| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Surgical Oncology (RFH, KKH, NM, FCA, BF, HMK, SES, GB, FM, JSA, JN, MIR) and Pathology (SK), The University of Texas M. D. Anderson Cancer Center, Houston, Texas.
Correspondence: Address correspondence and reprint requests to: Kelly K. Hunt, MD, Surgical Breast Section, UT M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Box 444, Houston, TX 77030; Fax: 713-792-4689; E-mail: khunt{at}mail.mdanderson.org
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: We reviewed the records of all patients with invasive breast cancer who underwent SLN biopsy at our institution between 1993 and August 2001. Multivariate analysis was used to identify clinicopathologic features in SLN-positive patients that predict involvement of NSLNs.
Results: A total of 131 patients had a positive SLN and underwent completion axillary lymph node dissection. Multivariate analysis revealed that primary tumor >2 cm (P = .009), SLN metastasis >2 mm (P = .024), and lymphovascular invasion (P = .028) were independent predictors of positive NSLNs. The number of SLNs harvested was a significant negative predictor (P = .04). In our model, based on the presence of these factors, the positive predictive value was 100% for a score of 4.
Conclusions: The likelihood of positive NSLNs correlates with primary tumor size, size of the largest SLN metastasis, and presence of lymphovascular invasion. A scoring system incorporating these factors may help determine which patients would benefit from additional axillary surgery.
Key Words: Sentinel lymph node • Predictive factors • Axillary dissection • Breast cancer
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
CLND has remained the standard of practice when the SLN is involved with tumor. However, in the 40% to 60% of patients in whom the SLN is the only positive lymph node,4 removal of further axillary nodes may not benefit the patient. In fact, given the added morbidity, operative time, and cost of formal axillary dissection,5 avoidance of unnecessary axillary dissection would be desirable. Furthermore, the National Surgical Adjuvant Breast and Bowel Project B-04 trial and other studies have not demonstrated a survival advantage from immediate axillary lymph node dissection in early-stage breast cancer,6,7 although this procedure may reduce the incidence of axillary recurrence. Previously, recommendations for adjuvant systemic therapy were based on lymph node status; more recently, however, adjuvant chemotherapy is recommended for most patients with tumors >1 cm regardless of axillary nodal involvement.8 Because lymphatic and hematogenous metastasis may occur via distinct and possibly unrelated pathways, the clinical relevance of axillary nodal status in comparison to that of hematogenous metastasis is now being questioned.9,10
Although there are arguments against CLND, some studies suggest that survival is worse with less than complete dissection of the axillary nodes. Patients in the Guys’ Hospital trial11 who were randomized to radical mastectomy had a significant survival advantage compared with patients who underwent lumpectomy and breast and axillary irradiation at doses now considered inadequate to treat the axilla. A meta-analysis of six randomized, controlled trials concluded that prophylactic axillary dissection for node-negative patients results in a 5.4% survival benefit. Thus, the role of further axillary dissection in patients with a positive SLN is still unclear. Moreover, the development of newer pathologic techniques to detect the presence of tumor in lymph nodes (immunohistochemistry [IHC], reverse transcriptase-polymerase chain reaction, touch imprint cytology, flow cytometry, and so on) adds to the dilemma of how to handle a positive SLN. In this study, we analyzed various clinicopathologic features in cases of SLN-positive breast cancers to determine factors that might help predict the involvement of nonsentinel axillary lymph nodes (NSLNs).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Technique of Intraoperative Lymphatic Mapping and SLN Biopsy
Our technique of SLN biopsy has been described elsewhere.12 Briefly, intraoperative lymphatic mapping was performed with injections of blue dye alone or in combination with 99mTc-labeled sulfur colloid. When lymphoscintigraphy was performed, patients received filtered 99mTc sulfur colloid, which was injected into the breast parenchyma surrounding the tumor or the biopsy cavity either on the day of surgery (.5 mCi) or the day before surgery (2.5 mCi). The injection was performed under mammographic or sonographic guidance when the tumor was not palpable. On the day of surgery, with the patient under general anesthesia, 5 mL of 1% isosulfan blue (Lymphazurin; US Surgical, Norwalk, CT) was injected peritumorally, and the breast was massaged for 5 minutes. In patients who underwent sulfur colloid injection, a handheld gamma detection probe (NeoProbe 2000; US Surgical) was used to scan the axilla transcutaneously and identify the most radioactive area. Through an axillary incision over this "hot spot," SLNs were identified as those with blue dye uptake, radiotracer uptake (more than two times the background count), or both.
During the SLN biopsy learning phase for each surgeon in this study, all SLN biopsies were followed by immediate level I and II axillary lymph node dissection (CLND) for the first 30 cases. Otherwise, CLND was performed whenever a SLN was found to be positive on pathologic examination. If the SLN appeared grossly suspicious for tumor and this was confirmed by frozen-section examination, CLND was performed during the same surgery.
Pathologic Evaluation of SLNs
The pathologic evaluation of SLNs at M. D. Anderson Cancer Center has evolved as a result of advancing technology. Before April 2000, SLNs were serially sectioned along the short axis at 2- to 3-mm intervals; sections were embedded in paraffin blocks, and one level from each block was stained with H&E. Beginning in April 2000, SLNs were grossly processed in the same manner as before; however, each paraffin block was serially sectioned at 5-µm intervals to produce 10 levels. Levels 1, 2, and 4 were evaluated by routine H&E staining, and level 3 was analyzed for cytokeratin by IHC.13
The maximum dimension of the metastasis in each SLN was measured. If more than one deposit of tumor was found in a SLN or if multiple SLNs were positive, the size of the largest tumor deposit was recorded. All SLNs were examined for extranodal extension. NSLNs from the CLND specimen were analyzed by H&E staining only. Each primary tumor was evaluated for size of the invasive component, histological type, Black’s nuclear grade, estrogen and progesterone receptor status, Her-2/neu status, and presence of lymphovascular invasion. For hormonal receptor status, >10% staining of cells by IHC was considered positive.
Statistical Methods
Descriptive statistics were used to assess the frequency distribution among the study population. Associations between the presence of positive NSLNs and various characteristics—including age, primary tumor size, use of neoadjuvant chemotherapy, histological subtype, estrogen and progesterone receptor status, Her-2/neu overexpression, Black’s nuclear grade, lymphovascular invasion in the primary tumor, number of SLNs removed, number of histologically positive SLNs, presence of grossly positive SLNs, size of the largest SLN metastasis, and extent of extranodal extension in the SLN—were analyzed with the 
2 test. P values of 
.05 were considered to be statistically significant.
Characteristics that were found to be significant by univariate analysis were then entered in a forward stepwise method in a logistical regression analysis to develop a multivariate model of independent factors for predicting the presence of positive NSLNs. Information on these 4 independent factors was available for 86 patients (65.6%) from 131 patients who underwent CLND. By using this information, a scoring system was then developed in the following manner: the ß coefficient obtained for each significant factor from the multivariate logistical regression model was rounded off to a whole integer. For example, for tumor size >2 cm, the ß coefficient of 1.41 was rounded off to 1.0, and for patients with tumor size <2 cm, a score of 0 was assigned. The scores for each significant factor were then added to obtain a total score for the individual patient. Thus, the final resulting scores ranged from -2 to 4. Scores were calculated for the patients in our series for whom all four characteristics found to be significant on multivariate analysis were known. By using the receiver operating characteristic curve method and various cutoff values, the sensitivity, specificity, and positive and negative predictive values of each score were determined. The SPSS 10.1 software package (SPSS Inc., Chicago, IL) was used for all statistical analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
Characteristics of SLNs and NSLNs
The median number of SLNs identified was 2 (range, 1–8; SD, 1.64). The median number of SLNs found to be positive was 1 (range, 1–5; Table 1). The median size of the largest SLN metastasis was 6 mm (range, .2–25 mm). Micrometastasis (2 mm) was identified in 30 cases (24.6%). Only 12 SLNs (9.2%) were grossly involved with tumor. Metastasis was identified in the SLN only by IHC for cytokeratin in 13 patients, of whom 3 underwent CLND; none of these were found to have a positive NSLN.
The median number of lymph nodes harvested with CLND was 15 (range, 3–37), and the median number of positive NSLNs was 0 (range, 0–29). Fifty-three (40.5%) of the 131 patients in our series were found to have positive NSLNs by standard H&E examination.
Correlation Between Clinicopathologic Features and Positive NSLNs
Table 2 summarizes the results of the statistical analyses to determine the relationship between clinicopathologic variables and positive NSLNs. Univariate analysis revealed that primary tumor size >2 cm, largest SLN metastasis >2 mm, and receipt of neoadjuvant chemotherapy were significantly associated with positive NSLNs. There was a trend toward an association between lymphovascular invasion and positive NSLNs (P = .057). When neoadjuvant chemotherapy was excluded from the multivariate analysis, primary tumor size, size of the largest SLN metastasis, and lymphovascular invasion were all significant independent predictors of positive NSLNs (Table 3). The risk of positive NSLNs was 4.1 times as great with primary tumors >2 cm as with smaller primary tumors (P = .009; 95% confidence interval, 1.42–11.89). When the largest SLN metastasis was >2 mm in maximum diameter, the risk of positive NSLNs was six times as high as the risk when the largest SLN metastasis was a micrometastasis (P = .024; 95% confidence interval, 1.26–28.70). The presence of lymphovascular invasion in the primary tumor increased the risk of positive NSLNs to 3.7 times the risk without lymphovascular invasion (P = .028; 95% confidence interval, 1.15–11.70). In addition, the number of SLNs harvested was inversely related to the presence of positive NSLNs. The likelihood of having a positive NSLN was lower in patients with three or more SLNs identified than in patients with fewer SLNs identified (odds ratio, .19; P = .04; 95% confidence interval, .04–.93). The ß coefficient associated with each significant characteristic in the multivariate model is shown in Table 3.
|
0 had additional positive NSLNs identified during CLND. Conversely, all nine patients with a score of 4 who had all three independent positive predictors present (i.e., tumor size >2 cm, largest SLN metastasis >2 mm, and presence of lymphovascular invasion) and fewer than three SLNs removed were found to have positive NSLNs.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Several other groups have also identified factors that seem to be associated with the presence of positive NSLNs.17–23 In five of these six studies, tumor size >2 cm was significantly associated with additional NSLN metastases. We found that primary tumors >2 cm were four times as likely as smaller tumors to be associated with disease in NSLNs. None of our patients with T1a tumors had positive NSLNs, whereas T2, T3, and T4 tumors were associated with positive NSLNs in 54%, 77%, and 80% of cases, respectively.
Other groups that included the size of the SLN tumor volume in their analyses found, as we did, that largest SLN metastasis >2 mm was a significant predictor of positive NSLNs. In the series of Turner et al.,17 47% of patients with largest SLN metastasis >2 mm were found to have positive NSLNs when NSLN disease status was determined by H&E. In fact, the amount of disease in the SLN was the strongest positive predictor of all the factors we analyzed, with an odds ratio of 6.0, which confirms the results by Chu et al.19
Lymphovascular invasion was another significant predictor of NSLN involvement; more than half of tumors with lymphovascular invasion had positive NSLNs, compared with one third of tumors without lymphovascular invasion. Abdessalam et al.18 and Turner et al.17 also demonstrated that peritumoral lymphovascular invasion was significantly associated with positive NSLNs. When this variable was present, 62% of the patients in the Abdessalam series had positive NSLNs, which was slightly higher than our rate of 52% for this same group of patients. In a review of 206 patients with positive SLNs who underwent CLND, Weiser et al.23 found that three variables were inversely related to the likelihood of positive NSLN. When tumor size was 1.0 cm, lymphovascular invasion was absent, and SLN metastasis size was 2 mm, no NSLN metastases were identified in 24 patients. Interestingly, we found that the number of SLNs removed was a significant negative predictive factor. The likelihood of finding additional positive NSLNs decreased as more SLNs were harvested, probably reflecting the reduced number of nodes remaining in the axilla.
We found that our scoring system, in which the ß coefficients for each significant variable are summed to produce a total score, may be used to predict the likelihood that NSLNs will be involved with tumor. However, because we did not have information on all four significant predictors for all of our patients, this scoring system is limited by the small number of patients on which it was based. Furthermore, the model remains to be tested on a larger group of patients. With these limitations in mind, the scoring system may be most useful at the two extremes of its scale. The lowest possible score, of -2, is achieved when the primary tumor is <2 cm, the tumor in the SLN is a micrometastasis only (<2 mm), lymphovascular invasion is not present, and more than three SLNs have been removed. Although the sensitivity associated with this score is 100%, because none of our patients with this score were found to have a positive NSLN, the specificity is very low because we had only two patients with a score of -2. Similarly, none of the patients in this study with a score <1 had an NSLN involved with tumor. However, all of the patients in this study with a score of 4 had residual axillary disease beyond the SLN, although again, only nine patients fell into this subgroup. Thus, for the patients with scores at either extreme of the scale (one third of our patients), our model was able to accurately predict axillary nodal status 100% of the time. For patients with scores between 1 and 3, it is difficult to predict which ones will have additional positive NSLNs, because these patients were divided almost equally between those who did and those who did not have further axillary disease. In summary, although this model needs to be validated in a larger patient population, it was able to correctly predict NSLN involvement in a substantial proportion of our patients. Additionally, although this study evaluated 14 different clinicopathologic features, there most likely are other variables not included here that are important factors in helping to predict which patients would benefit from further axillary dissection.
Analysis of the SLN by IHC for cytokeratin was performed in approximately 60% of the patients in our study. The presence of cytokeratin in a single cell alone was not classified as a positive sentinel node. Metastasis was detected in the SLN only by IHC, after negative H&E evaluation, in 13 (13%) of the 98 patients whose nodes were examined with IHC analysis. Other published series have reported between 7% and 26% of SLNs being upstaged on IHC examination after negative findings on H&E staining.17,20 Only 3 of the 13 patients in our study with SLNs positive only by IHC underwent CLND, and none of these was found to have additional positive nodes. Because the number of patients in this series with cytokeratin-positive sentinel nodes used in the multivariate analysis was so small, a separate analysis of their clinicopathologic features was not performed. The H. Lee Moffitt Cancer Center group20 reported that 7.7% of their patients with SLNs positive by IHC only had additional positive nodes, whereas the John Wayne Cancer Institute reported a much higher rate of 20%.17 In the John Wayne study, however, NSLNs were subjected to cytokeratin IHC when the nodes were negative by H&E. Of 124 patients who had NSLNs examined by IHC, micrometastases were identified in 18 (14.5%), with an overall NSLN positivity rate of 45% by either IHC or H&E. Thus, of the patients with IHC-only positive SLNs, only 4% had NSLNs positive by H&E. In our study, NSLNs were evaluated only by routine H&E; IHC was not performed. The routine use of IHC to detect metastatic disease remains controversial. Several groups have reported that the presence of occult metastases detected only by IHC has a negative effect on disease-free survival.24,25 However, on multivariate analysis, only Cote et al.26 found a poorer disease-free and overall survival, which was limited to postmenopausal patients.
In summary, our study demonstrated that the SLN was the only positive axillary node in 60% of our patients. Primary tumor size >2 cm, largest SLN metastasis >2 mm, and lymphovascular invasion were independent predictors of additional disease in the axilla. As the number of SLNs increased above 3, the likelihood of finding positive NSLNs decreased significantly. A scoring system based on these factors may be useful in determining the status of the NSLNs and could be helpful in deciding which patients would benefit from CLND. Two prospective, randomized trials currently in progress address the controversy surrounding the relevance of an IHC-positive SLN (American College of Surgeons Oncology Group Z0010) and the need for further axillary dissection in patients with positive SLNs (American College of Surgeons Oncology Group Z0011). In the Z0011 trial, patients with a positive SLN by H&E will be randomized to either CLND or observation. This study will help to determine whether axillary node dissection has any effect on survival in this group of pathologically node-positive patients.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
In breast cancer patients with positive sentinel nodes (SLNs), a scoring system was devised based on primary tumor size, SLN metastasis size, lymphovascular invasion, and number of SLNS harvested to help predict the presence of positive nonsentinel axillary nodes.
Received for publication May 15, 2002. Accepted for publication October 25, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Cho, W. Han, and D.-Y. Noh What Is the Best? Ann. Surg. Oncol., November 1, 2008; 15(11): 3317 - 3317. [Full Text] [PDF]
|
|
|
|
|
|
J. Cho, W. Han, J. W. Lee, E. Ko, S. Y. Kang, S.-Y. Jung, E.-K. Kim, W. K. Moon, N. Cho, I.-A. Park, et al. A Scoring System to Predict Nonsentinel Lymph Node Status in Breast Cancer Patients with Metastatic Sentinel Lymph Nodes: A Comparison with Other Scoring Systems Ann. Surg. Oncol., August 1, 2008; 15(8): 2278 - 2286. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Barranger, O. Morel, and C. Coutant Axilla Scoring Systems Predicting Risk of Non-sentinel-node Metastasis in Breast Cancer Patients with a Positive Sentinel Node Ann. Surg. Oncol., April 1, 2008; 15(4): 1261 - 1262. [Full Text] [PDF]
|
|
|
|
 |
|
 G. G. Van den Eynden, M. K. Vandenberghe, P.-J. H. van Dam, C. G. Colpaert, P. van Dam, L. Y. Dirix, P. B. Vermeulen, and E. A. Van Marck Increased Sentinel Lymph Node Lymphangiogenesis is Associated with Nonsentinel Axillary Lymph Node Involvement in Breast Cancer Patients with a Positive Sentinel Node Clin. Cancer Res., September 15, 2007; 13(18): 5391 - 5397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Samoilova, J. T. Davis, J. Hinson, Y. M. Brill, M. L. Cibull, P. McGrath, E. Romond, A. Moore, and L. M. Samayoa Size of Sentinel Node Tumor Deposits and Extent of Axillary Lymph Node Involvement: Which Breast Cancer Patients May Benefit From Less Aggressive Axillary Dissections? Ann. Surg. Oncol., August 1, 2007; 14(8): 2221 - 2227. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E. Dauphine, J. S. Haukoos, M. P. Vargas, N. M. Isaac, I. Khalkhali, and H. I. Vargas Evaluation of Three Scoring Systems Predicting Non Sentinel Node Metastasis in Breast Cancer Patients with a Positive Sentinel Node Biopsy Ann. Surg. Oncol., March 1, 2007; 14(3): 1014 - 1019. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Smidt, L. J. A. Strobbe, H. M. M. Groenewoud, G. J. der Wilt, K. J. Van Zee, and T. Wobbes Can Surgical Oncologists Reliably Predict the Likelihood for Non-SLN Metastases in Breast Cancer Patients? Ann. Surg. Oncol., February 1, 2007; 14(2): 615 - 620. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Motomura, C. Egawa, Y. Komoike, S. Nagumo, H. Koyama, and H. Inaji Three-Axillary Lymph Node Sampling for the Prediction of Nonsentinel Node Metastases in Breast Cancer Patients With Sentinel Node Metastases Ann. Surg. Oncol., July 1, 2006; 13(7): 985 - 989. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Houvenaeghel, C. Nos, H. Mignotte, J. M. Classe, S. Giard, P. Rouanet, F. P. Lorca, J. Jacquemier, and V. J. Bardou Micrometastases in Sentinel Lymph Node in a Multicentric Study: Predictive Factors of Nonsentinel Lymph Node Involvement--Groupe Des Chirurgiens De La Federation Des Centres De Lutte Contre Le Cancer J. Clin. Oncol., April 20, 2006; 24(12): 1814 - 1822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Lambert, G. D. Ayers, R. F. Hwang, K. K. Hunt, M. I. Ross, H. M. Kuerer, S. E. Singletary, G. V. Babiera, F. C. Ames, B. Feig, et al. Validation of a Breast Cancer Nomogram for Predicting Nonsentinel Lymph Node Metastases After a Positive Sentinel Node Biopsy Ann. Surg. Oncol., March 1, 2006; 13(3): 310 - 320. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Smidt, D. M. Kuster, G. J. van der Wilt, F. B. Thunnissen, K. J. Van Zee, and L. J. A. Strobbe Can the Memorial Sloan-Kettering Cancer Center Nomogram Predict the Likelihood of Nonsentinel Lymph Node Metastases in Breast Cancer Patients in The Netherlands? Ann. Surg. Oncol., December 1, 2005; 12(12): 1066 - 1072. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. E. Davidson, M. Morrow, D. B. Kopans, and F. C. Koerner Case 35-2005 -- A 56-Year-Old Woman with Breast Cancer and Isolated Tumor Cells in a Sentinel Lymph Node N. Engl. J. Med., November 17, 2005; 353(20): 2177 - 2185. [Full Text] [PDF]
|
|
|
|
|
|
H. M. Kuerer and L. A. Newman Lymphatic Mapping and Sentinel Lymph Node Biopsy for Breast Cancer: Developments and Resolving Controversies J. Clin. Oncol., March 10, 2005; 23(8): 1698 - 1705. [Full Text] [PDF]
|
|
|
|
|
|
R. J. Gray, B. A. Pockaj, and C. R. Conley Sentinel Lymph Node Metastases Detected by Immunohistochemistry Only Do Not Mandate Complete Axillary Lymph Node Dissection in Breast Cancer Ann. Surg. Oncol., December 1, 2004; 11(12): 1056 - 1060. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Cserni, T. Burzykowski, V. Vinh-Hung, L. Kocsis, G. Boross, M. Sinko, M. Tarjan, R. Bori, M. Rajtar, E. Tekle, et al. Axillary Sentinel Node and Tumour-related Factors Associated with Non-sentinel Node Involvement in Breast Cancer Jpn. J. Clin. Oncol., September 1, 2004; 34(9): 519 - 524. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G Cserni, I Amendoeira, N Apostolikas, J P Bellocq, S Bianchi, W Boecker, B Borisch, C E Connolly, T Decker, P Dervan, et al. Discrepancies in current practice of pathological evaluation of sentinel lymph nodes in breast cancer. Results of a questionnaire based survey by the European Working Group for Breast Screening Pathology J. Clin. Pathol., July 1, 2004; 57(7): 695 - 701. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
