Lymph Nodes in Breast Cancer: So Who’s Counting?

Abstract & Commentary

Source: Camp RL, et al. Cancer 2000;88:108-113.

Various cytokines may play a role in both local, tumor-induced angiogenesis as well as downstream lymph node genesis and hyperplasia. Camp and colleagues hypothesize that prognostic information might be available by evaluating the downstream effects of malignant tumors. Expression of changes in the lymph nodes could be characterized by either hyperplasia or by an increased number of tumor-free lymph nodes. Because information on the size of lymph nodes is not as readily available, a study was undertaken to evaluate the relationship between the number of uninvolved lymph nodes present and prognosis.

The study included 290 patients with breast cancer whose surgery occurred between July 1, 1983 and July 1, 1993 at Yale-New Haven Hospital. Only node-negative patients with T1-T2 lesions were included. With a median follow-up of 103 months, these early-stage patients had an overall survival of 86.3% and a five-year survival of 93.6%. Treatment was provided by 35 different surgeons, and the specimens were prosected by 77 surgical pathology residents. For this analysis, the tumor size and lymph node number were taken from the original description, but a single pathologist scored the histopathologic features of the primary specimen in a blinded fashion.

The relationship between the number of lymph nodes present and survival in these node-negative patients was examined. With a median follow-up of 103 months, the five-year survival was 84.7% for patients with 20 or more lymph nodes compared with 96.3% if less than 20 nodes were present. In addition to the number of lymph nodes, a univariate analysis also revealed that survival was related to tumor size, grade, and the presence of necrosis. Both tumor size and the presence of 20 or more lymph nodes retained significance in the multivariate analysis. The link between survival and negative nodes may be cytokines which act not only locally to adversely affect the biology of the tumor but also cause downstream lymphogenesis.

COMMENT By Kenneth W. Kotz, MDM

In this study, a statistical analysis suggested that tumor size (a known factor for survival) was not spuriously associated with the number of lymph nodes removed. For example, larger tumors were not associated with "large" axillary dissections or with an increased number of lymph nodes removed. (Based on recorded dimensions, the cubic volume of the axillary specimen was considered "large" if its size was in the top 25% of all specimens.) Furthermore, the number of lymph nodes removed was not related to either the size of the axillary dissection or the surgeon performing the operation. These results suggest that tumor size did not influence the surgeon to augment the axillary dissection.

Both the level of axillary dissection and the pathologist’s preparation of the specimen might affect the results of this study. In this study, a median of 15 lymph nodes were removed, but approximately one-fourth of patients had less than 11 lymph nodes removed. Some of these may have been limited dissections, which is associated with less accurate staging.1 For a clinically negative axilla, the standard level I-II dissection should include those nodes located lateral to and behind the pectoralis minor and located within the axillary triangle (bordered by the axillary vein superiorly, serratus anterior medially, and the latissimus dorsi laterally).2 Approximately 18 nodes would be expected from a level I-II dissection 3 although ranges of 0 to 70 have been reported.2

The routine technique to identify lymph nodes in the axilla for a pathologic assessment involves a manual dissection.2,4 The use of fat-clearing fixative, as was used in the majority of cases in this study, increases the number of lymph nodes identified, but has not been associated with significant upstaging.2,4 Because identifying low numbers of lymph nodes is associated with understaging, my institution performs lymph node enhancement only when less than 10 lymph nodes are found. As can be seen, the pathologist’s approach to evaluating the axilla can vary and might change any relationship between lymph node number and survival.

This analysis was restricted to breast cancer patients with T1-T2 tumors whose lymph nodes were negative for malignancy. It would be interesting to test whether the total number of lymph nodes would provide prognostic information for T3-T4 tumors, and compare the level of significance with less advanced tumors. Additionally, I wonder what relationship would emerge between the total number of lymph nodes and prognosis even if some of the lymph nodes contained metastatic deposits. Finally, is this phenonenom restricted to breast cancer? Colon cancer and melanoma are examples of other malignancies where lymph node dissections have been routinely performed and a similar retrospective analysis could be undertaken. Of course, the increasing use of sentinel lymph node biopsies may make unavailable any real prognostic information that the total number of lymph nodes provides.

So who’s counting? If, as the title of this article suggests, "a high number of tumor-free axillary lymph nodes from patients with lymph node negative breast carcinoma is associated with a poor outcome," we may be counting lymph nodes, even when node-negative.


1. Gadd M, et al. Diseases of the Breast Updates 1998;2: 1-10.

2. Silverberg S, et al. In: Silverberg S, ed. Principles and Practice of Surgical Pathology and Cytopathology. New York, NY: Churchill Livingstone Publishing; 1997: 578-579.

3. Veronesi U, et al. Eur J Surg Oncol 1990;16:127-133.

4. Tavassoli F. Pathology of the Breast. 2nd ed. Norwalk, CT: Appleton & Lange; 1999:107-108.

From the study by Camp et al on the relationship between lymph node number and survival in breast cancer, which of the following is true?

a. Larger axillary dissections contained an increased number of lymph nodes.

b. Patients were randomized to different levels of axillary lymph node dissection.

c. Patients with the most aggressive axillary lymph node dissections had the best survival.

d. Patients with less than 20 lymph nodes survived longer than patients with more than 20 lymph nodes.