Article of the Month -- April 1997

THE ROLE OF AXILLARY DISSECTION IN MAMMOGRAPHICALLY DETECTED CARCINOMA

Steven M. Pandelidis, MD,* Kristi L. Peters, MS,† Mbaga S. Walusimbi, MD,* Roger L. Casady, MD,* Shelli V. Laux, RRA, CTR,‡ Sally H. Cavanaugh, PhD,† and Thomas L. Bauer, MD, FACS*

BACKGROUND
Axillary dissection remains a standard component of the treatment of invasive carcinoma of the breast. The presence of metastases to the regional lymph nodes guides adjuvant therapy and aids in determining prognosis. Mammography results in the discovery of small and often node-negative carcinomas of the breast.

STUDY DESIGN
This 15-year, retrospective analysis investigated whether certain patients with small tumors could be spared the morbidity of axillary dissection.

RESULTS
Medical records showed that from January 1980 to May 1995, 4,543 needle localization biopsies were done at York Hospital because of abnormalities detected on mammograms. Of these, 703 (15.5 percent) proved to be carcinoma. Of the carcinomas, 68 percent were infiltrating ductal carcinoma, 26 percent were ductal carcinoma in situ, and 5.4 percent were infiltrating lobular carcinoma. Axillary dissection was done on 588 patients, and 88.1 percent of the patients had no metastases to axillary lymph nodes. No axillary metastases were present in 109 patients with ductal carcinoma in situ who underwent axillary lymph node dissection or in 21 patients with microscopic invasive tumors. Only two of 54 patients with a T1a tumor (tumor [T], £0.5 cm) had positive axillary nodes. Only one of 29 patients with a well-differentiated T1b tumor (T, >0.5 to £1 cm) had metastatic axillary nodes. In the presence of negative axillary lymph nodes, 19.2 percent of patients with a T1a tumor, 33.7 percent of patients with a T1b tumor, 60 percent of patients with a T1c tumor (T, >1 to £2 cm), and 78.9 percent of patients with a T2 tumor (T, >2 cm) were given adjuvant chemotherapy or hormonal therapy.

CONCLUSIONS
Patients with ductal carcinoma in situ and microscopic invasive tumors do not require node dissections. Possibly patients with T1a tumors and patients with well-differentiated, estrogen-receptor positive, progesterone-receptor positive, T1b tumors can also be spared axillary node dissection. By following this approach on occasion, patients with positive nodes might not undergo axillary lymph node dissection, but they may still be offered adjuvant therapy. J. Am. Coll. Surg., 1997, 184: 341-345.

Axillary lymph node dissection is traditionally a necessary component of the treatment of carcinoma of the breast. Knowledge of the status of the lymph nodes aids in prognosis and helps determine the need for adjuvant chemotherapy or adjuvant hormonal therapy. Additionally, the resection of involved lymph nodes may effect a surgical cure, without the aid of adjuvant therapy (1, 2).

The widespread use of mammography has resulted in the detection of small and even microscopic invasive tumors (3, 4). In our previously reported series, axillary lymph nodes contained metastases in only 11.6 percent of all cases (4).

Before 1989, chemotherapy usually was not offered to patients with axillary lymph nodes that were negative for metastases. The results of the National Surgical Adjuvant Breast and Bowel Project (NSABP)-13 and NSABP-14 demonstrated a significant benefit from adjuvant chemotherapy or hormonal therapy in certain patients with uninvolved lymph nodes (5, 6). The increasing number of small, mammographically detected, and often node-negative tumors, along with the tendency to offer adjuvant therapy to patients without positive nodes, calls into question the routine use of axillary lymph node dissection. The current retrospective study was undertaken to try to identify subgroups of patients who could be spared the morbidity of axillary lymph node dissection without compromising treatment outcome.

METHODS
From January 1980 until May 1995, the surgical staff and residents of the York Hospital performed biopsies -- after the patient underwent needle localization by the radiologists -- of lesions detected on mammograms. During the first 11 years of the study, the localization was performed by the free-hand technique. Since 1992, the free-hand and grid methods have been used for localization.

The mammographic abnormalities were categorized as calcifications, nodules or masses, densities, or spiculated lesions. The method of needle localization biopsy has been previously described (3). If calcifications, spiculated lesions, nodules, or masses were the indication for biopsy, specimen mammography was performed to ensure accuracy of the biopsy.

Specimens were oriented by the surgeon and dyed by the pathologist so that margins could be assessed. For carcinomatous lesions, histology, tumor size, histologic grade, nuclear grade, estrogen receptor (ER) status, progesterone receptor (PR) status, and the status of the axillary lymph nodes (if available) were recorded. To evaluate the lymph nodes, pathologists made one section of grossly normal nodes and multiple sections of grossly abnormal nodes. Only those carcinomas discovered by true screening mammograms were included in our study.

After the diagnosis was made, patients were treated according to the standard of care at the time. As expected, in the early years, fewer patients were treated with breast-sparing operative treatment than in latter years. During the early years of the study, patients with a diagnosis of ductal carcinoma in situ (DCIS) often underwent axillary lymph node dissection, but during the final 7 years of the study, only rarely did patients undergo axillary dissection. During the final 6 years of the study, more patients with nodes negative for metastases were treated with adjuvant chemotherapy or hormonal therapy than in earlier years. The clinical alert issued by the National Cancer Institute in May 1988 and the subsequent publication of the NSABP-13 and NSABP-14 no doubt influenced the decision of medical oncologists to offer adjuvant therapy (5, 6).

RESULTS
From January 1980 to May 1995, 4,543 needle localization biopsies were performed at York Hospital or its affiliated outpatient surgery facility. The mean age of patients undergoing the procedure was 62.3 years. Seven hundred three carcinomas were diagnosed by needle-localization biopsy of lesions detected on screening mammography. Four hundred eighty-one tumors were infiltrating ductal carcinoma; 184 were DCIS, and 38 were infiltrating lobular carcinoma. Most carcinomas (84.2 percent) were detected in patients 50 years of age or older. Of the 517 invasive tumors, only 31 were larger than 2 cm (Table I).

Medical records from the early years of the study showed that women with DCIS were often treated by axillary dissection in addition to wide excision and radiation or mastectomy. Of the 184 patients treated for DCIS, 109 underwent axillary lymph node dissection; no metastatic nodes were detected in these patients.

Of the 517 patients with invasive tumors, 476 had an axillary lymph node dissection. Most patients (65 percent) underwent total mastectomy and axillary lymph node dissection rather than wide excision, axillary lymph node dissection, and radiation therapy. All 21 patients with microscopic invasive tumors underwent axillary lymph node dissection; no metastatic nodes were detected in their axillary specimens. The association between size of the tumor and status of the axillary nodes is summarized in Table II.

Twenty-nine patients with well-differentiated T1b tumors (tumor [T], >0.5 to £1 cm) underwent axillary lymph node dissection. Metastatic axillary lymph nodes were found in only one of these patients, whose tumor was ER and PR negative.

Among the patients with well-differentiated T1b tumors, only 4 patients tested negative for ER and PR. The nodes were negative for metastases in the 16 patients with T1b, ER-, PR-positive well-differentiated tumors. The ER and PR status was not known in the remaining 9 patients with well-differentiated T1b tumors. No axillary recurrences have occurred among the 75 patients with DCIS and the 35 patients with infiltrating ductal carcinoma who did not undergo axillary lymph node dissection.

The anatomic distribution of the invasive carcinomas along with information about the status of the axillary lymph nodes is summarized in Table III. As expected, most carcinomas (57 percent) were in the upper outer quadrant. Lower inner quadrant lesions were the least likely to metastasize to axillary lymph nodes, but they still metastasized nearly 10 percent (3 of 32 patients) of the time. The difference in the rates of axillary lymph node metastases on the basis of tumor location was not statistically significant (chi-square test, p=.793).

Patients with microscopic invasive tumors did not receive adjuvant chemotherapy or tamoxifen citrate (Nolvadex, Zeneca Pharmaceuticals, Wilmington, Del). Usually, patients with positive nodes received adjuvant therapy. Some patients with negative axillary lymph nodes received chemotherapy or tamoxifen (Table IV).

DISCUSSION
Axillary dissection potentially benefits patients with carcinoma of the breast for several reasons. The status of the lymph nodes remains the strongest prognostic indicator. The presence of involved axillary lymph nodes usually leads to a recommendation for adjuvant chemotherapy or hormonal therapy. Last, resection of axillary nodes improves local control and may, in fact, increase overall survival.

In a series of 1,624 patients, Recht and colleagues reported an axillary recurrence rate of 2.2 percent in patients treated with level I and II lymph node dissections (7). In the NSABP-4, there was an 18 percent axillary recurrence rate at 10 years in the group of patients treated with total mastectomy only. The investigators in the NSABP-4 concluded that although omission of axillary lymph node dissection led to a higher axillary failure rate, overall survival remained unaffected (8).

Hellman and Harris argued that the NSABP-4 data are flawed because 35 percent of the patients who underwent "total mastectomy" had between 1 and 10 nodes incidentally removed during the procedure (2). Elimination of these patients from the study would weaken the statistical evidence that total mastectomy and total mastectomy plus axillary lymph node dissection produce equivalent survival results (2). Additionally, Feintman and coauthors reported that among long-term survivors (longer than 20 years) of carcinoma of the breast, 35 percent had positive nodes at the time of treatment (1). In their series, patients were treated with operation only. The authors suggest that the resection of the involved nodes leads to the cure (1).

While controversy exists about the contribution of axillary dissection to the treatment of carcinoma of the breast, few would disagree that the procedure adds to the morbidity of operative treatment. Among patients with mammographically detected carcinomas, subgroups may exist in whom the risk of axillary metastases is so small that they can be spared the morbidity of node dissection.

We and other authors have shown that axillary lymph node dissection is unnecessary in mammographically detected DCIS (3, 4, 9). In the York Hospital experience, 109 of 184 patients with DCIS underwent axillary lymph node dissection. No specimens revealed metastatic nodes. All 21 of our patients with invasive microscopic tumors underwent node dissection; no metastatic nodes were detected. Patients with T1a tumors (£0.5 cm.) were only rarely found to have positive nodes (3.7 percent). Silverstein and associates reported a similar rate (3.8 percent) (9).

Among our 54 patients with T1b tumors, 11.7 percent had positive nodes; however, in only one instance did a well-differentiated T1b lesion metastasize to axillary nodes. The tumor was 0.6 cm and ER and PR negative. A patient with such a tumor may well be offered adjuvant chemotherapy even in the absence of positive nodes.

The location of the primary tumor did not aid in predicting the presence or absence of metastases to axillary lymph nodes (Table III).

Table IV illustrates that medical oncologists offer adjuvant therapy to certain patients with node-negative tumors. The larger the tumor, the more likely they are to offer chemotherapy or hormonal therapy. Almost 20 percent of the 52 patients with T1aN0 tumors and almost 34 percent of 166 patients with T1bN0 tumors received adjuvant therapy. Most patients (60 percent) with T1cN0 tumors received chemotherapy or hormonal therapy. Patients with negative nodes were more likely to receive adjuvant hormonal therapy than adjuvant cytotoxic therapy (Table IV). Strict criteria were not used to make the recommendation for adjuvant therapy, but rather those patients with tumors showing poor prognostic features were advised of the risks and benefits of specific types of adjuvant therapy.

Patient age, performance status, and, of course, desires, influence the decision to offer adjuvant therapy. The average age of patients with T1a or T1b tumors who received chemotherapy was 53 years. No patients had well-differentiated tumors, and most tumors were ER or PR negative (75 percent) and aneuploid (83 percent). When considering differentiation, ER status, PR status, ploidy, and S phase, all patients who received chemotherapy had at least three poor prognostic features.

Given the lesser risks of tamoxifen therapy compared with cytotoxic therapy, tamoxifen was more often offered to patients with nodes negative for metastasis. Almost all such patients had ER-positive tumors (98 percent). Because tamoxifen therapy decreases the risk of the development of contralateral carcinoma by 39 percent, and that in the Breast Cancer Prevention Trial high-risk patients without carcinoma are being offered tamoxifen, it does not seem unreasonable to offer tamoxifen to patients with T1a or T1b tumors (10).

Despite the usual good prognosis for patients with T1aN0M0 or T1bN0M0 tumors, approximately 10 percent of such patients still die of metastatic disease (11). By analyzing prognostic features of primary tumors, we tried to offer adjuvant therapy only to those patients at high risk for recurrence.

Based on our series of 517 mammographically detected invasive carcinomas, we can recommend that microscopic invasive tumors (n=21), and perhaps T1a tumors (n=61) and well-differentiated ER- and PR-positive T1b tumors (n=16), do not require an axillary lymph node dissection. Such a policy would have eliminated 98 axillary lymph node dissections (19 percent), and in two patients involved axillary lymph nodes would not have been detected. A larger confirmatory study is necessary to make any clear recommendations about the elimination of axillary dissection in well-differentiated, ER- and PR-positive T1b tumors.

In any large series, on occasion, a small tumor that has regionally metastasized might be missed following a policy of selective lymph node dissection. Given the tendency for our medical oncologists to offer adjuvant treatment to patients with nodes negative for metastases and T1a or T1b tumors with poor prognostic indexes, such patients would probably receive chemotherapy or tamoxifen. The adjuvant therapy may control disease in the axilla as well as operative therapy does.

Treating 98 women with microscopic invasive tumors, T1a tumors, or well-differentiated, ER- and PR-positive T1b tumors by axillary lymph node dissection for the sake of 2 women who have positive nodes seems ill-advised. Even if both women received adjuvant chemotherapy, as a group they would realize a 16 percent reduction in the risk of death from carcinoma of the breast.

REFERENCES

1. Feintman, I. S., Cuzick, J., Millis, R. R., and Hayward, J. L. Which patients are cured of breast cancer. B. M. J., 1984, 289: 1108-1111.
2. Hellman, S., and Harris, J. R. The appropriate breast cancer paradigm. Cancer Res., 1987, 2: 339-342.
3. Bauer, T. L., Pandelidis, S. M., Rhoads Jr, J. E., and Owens, R. S. Mammographically detected carcinoma of the breast. Surg. Gynecol. Obstet., 1991, 173: 482-486.
4. Bauer, T. L., Pandelidis, S. M., and Rhoads Jr, J. E. Five-year survival of 100 women with carcinoma of the breast diagnosed by screening mammography and needle localization biopsy. J. Am. Coll. Surg., 1994, 178: 427-430.
5. Fisher, B., Constantino, J., Redmond, C., et al. A random and clinical trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have estrogen-receptor-positive tumors. N. Engl. J. Med., 1989, 320: 479-484.
6. Fisher, B., Redmond, C., Dimitrov, N., et al. A randomized clinical trial evaluating sequential metho<trexate and fluorouracil in the treatment of patients with node-negative breast cancer who have estrogen receptor negative tumors. N. Engl. J. Med., 1989, 320: 473-478.
7. Recht, A., Pierce, S. M., Abner, A., et al. Regional nodal failure after conservative surgery and radiotherapy for early stage breast carcinoma. J. Clin. Oncol., 1991, 9: 988-996.
8. Fisher, B., Redmond C., Fisher, E. R., et al. Ten year results of a randomized clinical trial comparing radicalmastectomy and total mastectomy with or without radiation. N. Engl. J. Med., 1985, 312: 674-681.
9. Silverstein, M. J., Giersen, E. D., Waisman, J. R., et al. Axillary lymph node dissection for T1a breast carcinoma. Is it indicated? Cancer, 1994, 73: 664-667.
10. Early Breast Cancer Trialists Collaborative Group. Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy. Lancet, 1992, 339: 1-15, 71-85.
11. Rosen, P. P., Groshen, S., Saigo, E., et al. A long term follow-up study of survival in stage I (T1N0M0) and stage II (T1N1M0) breast carcinoma. J. Clin. Oncol., 1989, 7: 355-366.

_________________________

Journal of the American College of Surgeons
April 1997, Volume 184, Number 4

Back to JACS Menu