| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Divisions of Surgical Oncology (JMK, SBE, NW, TCH), and Pathology (JSW), Roswell Park Cancer Institute, Buffalo, New York.
Correspondence: Address correspondence and reprint requests to: Thelma C. Hurd, MD, Surgical Oncology, Roswell Park Cancer Institute, Elm & Carlton Street, Buffalo, NY 14263; Fax: 716-845-3434; E-mail: thelma.hurd{at}roswellpark.org
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: Intraoperative gross examination/TPA were performed on the SLN of consecutive breast cancer patients from 1997 to 2000. Patients with an intraoperative "positive" SLN underwent synchronous ALND. Intraoperative results were compared with the final pathology.
Results: Thirty-seven of 150 patients had a positive SLN on final pathology. Intraoperative gross examination/TPA identified 54% (20 of 37) of these patients. All intraoperative "positive" patients underwent synchronous ALND. Of 17 "false-negative" findings, 53% (9 of 17) had micrometastatic disease. There were no "false-positive" results. Overall sensitivity and specificity were 54% and 100%, respectively.
Conclusions: Gross examination/TPA are simple, rapid techniques for the intraoperative evaluation of a breast cancer SLN. As there were no false-positive results, the rationale behind SLN biopsy was preserved. These techniques permitted synchronous ALND in over half of all patients with a positive SLN. This represents a potential benefit to the patient by eliminating a second hospitalization for delayed ALND.
Key Words: Breast cancer • Sentinel lymph node • Touch prep analysis • Intraoperative evaluation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Surgical lymphadenectomy for pathologic evaluation is the only accurate method of determining nodal metastases. Until recently, this required removal of the entire axillary contents. Although safe, ALND carries a risk for long-term morbidity. Initially developed for melanoma, sentinel lymph node (SLN) biopsy has become a technique for accurately identifying breast cancer nodal metastases with morbidity markedly less than for ALND. The theory of SLN biopsy is that the pathologic status of the first or "sentinel" lymph node in the nodal drainage basin for a given breast tumor is representative of the remainder of the nodal basin. A positive SLN biopsy currently mandates formal ALND, whereas dissection of a nodal basin associated with a negative SLN biopsy can be omitted. At the present time, SLN biopsy has pervaded the treatment of breast cancer with multiple studies supporting the accuracy and safety of this technique for axillary staging.2–7
If ALND is to be performed for cases with a positive SLN, then it would be ideal to identify the positive node during the initial surgery. Patients with an intraoperatively positive SLN biopsy could then undergo synchronous ALND. Pathologic techniques that could accurately detect SLN metastases in a timely manner would be essential to this treatment algorithm. Although frequently used for intraoperative pathologic evaluation, frozen section analysis may destroy tissue necessary for permanent histology, is time consuming when multiple samples are present, and can be expensive. Touch prep analysis (TPA), also known as "imprint cytology," is already used for the rapid intraoperative diagnosis of breast cancer and the evaluation of surgical resection margins.8–10 The purpose of this study was to evaluate the sensitivity of gross examination and TPA for the intraoperative assessment of the metastatic status of a breast cancer SLN biopsy and to use this information as a determinant for performing a synchronous ALND.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Sentinel Lymph Node Biopsy Technique
The breast cancer SLN biopsy technique at our institution had been previously validated with completion ALND and was comparable to the results of other studies with regard to sensitivity and specificity.5 The tissues surrounding the site of the primary tumor or prior excision cavity were injected in four quadrants with 1 mCi of 99mTc-sulfur colloid (Syncor, Buffalo, NY) in a 4-cc total volume. Approximately 2 hours after injection, lymphoscintigraphy was performed and interpreted by the nuclear medicine radiologist.
Approximately 4 hours after injection of the 99mTc-sulfur colloid, the patient was taken to the operating room for SLN biopsy and surgical treatment of the primary tumor site if indicated. The majority of cases were performed under general anesthesia. Peritumoral or pericavity injection of 1% isosulfan blue dye (Lymphazurin, US Surgical Corporation, Norwalk, CT) was performed to a total volume of 3–5 cc. Using a hand-held gamma probe (Neoprobe 1000, Neoprobe Corporation, Dublin, OH, or Navigator, US Surgical Corporation), the SLN(s) were identified using a combination of radioactivity and blue dye staining. All SLNs were examined with the gamma probe ex vivo in comparison to normal tissues and sent to pathology for intraoperative and permanent pathologic evaluation. While the SLNs were being examined intraoperatively, the axillary incision was closed and surgical treatment of the primary tumor site, if indicated, was performed using separate instruments.
Intraoperative Pathologic Evaluation
Only lymph nodes clearly identified as the SLNs were bivalved and grossly examined for macroscopic foci of tumor. A lymph node was considered to be grossly positive if there were obvious tumor nodules within the lymph node. The presence of metastasis was confirmed by TPA or frozen section. If gross examination was negative, TPA was performed by pressing each cut half of the specimen onto a glass slide and rapidly staining the "imprint" with hematoxylin and eosin (H&E). The touch preparation was evaluated by the surgical pathologist assigned to frozen section. A cytopathologist was not required for TPA interpretation, because the intraoperative evaluation only requires the identification of malignant epithelial cells within a background of lymphocytes. At the pathologist’s discretion, the presence of either gross tumor or a positive TPA could be confirmed with frozen section analysis. A definitive finding of tumor on gross examination or TPA was considered an intraoperative "positive" result. Questionable or negative findings were considered an intraoperative "negative" result and deferred to permanent evaluation. The intraoperative finding of "positive" or "negative" was conveyed to the operating room, and all specimens were then processed for permanent pathologic evaluation. On average, the intraoperative evaluation of the sentinel lymph node requires approximately 15 minutes.
Permanent Pathologic Evaluation
Formal processing of each half of the previously bivalved SLN specimen consisted of 10% formalin fixation, paraffin imbedding, and serial step-sectioned at three levels. All sections were stained with H&E and examined for evidence of metastatic tumor. A "micrometastasis" was defined as a metastatic tumor focus < 2 mm in greatest diameter. Although routine immunohistochemistry (IHC) was not performed, suspicious findings on H&E were further evaluated with IHC using the monoclonal antibody for cytokeratin AE1/3.
Axillary Lymph Node Dissection
Patients with a "positive" SLN on intraoperative pathologic evaluation underwent a synchronous standard level I/II ALND under general anesthesia.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Two patients were excluded from this analysis due to deviations from the predetermined intraoperative algorithm. The SLN of one patient underwent gross examination only (no TPA), and one patient had TPA performed on only one of two identified sentinel lymph nodes. Both of these patients had a positive SLN biopsy on final pathology. Exclusion of these two patients resulted in 150 patients eligible for evaluation after intraoperative pathologic evaluation. Thirty-seven of these 150 patients (28%) had a positive SLN biopsy based on final permanent pathology.
Twenty-two percent (8 of 37) of all patients with a positive SLN biopsy were identified intraoperatively by gross examination of the SLN. An additional 32% (12 of 37) were identified by TPA after a negative gross examination. The remaining 46% (17 of 37) had both a negative gross examination and TPA. Fifty-three percent (9 of 17) of the patients with a "false-negative" intraoperative evaluation had micrometastatic disease on final permanent pathology detected only after serial sectioning.
All 20 patients with a positive gross examination or TPA underwent synchronous ALND. This accounted for 54% (20 of 37) of all patients with a positive SLN biopsy. The majority of patients with a positive SLN biopsy but a "negative" intraoperative pathologic evaluation underwent delayed ALND. There were no "false-positive" intraoperative pathologic findings. Therefore, no synchronous ALND was performed for a negative SLN biopsy.
The results of the intraoperative pathologic evaluation with gross examination alone, TPA alone after the exclusion of grossly negative patients, and the combination of gross examination and TPA are summarized in Tables 1–3. For gross examination or TPA alone, sensitivity was 22% and 41%, respectively. Sensitivity rose to 54% for gross examination combined with TPA. Specificity, positive predictive value, and negative predictive value were 100%, 100%, and 87%, respectively, for the gross examination/TPA combination.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The technique of "axillary sampling" was initially developed to stage the axilla with less morbidity. Although some studies showed promising results, nodal selection was too inconsistent for accurate and reproducible axillary staging.14–16 Extrapolating from its role in the treatment of melanoma, SLN biopsy was subsequently applied to breast cancer axillary staging. The results of breast cancer SLN biopsy have been dramatic. Meta-analysis of multiple studies has shown an overall success rate of 84%, concordance with ALND of 98%, and a false-negative rate of only 5%.17 In addition, the morbidity associated with SLN biopsy has been extremely low.17,18
Traditional pathologic evaluation of breast cancer axillary lymph nodes has consisted of bivalving all identified nodes with subsequent microscopic examination. Although this approach is efficient for the processing of large numbers of nodes, random sectioning of the lymph nodes reduces the chance of detecting micrometastatic disease. The advent of SLN biopsy has dramatically changed the pathologic evaluation for nodal metastases. The limited number of identified nodes allows intensive pathologic review of the entire specimen by hematoxylin and eosin staining of serial sections. Several institutions have also used IHC for epithelial antigens in an effort to further improve the identification of micrometastatic disease.19–22
There is an obligate delay of several days from the time of SLN biopsy until the final determination of the sentinel nodal status. This "lag" time in SLN biopsy processing precludes the performance of a synchronous ALND. Therefore, completion ALND for patients with a positive SLN biopsy usually requires a separate hospitalization, a second operative procedure with its attendant anesthetic risk and cost, increased surgical risk due to the distortion of normal axillary anatomy as a consequence of the previous biopsy, additional time demands on personal or professional commitments, and prolonged anxiety over the biopsy results and the possible need for additional surgery. These factors represent potentially increased emotional and financial costs for the patient.
This study was conducted to determine whether gross examination and TPA could identify patients with a positive SLN biopsy at the time of surgery, allowing for synchronous ALND. We found that 54% of all patients with a positive breast cancer SLN biopsy were successfully identified intraoperatively by combining these two pathologic techniques. The sensitivity of gross examination alone was relatively low at 22%. However, the sensitivity of TPA on grossly negative specimens was 41%. By combining these two simple techniques, over half of all women with a positive SLN biopsy at our institution were able to undergo synchronous ALND.
The specificity of intraoperative gross examination combined with TPA was 100%. Therefore, none of the patients underwent a synchronous ALND for a negative SLN biopsy on permanent pathology. This approach preserves the rationale behind the development of the SLN technique; ALND is performed only for pathologically proven lymph node metastases.
Several authors have investigated the accuracy of intraoperative pathologic evaluation of breast cancer SLN biopsies through the use of frozen section analysis. In 1997, Veronesi et al.6 reported on 160 women undergoing successful SLN biopsy. In 107 of the patients, intraoperative frozen section analysis was performed on the SLN biopsy. Frozen section correctly identified 32 of the 50 patients with positive final pathology for a sensitivity of 64%. There were no false-positive results. Of the "negative" intraoperative findings, 24% (18 of 75) were false negatives. All of these were in patients with micrometastatic disease. In a similar study by Flett et al.,7 intraoperative frozen section analysis of SLN biopsies corresponded to the final pathology in 53 of 56 patients (95%). There were no false positives, two suspicious findings, and one false-negative result. Finally, an intensive study by Viale et al.23 examined intraoperative frozen section serial sectioning with both H&E and rapid IHC. Sentinel lymph node metastases were detected in 70 of 155 patients (45%). Interestingly, 17 of the 70 patients (24%) with a positive SLN had micrometastatic disease that was identified by these techniques. Rapid IHC did not detect additional SLN metastases that were not already proven or suspicious by H&E. Unfortunately, the time to perform this intraoperative evaluation typically approached 1 hour.
Although intraoperative frozen section analysis preserves the nodal architecture, there are several disadvantages to this technique. It is time consuming, potentially inaccurate, may cause freezing artifact, and can destroy valuable specimen. The latter is especially important given the increasing prevalence of isolated micrometastatic disease in SLN specimens. In an effort to decrease processing time and preserve the entire lymph node, several studies have examined TPA for the intraoperative evaluation of breast cancer lymph node biopsies.
The earliest studies of TPA in breast cancer examined its role in axillary lymph node sampling. In 1974, Rimsten et al.24 reported on 50 cases in which TPA of sampled axillary lymph nodes was used to determine the need for ALND at the time of mastectomy. Intraoperative TPA was 100% accurate in predicting the presence or absence of nodal metastases. In a similar study, Quill et al.25 used intraoperative TPA to determine the need for perioperative chemotherapy in 13 women undergoing simple mastectomy with axillary lymph node sampling. The overall accuracy was 95% (82 of 86 lymph nodes). There were three false negatives and one false positive compared with the final permanent pathology. Sensitivity and specificity were 93% and 98%, respectively. Finally, a large series by Hadjiminas and Burke26 used intraoperative TPA to determine the need for ALND in 144 consecutive women undergoing axillary sampling. There were five patients with false-negative TPA findings and three with false-positive results. Sensitivity was 89% and specificity was 96%.
Recently, TPA has also been applied to the intraoperative assessment of breast cancer SLN biopsies. A study by Rubio et al.27 examined the accuracy of intraoperative TPA on 124 radiolabeled lymph nodes in 53 patients undergoing successful SLN biopsy. The results of TPA correlated with the final pathology in 99% of cases. There were no false-positive and one false-negative finding. Sensitivity and specificity were 96% and 100%, respectively. These results were similar to the findings of Ratanawichitrasin et al.28 There was a 98% concordance rate between intraoperative TPA and the final pathology in 55 patients undergoing SLN biopsy. The only false-negative finding occurred in a specimen with micrometastatic disease. Sensitivity was 93% and specificity was 100%. A study by Turner et al.29 compared intraoperative TPA combined with frozen section to permanent H&E and IHC in a large study of 278 SLN biopsy patients. Overall accuracy was 93%. However, the sensitivity of TPA and frozen section varied based on the size and method of detection of the SLN metastases. The sensitivity of the combined techniques was 98% (46 of 47) for macroscopic nodal metastases on H&E. This rate dropped to 28% (7 of 25) for micrometastases on H&E. None of the micrometastases seen on IHC were detected by TPA/frozen section (0 of 39). There is no direct analysis of the additional benefit of frozen section examination of TPA "negative" specimens. However, the authors state that a retrospective review of the TPA results alone closely paralleled the findings of TPA combined with frozen section in 52 randomly selected cases.
Although many of the previously mentioned studies have validated the accuracy of intraoperative pathologic SLN evaluation, its use as a determinant for synchronous ALND has only been inferred as routine completion ALND was typically performed at the time of SLN biopsy or deferred until the final permanent pathology.6,7,23,27–29 However, the Moffit Cancer Center (MCC) has examined the utility of synchronous ALND for intraoperative "positive" TPA results.30 In this series, 47 of 210 patients had a positive SLN biopsy on permanent pathology. TPA identified 50% of all metastatic disease and synchronous ALND was performed in this group of patients. In a separate report on the accuracy of intraoperative "imprint cytology" on 381 grossly negative breast cancer sentinel lymph nodes, there were 15 true positive, one false-positive, and 35 false-negative findings.31 Sensitivity and specificity were 30% and 99.6%, respectively. Based on the MCC experience, false-negative results were fairly common in grossly negative nodes or in the presence of only micrometastatic disease.
The false negative rate for the combined intraoperative techniques in our series was 46%. Although this may seem relatively high, over half of these cases had isolated foci of micrometastatic disease seen only on thorough step-sectioning of the SLN. The decreased ability of TPA to detect micrometastases is consistent with the experience of MCC and several of the studies cited above.6,28,29,31 Therefore, gross examination and TPA seem to be less sensitive for small volumes or peripherally located foci of metastatic disease. Given that gross examination and TPA sampled only the midportion of each SLN, more intensive intraoperative sectioning may have reduced the false negative rate in our study. However, the thin sections required to accurately assess for micrometastases would have also been more time-consuming and may have potentially destroyed the specimen. Although the use of rapid IHC techniques for evaluating breast cancer lymph node specimens has been examined, the results are too preliminary to use this approach for intraoperative decision-making at the present time.30,31 As the time for our intraoperative pathologic evaluation was comparable to the time necessary to close the biopsy incision, we believe that our intraoperative algorithm is an acceptable balance between sensitivity and the effective use of operative time.
In the present study, intraoperative gross examination/TPA correctly identified 20 of 37 patients with a positive SLN. This identification allowed for synchronous ALND in 54% of patients who would have required a delayed ALND based on the final pathology. If one also considers the 130 patients who had a true negative or a false negative on the intraoperative evaluation, then gross examination/TPA altered the immediate operative plan in 13% (20 of 150) of all patients undergoing SLN biopsy.
Although some may argue that this rate of change may not justify the routine use of intraoperative pathologic evaluation, we believe that this approach is effective for several reasons. First, many current medical interventions are considered "appropriate" if the overall perceived benefit to the population at risk is > 10%. Second, the intraoperative pathologic evaluation is extremely rapid and simple to perform. It does not require any resources other than those readily available in any pathology frozen section room. As stated previously, the results of the intraoperative evaluation were typically known before the complete closure of the axillary incision. Therefore, it did not prolong the operative time in any patient with an intraoperative negative result. Finally, although there was no attempt to perform a direct cost analysis in this study, the cost of an intraoperative pathologic evaluation would seem to be far less than one delayed ALND (based on the 13% change in our study) and its associated costs. However, this is an area that would definitely warrant further evaluation from a fiscal standpoint.
In summary, intraoperative gross examination and TPA are simple, rapid methods for detecting nodal metastases in patients undergoing a breast cancer SLN biopsy. In our study, TPA was complimentary to the results of gross examination alone. However, these methods are less sensitive for the presence of micrometastatic disease. Despite these limitations, intraoperative gross examination and TPA allowed for synchronous ALND to be performed in over half of patients with a positive SLN biopsy at our institution. This approach represents a potential emotional and time benefit for the patient by eliminating the need for a second hospitalization for delayed ALND. In addition, the high positive predictive value of these methods preserves the rationale behind the development of the SLN biopsy technique, i.e., ALND is reserved for patients with pathologically documented nodal metastases.
|
Footnotes |
|---|
Received for publication July 24, 2000. Accepted for publication December 5, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Gretschel, A. Bembenek, M. Hunerbein, S. Dresel, W. Schneider, and P. M. Schlag Efficacy of Different Technical Procedures for Sentinel Lymph Node Biopsy in Gastric Cancer Staging Ann. Surg. Oncol., July 1, 2007; 14(7): 2028 - 2035. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Pugliese, R. Tickman, N. P. Wang, M. Atwood, and J D. Beatty The Utility of Intraoperative Evaluation of Sentinel Lymph Nodes in Breast Cancer Ann. Surg. Oncol., March 1, 2007; 14(3): 1024 - 1030. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Perez, S. Vidal-Sicart, G. Zanon, M. Velasco, G. Santamaria, A. Palacin, E. Campo, A. Cardesa, and P. L. Fernandez A Practical Approach to Intraoperative Evaluation of Sentinel Lymph Node Biopsy in Breast Carcinoma and Review of the Current Methods Ann. Surg. Oncol., April 1, 2005; 12(4): 313 - 321. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Brogi, E. Torres-Matundan, L. K. Tan, and H. S. Cody III The Results of Frozen Section, Touch Preparation, and Cytological Smear Are Comparable for Intraoperative Examination of Sentinel Lymph Nodes: A Study in 133 Breast Cancer Patients Ann. Surg. Oncol., February 1, 2005; 12(2): 173 - 180. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Lucci, A. Shoher, M. O. Sherman, and A. Azzizadeh Assessment of the Current Medicare Reimbursement System for Breast Cancer Operations Ann. Surg. Oncol., December 1, 2004; 11(12): 1037 - 1044. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Fortunato, M. Amini, M. Farina, S. Rapacchietta, L. Costarelli, F. R. Piro, G. Alessi, P. Pompili, S. Bianca, and C. E. Vitelli Intraoperative Examination of Sentinel Nodes in Breast Cancer: Is the Glass Half Full or Half Empty? Ann. Surg. Oncol., November 1, 2004; 11(11): 1005 - 1010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Bedrosian, D. Bedi, H. M. Kuerer, B. D. Fornage, L. Harker, M. I. Ross, F. C. Ames, S. Krishnamurthy, B. S. Edeiken-Monroe, F. Meric, et al. Impact of Clinicopathological Factors on Sensitivity of Axillary Ultrasonography in the Detection of Axillary Nodal Metastases in Patients With Breast Cancer Ann. Surg. Oncol., November 1, 2003; 10(9): 1025 - 1030. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Lardinois, T. Brack, A. Gaspert, T. Spahr, D. Schneiter, H.C. Steinert, and W. Weder Bronchoscopic radioisotope injection for sentinel lymph-node mapping in potentially resectable non-small-cell lung cancer Eur. J. Cardiothorac. Surg., May 1, 2003; 23(5): 824 - 827. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. McCready Intraoperative Sentinel Node Assessment Is a Touchy Subject Ann. Surg. Oncol., May 1, 2002; 9(4): 321 - 323. [Full Text] [PDF]
|
|
|
|
|
|
A. R. Miller, V. E. Thomason, I-T. Yeh, A. Alrahwan, F. E. Sharkey, J. Stauffer, P. M. Otto, C. McKay, M. S. Kahlenberg, W. T. Phillips, et al. Analysis of Sentinel Lymph Node Mapping With Immediate Pathologic Review in Patients Receiving Preoperative Chemotherapy for Breast Carcinoma Ann. Surg. Oncol., April 1, 2002; 9(3): 243 - 247. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
