|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Prognostic Significance of the Number of Axillary Lymph Nodes Removed in Patients With Node-Negative Breast CancerByFrom the Radiation Therapy Program, Breast Cancer Outcomes Unit, and Population and Preventive Oncology, British Columbia Cancer Agency and University of British Columbia, Vancouver, British Columbia, Canada. Address reprint requests to L. Weir, MD, British Columbia Cancer Agency, 600 West 10th Ave, Vancouver, British Columbia V5Z 4E6, Canada; email: lweir{at}bccancer.bc.ca
PURPOSE: The objective of the study was to evaluate the association between the number of lymph nodes removed at axillary dissection and recurrence and survival for patients with node-negative invasive breast cancer.
PATIENTS AND METHODS: Subjects were 2,278 women with pathologically node-negative invasive breast cancer, diagnosed from 1989 to 1993 in British Columbia, Canada. Women aged RESULTS: For patients not receiving systemic therapy, regional relapse was significantly increased with smaller numbers of nodes removed (P = .03). There was a trend toward shorter overall survival with fewer nodes removed (P = .06). Node-negative patients who received systemic therapy did not have a higher regional relapse rate or shorter overall survival when fewer nodes were recovered. CONCLUSION: Recovery of a small number of negative lymph nodes at axillary dissection likely understages patients and leads to undertreatment, resulting in an increased regional relapse rate and poorer survival. The use of systemic therapy may overcome this effect. The number of nodes removed, in conjunction with other prognostic factors, may be useful in selecting node-negative patients for systemic therapy.
AXILLARY LYMPH NODE dissection (ALND) is still considered to be a standard part of the management of operable breast cancer because of the significant prognostic information it imparts and the increasing refinement in the use of adjuvant therapies.1 The development of sentinel node biopsy and the increasing use of adjuvant systemic therapy provide a rationale for omission of ALND, but it has not yet been proven that these approaches result in equivalent survival and axillary control. Currently, the majority of women still undergo ALND as part of their initial management. There is debate about the optimum extent of axillary surgery to provide the necessary prognostic information and axillary control. Previously published studies suggest that the recovery of  10 nodes is adequate,2,3 whereas others show that a more limited axillary sampling that recovers a mean number of nodes of four to six is sufficient.4,5 It has also been shown that patients with positive nodes have a higher locoregional relapse rate when few nodes are examined.6 The present study evaluates the association between the number of axillary nodes recovered and outcomes for patients with pathologically negative nodes. A unique aspect of this study is that two groups of node-negative patients were analyzed, those receiving and those not receiving systemic therapy.
Subjects were node-negative patients with invasive breast cancer diagnosed between January 1, 1989, and December 31, 1993, that were identified from the Breast Cancer Outcomes Unit database of the British Columbia Cancer Agency (BCCA). This database contains detailed demographic, staging, treatment, and outcome information for women referred to the BCCA. Eligible patients were younger than 90 years of age at the time of diagnosis, were not clinically or pathologically pure in situ, T4, N1, N2, or M1, had survived at least 30 days from the time of diagnosis, had an axillary dissection with at least one node recovered, did not have bilateral breast cancer, and did not have axillary irradiation. Patients referred to the BCCA had surgery performed in approximately 30 different hospitals in British Columbia. Processing of the pathologic specimen was not standardized throughout the province but was performed in accordance with usual surgical pathology practice. The pathologists examined the gross specimen for lymph nodes, and all possible nodes were submitted for microscopy. Generally, nodes smaller than 4 mm were submitted unsectioned and larger nodes were sliced at 3- to 5-mm increments and submitted for microscopy. Immunohistochemical analysis for occult metastases was not performed routinely. The majority of patients also had a central pathology review performed by a BCCA pathologist. Sentinel node biopsies were not performed on any cases during the study period. Local relapse was defined as any relapse within the breast or chest wall. Axillary relapse was defined as any relapse within the ipsilateral axilla. Regional relapse was defined as any relapse involving the ipsilateral regional nodal areas, including the axilla, supraclavicular, and internal mammary. Systemic relapse was defined as any distant relapse or death from breast cancer, and overall survival was defined as not dead from any cause. Relapse-free and survival time were calculated from the date of diagnosis. Prognostic factors abstracted were age at diagnosis, tumor grade (nuclear grade or histologic grade using the modified Scarff-Bloom-Richardson system),7 size of the primary tumor (maximum histologic or gross pathologic size in millimeters or, if unavailable, the clinical size from a preoperative mammogram or notes of the referring surgeon), lymphatic, vascular, or perineural space (LVN) invasion in the tumor (absent or present), and estrogen receptor (ER) status (negative or positive).
Relapse rates were estimated using the Kaplan-Meier method. The statistical significance of the difference between survival times was determined by the log-rank test8 in univariate analysis. Multivariate tests of the effect of number of nodes examined were performed with Cox proportional hazards analysis. The number of nodes recovered was categorized as one to three, four to nine, 10 to 19, and more than 19 for the tables and log-rank test. In contrast, in the Cox analyses, the number of nodes recovered was used as a continuous variable. Cox analyses were conducted in two stages.9 In the first stage, all variables selected as potential prognostic factors of relapse and overall survival were included in the Cox survival model. Variables included were age at diagnosis and tumor size (continuous variables), LVN invasion, tumor grade, and ER status (categorical variables). The distribution of the number of nodes removed and tumor size were not normally distributed; therefore, a natural log transformation of these continuous variables was performed. Using forward selection, the best-fitting prognostic model for each index of outcome was constructed. In the Cox model, cases had to be removed from analyses when there were missing values in at least one variable. As there were a number of missing values in some of the prognostic variables, the model was rebuilt after the variables that were not significant (P > .10) were omitted. Assumptions of the Cox regression analyses were verified. In the second stage of the analysis, a model containing significant factors and the number of nodes removed was fitted. The difference in the fit to the data of the two models provided a measure of statistical significance of whether the number of nodes removed contained additional prognostic information. The a priori assumption was that the number of nodes examined and outcome were related if there was a statistically significant (P
There were 2,278 patients eligible for analysis. There were 1,468 patients who did not receive systemic therapy (NST group), and 810 who received systemic therapy (ST group). Median follow-up was 7.5 years. The frequency of the univariate variables is listed in Table 1. Node-negative women who received systemic therapy were younger, had larger tumors, and were more likely to have LVN invasion and grade 3 histology, consistent with provincial treatment guidelines in effect at the time.10,11 Table 2 contains a summary of results. The median number of nodes examined was 10, with a range of one to 38.
Local Relapse The local relapse rate for the entire group at 5 years was 5.3%. Approximately half of the local recurrences were in the breast, and the other half were on the chest wall (data not shown). Increasing age was associated with a lower rate of local relapse in the ST group. The number of nodes removed was not significantly associated with local relapse in either the ST group or the NST group.
Axillary Relapse
Regional Relapse
Systemic Relapse The systemic relapse rate for the entire group at 5 years was 10.5%. In the NST group, higher systemic relapse rates were associated with increasing tumor size (P < .001), LVN invasion (P = .05), and poorly differentiated histology (borderline significance, P = .06). In the ST group, systemic failure was associated with poorly differentiated histology (P = .02), and increasing tumor size was of borderline significance (P = .06). Number of nodes examined was not significantly associated with rates of systemic relapse in either the ST group or the NST group (Fig 2).
Overall Survival Overall survival for the entire group at 5 years was 89.4%. Shorter overall survival was associated with fewer nodes removed (Fig 3A) (P = .03). In the NST group, shorter overall survival was associated with increasing tumor size (P < .001) and higher age at diagnosis (P < .001). There was a trend to shorter overall survival with fewer nodes examined (Fig 3B) (P = .06). In the ST group, shorter overall survival was associated with higher age at diagnosis (P < .001) and negative ER status (P = .001), but number of nodes examined was not significant (Fig 3C) (P = .571).
The importance of axillary lymph node status as a prognostic factor has long been recognized. The usefulness of this information must be balanced against the morbidity of ALND. Hack et al12 studied 222 women with a previous ALND. A minimum of 6 months had passed since the surgery. They found that 72% experienced arm/shoulder pain, weakness, or numbness. Range of motion of the arm/shoulder was impaired in 73% of the women. There is evidence that the use of ALND is declining in the United States and that its omission may be associated with poorer outcomes. Du et al13 examined Surveillance, Epidemiology, and End-Results program data and found that there was an increase in the number of women not having an ALND over the decade from 1983 to 1993. They found 19.3% of women had neither an ALND nor radiation after breast-conserving surgery. Bland et al14 conducted a retrospective review of the National Cancer Data Base and found that women with stage I disease had poorer relative 10-year survival when ALND was not performed. Orr15 performed a meta-analysis of six randomized trials comparing standard surgery (mastectomy/ALND or segmentectomy/ALND plus breast irradiation) with standard treatment without ALND. Trials comparing axillary radiation to no axillary treatment were not included. A survival advantage from axillary dissection was seen in all six trials, ranging from 4% to 16%. A Bayesian combination of the six trials demonstrated that ALND confers an absolute survival advantage of 5.4%, with a probability of benefit of greater than 95.5%. This was equivalent to a relative reduction in deaths ranging from 7% to 46%.
Strategies for obtaining prognostic information without an ALND are being used increasingly. Sentinel node biopsy, when performed by experienced surgeons, has a high accuracy. Krag et al16 reported an accuracy of 97% in a series of 443 patients. However, the false-negative rate for the 11 surgeons in this study varied from 0% to 28.6%. It is also possible to predict the likelihood of axillary node positivity using pathologic information from the primary tumor. However, even with primary tumors Currently, therefore, a significant number of women continue to have an ALND. The number of lymph nodes needed to provide an accurate assessment of the axilla, however, has been debated. Steele et al4 have published results of a randomized study comparing lower axillary sampling using a standard technique versus total axillary clearance in patients undergoing total mastectomy. In the node-sampling group, the surgeon sought nodes by inspection and palpation of the axillary tail and contiguous fat. The dissection continued until four nodes were identified and submitted separately for histologic examination. In the node-sampling group, the mean number of nodes identified was 4.8, and 42% of patients had positive nodes. In the axillary clearance group, the mean number of nodes identified was 20.6, and 40% of patients had positive nodes. A subgroup of 67 patients who had sampling performed went on to have a clearance, which resulted in a mean number of additional nodes found of 15. This additional surgery did not change the nodal status in any patient. The authors conclude that when a standard protocol is used for sampling of four nodes, the accuracy of staging the axilla is equivalent to full axillary clearance.
Investigators in Denmark examined the effect of extent of axillary node dissection in a group of 7,145 patients enrolled onto their "low-risk" protocols.2 All patients had negative nodes and received no adjuvant systemic therapy. They found a highly significant correlation between the number of nodes examined and axillary recurrence-free survival, overall recurrence-free survival, and overall survival at a median observation time of 76 months. With
Recht et al6 reported locoregional failure rates in 2,016 patients with positive nodes who received systemic therapy without radiation. On multivariate analysis, number of nodes examined (two to five, six to 10, and The present study evaluates pathologically node-negative patients in two groups: those who did not receive systemic therapy, and those who did. In the NST group, there is a statistically significant increase in rates of regional recurrence for patients with fewer nodes removed as compared with those with more nodes removed. When the number of nodes removed was analyzed independently, there was a strong trend toward lower overall survival as well (P = .06). This effect of number of nodes examined is not seen in patients receiving systemic therapy, suggesting that the lack of prognostic accuracy with small numbers of nodes examined is compensated for by the use of adjuvant systemic therapy. When few nodes are found in the surgical specimen, it is either because an adequate dissection was not performed, the nodes were not found at gross pathologic examination, or because of anatomic variation only a few nodes were actually present. The patients in this study had their initial pathology reported at a variety of hospitals; therefore, some variation in the extent of surgical dissection and in the thoroughness of evaluation of the surgical specimen is expected. This is reflected in the fairly low median number of nodes examined, which was 10 (range, one to 38). With very detailed examination of axillary dissection specimens, a mean 20 to 25 nodes can be found in node-negative patients.21 Also, in this study, pathologic evaluation of lymph nodes did not routinely include immunohistochemistry. It is well known that more detailed evaluation of lymph nodes with immunohistochemistry will identify tumor cells missed on routine histologic assessment and result in upstaging. The significance of these missed micrometastases, however, remains unclear.22,23 The implication of a small number of nodes examined is that it reduces the prognostic value of the negative nodal status. For a variety of reasons, positive nodes have likely been missed in some of these patients, resulting in more regional relapses and shorter survival. This effect seemed to be largely overcome by the addition of systemic therapy. The number of nodes retrieved should therefore be considered along with other important prognostic factors such as tumor size, grade, and lymphatic or venous invasion, when making decisions about the use of systemic therapy.
1. Clinical practice guidelines for the care and treatment of breast cancer: A Canadian consensus document. Can Med Assoc J 158:S22-S26, 1998 (3 suppl) 2. Axelsson C, Mouridsen H, Zedeler K: Axillary dissection of level I and II lymph nodes is important in breast cancer classification. Eur J Cancer 28A: 1415-1418, 1992 3. Sosa J, Diener-West M, Gusev Y, et al: Association between extent of axillary lymph node dissection and survival in patients with stage I breast cancer. Ann Surg Oncol 5: 140-149, 1998[Abstract] 4. Steele R, Forrest A, Gibson T, et al: The efficacy of lower axillary sampling in obtaining lymph node status in breast cancer: A controlled randomized trial. Br J Surg 72: 368-369, 1985[Medline] 5. Forrest A, Everington D, McDonald C, et al: The Edinburgh randomized trial of axillary sampling or clearance after mastectomy. Br J Surg 82: 1504-1508, 1995[Medline] 6. Recht A, Gray R, Davidson N, et al: Locoregional failure 10 years after mastectomy and adjuvant chemotherapy with or without tamoxifen without irradiation: Experience in the Eastern Cooperative Oncology Group. J Clin Oncol 17: 1698-1700, 1999 7. Elston C, Ellis I: Pathological prognostic factors in breast cancer: I. The value of histopathological grade in breast cancer—Experience from a large study with long-term follow-up. Histopathology 19: 403-410, 1991[Medline] 8. Peto R, Pike M, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Br J Cancer 34: 585-612, 1976[Medline] 9. Simon R, Altman D: Statistical aspects of prognostic factor studies in oncology. Br J Cancer 69: 979-985, 1994[Medline]
10.
Olivotto I, Coldman A, Hislop G, et al: Compliance with practice guidelines for node-negative breast cancer. J Clin Oncol 15: 216-222, 1997 11. Sawka C, Olivotto I, Coldman A, et al: The association between population-based treatment guidelines and adjuvant therapy for node-negative breast cancer. Br J Cancer 75: 1534-1542, 1997[Medline]
12.
Hack T, Cohen L, Katz J, et al: Physical and psychological morbidity after axillary lymph node dissection for breast cancer. J Clin Oncol 17: 143-149, 1999 13. Du X, Freeman J, Goodwin J: The declining use of axillary dissection in patients with early stage breast cancer. Breast Cancer Res Treat 53: 137-144, 1999[CrossRef][Medline] 14. Bland K, Scott-Connor C, Menck H, et al: Axillary dissection in breast-conserving surgery for stage I and II breast cancer: A National Cancer Data Base study of patterns of omission and implications for survival. J Am Coll Surg 188: 586-596, 1999[CrossRef][Medline] 15. Orr RK: The impact of prophylactic axillary node dissection on breast cancer survival: A Bayesian meta-analysis. Ann Surg Oncol 6: 109-116, 1999[Abstract]
16.
Krag D, Weaver D, Ashikaga T, et al: The sentinel node in breast cancer. N Engl J Med 339: 941-946, 1998 17. Fein D, Fowble B, Hanlon A, et al: Identification of women with T1-T2 breast cancer at low risk of positive axillary nodes. J Surg Oncol 65: 34-39, 1997[CrossRef][Medline] 18. Dees C, Shulman L, Souba W, et al: Does information from axillary node dissection change treatment in clinically node negative patients with breast cancer? Ann Surg 226: 279-287, 1997[CrossRef][Medline] 19. Mustafa I, Bland K: Indications for axillary dissection in T1 breast cancer. Ann Surg Oncol 5: 4-8, 1998[Abstract] 20. Olivotto I, Jackson J, Mates D, et al: Prediction of axillary lymph node involvement of women with invasive breast cancer. Cancer 83: 948-955, 1998[CrossRef][Medline] 21. Morrow M, Evans J, Rosen P, et al: Does clearing of axillary lymph nodes contribute to accurate staging of breast carcinoma? Cancer 53: 1329-1332, 1984[CrossRef][Medline] 22. Hermanek P, Hutter R, Sobin L, et al: Classification of isolated tumor cells and micrometastasis. Cancer 86: 2668-2673, 1999[CrossRef][Medline] 23. Page D, Anderson T, Carter B: Minimal solid tumor in-volvement of regional and distant sites. Cancer 86: 2589-2592, 1999[CrossRef][Medline] Submitted August 1, 2000; accepted December 20, 2001.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
.05) improvement in model fit resulting from the entry of the number of nodes examined. All reported P values were two-sided. Because the prognostic effect of the number of nodes recovered was hypothesized to be different for patients receiving systemic therapy, models were constructed separately for cases with and without systemic therapy. 
