentinel node metastases, 28 (70%) of 40 sentinel nodes removed contained tumor. Axillary dissection was performed in 17 of these 18 patients, and in nine (53%) of them, additional axillary metastases were found. Ninety-six (97%) of the 99 successful sentinel node biopsies had intraoperative frozen-section analysis. The sentinel node findings confirmed on permanent sections stained with H and E were positive in 15 (16%) and negative in 81. There were no cases of a false-negative or false-positive findings on intraoperative frozen-section analysis.
In five (5%) of the 104 attempted sentinel node biopsies, no sentinel nodes were identified at surgery despite the use of the gamma probe and isosulfan blue dye. Lymphoscintigraphic mapping was unsuccessful in four of the five patients. Axillary dissection was performed in all five patients with 129 axillary nodes removed. Axillary metastases were found in two patients. In one patient, 19 of 32 nodes contained tumor, and in the second patient, one of 23 nodes contained tumor.
In 115 (96%) of 120 attempted sentinel node biopsies, sentinel nodes were identified at surgery. A total of 242 sentinel nodes were removed at surgery (range, 1-6 nodes; median, 2.00 nodes; mean, 2.10 nodes). In 23 (20%) of the 115 patients, 37 (15%) of 242 sentinel nodes contained metastases (range, 1-6 nodes; median, 1.00 nodes; mean, 1.61 nodes). Axillary dissection was performed in 21 of these patients, and additional metastases were found in 12 (57%) of the 21 patients. All patients with metastases in nonsentinel nodes also had metastases in the sentinel nodes. There was no association between the histology and size of the tumor and a failed sentinel node biopsy (p = 0.46 and p = 0.1, respectively) (Table 1).
|
Of the 115 patients, 40 had surgical excisional biopsies, 74 had nonsurgical biopsies or no biopsy, and in one patient, the type of biopsy was unknown. Of the five patients with unsuccessful sentinel node biopsies, two had excisional biopsies (Table 1). A prior surgical excisional biopsy was not significantly associated with a lower success rate of sentinel node biopsy (p = 0.750).
Considering the 21 patients with tumor in sentinel nodes who also underwent axillary dissection, we analyzed the data in subgroups on the basis of tumor size and histology. Of the nine patients with positive findings on sentinel nodes only, none had Tis-T1a tumors, four had T1b tumors, four had T1c tumors, and one had a T2 tumor. Of the 12 patients with tumor found in both sentinel and nonsentinel nodes, none had Tis-T1b tumors, one had a T1c tumor, and 11 had T2-T4 tumors. A significant association existed between the size of the tumor and the likelihood of finding tumor in nonsentinel nodes (p < 0.0001).
Twenty-five patients had tumor in either sentinel or nonsentinel nodes. The histology in these patients was infiltrating ductal carcinoma in 22, infiltrating lobular carcinoma in two, and mixed infiltrating ductal and infiltrating lobular carcinoma in one. No nodal metastases were found in the six patients with tumors of more favorable histologic type (tubular, tubulolobular, mucinous, or medullary), in the nine patients with ductal carcinoma in situ, or in the four patients with ductal carcinoma in situ with microinvasion (p = 0.001).
Lymphoscintigraphy
Preoperative lymphoscintigraphic mapping was considered successful in 106
(88%) of the 120 attempted mapping procedures. Sentinel nodes were
successfully identified at surgery in 105 of these 106 women. A total of 215
sentinel nodes were removed at surgery (range, 1-6 nodes; median, 2.00 nodes;
mean, 2.05 nodes). In one patient, lymphoscintigraphic mapping was considered
successful, but no sentinel node was identified at surgery. At conventional
axillary dissection, one node contained metastases and 23 nodes were removed.
No association was found between histology and size of the tumor and failed
lymphoscintigraphy (p = 0.36 and p = 0.47, respectively)
(Table 2).
|
Preoperative lymphoscintigraphic mapping was unsuccessful in 14 (12%) of the 120 attempted procedures. Sentinel nodes were identified at surgery in 10 (71%) of these 14 women. In these 10 patients, only the gamma probe identified the sentinel nodes in three; only the isosulfan blue dye, in three; and a combination of the gamma probe and isosulfan blue dye, in four. A total of 27 sentinel nodes were removed (range, 1-6 nodes; median, 2.00 nodes; mean, 2.70 nodes). In two of the 10 women, metastases were found in seven sentinel nodes (26%). In the first patient, a single sentinel node identified only by gamma probe contained tumor, and 32 of 36 additional axillary nodes had positive findings for metastases. In the second patient, six sentinel nodes were identified by isosulfan blue dye only. All six sentinel nodes and 24 of 25 additional axillary nodes contained tumor.
Sentinel nodes were not identified at surgery in four (29%) of the 14 cases of unsuccessful lymphoscintigraphic mapping. Conventional axillary dissections were performed, and a total of 106 lymph nodes were removed (range, 14-32 nodes; median, 30.00 nodes; mean, 26.50 nodes). In three women, no axillary metastases were found. In one woman, 19 of 32 nodes removed contained metastases. A positive finding on lymphoscintigraphy was significantly associated with the successful identification of a sentinel node at surgery (p < 0.0001).
Of the 106 patients with successful lymphoscintigraphic mapping, 34 had excisional biopsies, 69 had nonsurgical biopsies, two had no biopsy, and in one patient, the type of biopsy was unknown. Of the 14 patients with unsuccessful lymphoscintigraphic mapping, eight had excisional biopsies, and six had nonsurgical biopsies (Table 2). A statistically significant decrease occurred in the success rate of lymphoscintigraphic mapping in patients who had undergone prior surgical biopsy (p = 0.01)
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Twenty percent of the patients in whom sentinel nodes were identified had sentinel node metastases. This is lower than that observed by other investigators who report a range of 23-45.3% for sentinel node metastases [5,6,7, 9,10,11]. This finding was partly due to the inclusion of nine (8%) patients with ductal carcinoma in situ in our study. If these nine patients were excluded, the percentage of patients with sentinel node metastases would increase to 21.5%. Fifty-seven percent of women with sentinel node metastases had additional nonsentinel node metastases at axillary dissection. This result confirmed the findings of previous investigators who report a range of 39-62% for additional nonsentinel node metastases and also supported the sentinel lymph node hypothesis [5,6,7,8,9]. The accuracy of sentinel node biopsy in our study was 13 (100%) of 13 tumors, with no false-negative findings. Previous studies have reported accuracy rates for sentinel node biopsy of 95-97% [5, 6, 8,9,10].
Intraoperative frozen-section analysis was performed in 96 sentinel node biopsies. The overall concordance of frozen-section analysis and permanent sections was 96 (100%) of 96. The sentinel node findings were negative in 81 of 96, with no false-negative or false-positive findings. Other authors have reported a higher false-negative rate of 13-36% for frozen-section analysis [7, 11, 13, 14]. In 81 (78%) of 104 patients in our study, a negative finding at sentinel node biopsy spared the patient an axillary dissection. Liberman et al. [11] reported that sentinel node biopsy obviated axillary dissection in 70% of patients in their study, and Borgstein et al. [15] found that sentinel node biopsy alone would have been sufficient in 75 (62%) of 122 of their patients.
It has been hypothesized that prior excisional biopsy may lower the success rate of sentinel node biopsy and lymphoscintigraphic mapping by disrupting the lymphatic pathways that drain the tumor [11, 15, 16]. In our study, the lymphoscintigraphic failure rate was significantly higher in patients who had undergone previous excisional biopsy (p = 0.01). However, no decrease existed in the success rate of sentinel node biopsy in women with prior excisional biopsy compared with women who underwent nonsurgical biopsy or no biopsy (p = 0.750). Hill et al. [7], O'Hea et al. [9], and Cody [17] also found no association between prior surgical biopsy and failed sentinel node biopsy.
Presurgical lymphoscintigraphy was considered successful in 106 (88%) of 120 attempted mapping procedures. Variable success rates of lymphoscintigraphy have been reported by other investigators including 36% by Liberman et al. [11], 40% by Hill et al. [7], 75% by O'Hea et al. [9], and 98.7% by Veronesi et al. [5]. In our study, there was a positive association between technically successful presurgical lymphoscintigraphic mapping and successfully identifying sentinel nodes at surgery (p < 0.001). Hill et al. also report a statistically significant association between a positive result on a lymphoscintigram and successful localization of a sentinel node at surgery (p = 0.0001).
Linehan et al. [12] concluded that routine presurgical lymphoscintigraphy was not valuable in sentinel node biopsy. In their study, there were no cases in which lymphoscintigraphy showed an unanticipated pattern of lymphatic drainage, such as to internal mammary nodes, and there were no cases in which the surgical approach for sentinel node biopsy was changed as a result of lymphoscintigraphy. In addition, Linehan et al. believe that if there is an aberrant pattern of lymphatic drainage, this pattern can also be readily identified by the use of the gamma probe intraoperatively without the guidance of lymphoscintigraphy. These researchers report a lower success rate for lymphoscintigraphy (52% with filtered 99mTc sulfur colloid and 65% with unfiltered 99mTc sulfur colloid) than that found in our study. They conclude that at their institution, operative radiolocalization, but not lymphoscintigraphy, is valuable in guiding sentinel node biopsy by successfully locating sentinel nodes in more than half the patients with negative findings on lymphoscintigraphy.
The association between successful lymphoscintigraphy and locating sentinel nodes at surgery in our study may be the result of the technique used for lymphoscintigraphy. Images were obtained immediately, at 15 min, and at 30 min after injection. If migration of the radioisotope was observed on any of these initial images, we took the patient to surgery immediately, and no follow-up images were obtained.
If the sentinel node was not identified on these initial images, then a delayed 60-min image was obtained. This sequence allowed the patient to be taken to surgery immediately after migration of the radioisotope occurred and decreased the risk of a diffusely "hot" axillary bed in which sentinel nodes cannot be distinguished from nonsentinel nodes. Presurgical lymphoscintigraphy may be beneficial by identifying the optimal interval between isotope injection and surgery. If normal migration of the radioisotope had occurred on the initial images, the patient was taken to surgery. If not, delayed images could be obtained until migration was verified.
Borgstein et al. [15] report that presurgical lymphoscintigraphy can provide useful information before biopsy by showing the sufficient uptake of the radiocolloid by sentinel nodes to guarantee successful surgical identification of the correct sentinel node. These researchers also report a significant decrease in the success rate of lymphoscintigraphy after excisional biopsy (chi-square test; p < 0.005), but sentinel nodes are often still identified at surgery with intraoperative gamma probe or isosulfan blue dye. Our experience was similar. We found a statistically significant decrease in the success rate of lymphoscintigraphic mapping in women who had undergone excisional biopsies (p = 0.01). In the 14 (12%) patients in whom lymphoscintigraphy was unsuccessful, eight had undergone prior excisional biopsies. Of these 14 patients, sentinel nodes were identified at surgery in 10 (71%) either by gamma probe alone (3/10), isosulfan blue dye alone (3/10), or a combination of both (4/10). The high success rate of lymphoscintigraphy in our study and its role in successful sentinel node biopsy support a role for routine lymphoscintigraphy before sentinel node biopsy in patients with breast cancer.
Currently, the status of axillary lymph node involvement in women with metastatic breast cancer remains the most important prognostic determinant [2]. Small impalpable percutaneously diagnosed breast cancers would appear to be ideally suited for sentinel node biopsy because of the low presurgical probability for axillary metastases. In our study, we found a significant association between the size of the tumor and the likelihood of finding tumor in nonsentinel nodes in addition to sentinel nodes. Of the 21 patients with tumor in sentinel nodes who also underwent axillary dissection, nine had positive sentinel nodes only. These nine patients had no tumors T1a or smaller and no tumors larger than T2. Of the 12 patients with tumor in both sentinel and nonsentinel nodes, one had a T1c tumor and 11 had T2-T4 tumors. No tumors were smaller than T1c.
In addition, we found a statistically significant association between the histology of the breast cancer and nodal metastases (p = 0.001). All patients with nodal metastases had infiltrating ductal carcinoma (22/25), infiltrating lobular carcinoma (2/25), or mixed infiltrating ductal and lobular carcinoma (1/25). No nodal metastases were found in the patients with ductal carcinoma in situ, ductal carcinoma in situ with microinvasion, or in patients with tumors of more favorable histologic type (tubular, tubulolobular, mucinous, or medullary). Although we found no sentinel node metastases in patients with ductal carcinoma in situ or ductal carcinoma in situ with microinvasion, Pendas et al. [18] found that 5 (6%) of 87 patients with ductal carcinoma in situ without microinvasion had sentinel node metastases, and Zavotsky et al. [19] found that 2 (14%) of 14 patients with ductal carcinoma in situ with microinvasion had sentinel node metastases. In both studies, the sentinel nodes were the only nodes that contained tumor.
In our experience, no association was found between the size of the tumor and failed sentinel node biopsy or lymphoscintigraphy (p = 0.1 and p = 0.47, respectively), and no association was found between the histology of the tumor and failed sentinel node biopsy or failed lymphoscintigraphy (p = 0.46 and p = 0.36, respectively). Other authors have also found no correlation between the success of locating sentinel nodes and the size of the primary tumor [7, 10].
In summary, our data indicate that sentinel node biopsy guided by lymphoscintigraphy, intraoperative gamma probe, and isosulfan blue dye can accurately and safely determine the axillary nodal status of patients with clinically nodenegative invasive breast cancer. Large tumor size or prior excisional biopsy should not prevent patients from having sentinel node biopsy. Given the changing presentation of breast cancer in which small tumors are being discovered with increasing frequency and the decreasing numbers of patients with axillary metastases at presentation, we believe that sentinel node biopsy will help to eliminate dissection of healthy axillary nodes and substantially reduce the morbidity and cost of breast cancer treatment. In addition, sentinel node biopsy can stage the axilla of patients who have larger tumors and are candidates for primary chemotherapy because axillary node sampling after treatment may not accurately stage these patients with more advanced disease [13].
Acknowledgments
We thank Valerie P. Jackson for her insightful review of this
manuscript.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  F. M. Hall and H. E. Reynolds Lymphoscintigraphic Mapping of Sentinel Lymph Nodes Am. J. Roentgenol., October 1, 2001; 177(4): 945 - 945. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
