Japanese Journal of Clinical Oncology

INTRODUCTION

Axillary lymph node dissection plays a central role in the treatment of patients with invasive breast cancer. However, the clinical behavior of untreated lymph nodes is still unclear. Axillary lymph node dissection is now no longer considered to be the standard treatment in all patients with invasive breast cancer (1,2). The purpose of this study was to identify whether there was a subgroup of patients with invasive breast cancer who may not need to undergo axillary lymph node dissection.

SUBJECTS AND METHODS

Patients (n = 823) with T1N0M0 invasive breast cancer who underwent radical mastectomy or conservative breast surgery at our hospital between 1970 and 1994 were studied. Non-palpable tumors were excluded from the present study. In the subjects studied, we investigated the relationship between axillary lymph node positivity and the following clinico-pathological factors: patient age, menopausal status, contralateral breast cancer (synchronous or asynchronous), tumor location, tumor size (T: cm), histopathology, histological grade, the presence or absence of clustered and/or linear microcalcification or spiculation on mammography and estrogen receptor (ER) status. The histopathology was categorized as invasive ductal carcinoma, invasive lobular carcinoma and the `special' types of carcinoma (medullary, mucinous, tubular and others). All the data including pathological and radiological reports were retrieved from the medical charts of the patients. ER assays were performed using either the dextran-coated charcoal technique (cut-off 10 fmol/mg protein) or enzyme immunoassay (3). The carcinomas were classified as histological grade 1, 2 or 3 according to the grading system reported previously (4). However, not all the data were available for every patient either because the data were missing or the patient had not been tested. The chi-squared test and Student's t-test were used for assessing statistical significance.

RESULTS

In the 823 patients, the mean number of axillary lymph nodes that were dissected was 17.5 (median 16; range 0-99). The incidence of axillary lymph node metastases in T1N0M0 invasive breast cancer was 25% (208/823). The mean patient age at the time of surgery was 50.9 (±10.7) years. The age of the patients was significantly higher in the node-negative group than in the node-positive group (Table 1). Patients who were premenopausal showed a non-significant trend towards a higher rate of lymph node metastases (Table 2). The relationship between tumor size and lymph node involvement is shown in Table 3. The frequency of nodal metastases as related to tumor size was as follows: T [le] 1.0 cm, 17%; T [le] 1.5 cm, 25%; and T [le] 2.0 cm, 29%. The incidence of lymph node metastases ranged from 17 to 29% and was as high as 17% even for tumors that were [le]1cm in diameter. The number of tumors with a diameter of <0.5 cm was too small to be analyzed in this series. Mammography revealed that patients with malignant calcification or spiculation showed a significantly higher rate of nodal metastases than those without these radiographic findings (Table 4). When related to the histopathology, the `special' types of tumors (medullary, mucinous and tubular carcinomas) tended to have a lower rate of lymph node involvement than the common types such as invasive ductal carcinoma (Table 5). With regard to the histological grade, lymph node positivity significantly increased with high-grade tumors (Table 6). Lymph node metastasis could not be detected in any of postmenopausal patients where the histology was favorable and the tumor diameter was [le]1.0 cm without microcalcification or spiculation on mammography. However, the frequency of patients who satisfied this criterion was only 6% (50/823) among T1N0M0 breast cancers. No correlation was observed between any of the other factors (the presence of contralateral breast cancer, tumor location and ER status) and lymph node metastases (data not shown).

Table 1. Relationship between patient age at surgery and axillary lymph node metastasis

	Axillary lymph node metastasis
	Negative	Positive
Patient age at surgery (years)	51.5 ± 11.1	49.4 ± 9.8

P < 0.01.

Table 2. Relationship between patients' menopausal status and axillary lymph node positivity

Menopausal status	Axillary lymph node metastasis
	Negative	Positive
Premenopause	330 (73%)	122 (27%)
Postmenopause	279 (78%)	79 (22%)

NS.

Table 3. Relationship between tumor diameter and axillary lymph node positivity

Tumor diameter (cm)	Axillary lymph node metastasis
	Negative	Positive
[le]1.0	124 (83%)	25 (17%)a
1.0<-[le]1.5	236 (75%)	79 (25%)a,b
1.5<-[le]2.0	255 (71%)	104 (29%)b
[le]2.0	615 (75%)	208 (25%)

^a,bP < 0.01.

Table 4. Relationship between mammographic findings and axillary lymph node positivity

Mammography findings	Axillary lymph node metastasis
	Negative	Positive
None or tumor shadow only	296 (82%)	67 (18%)a,b
Spiculated tumor shadow only	170 (69%)	77 (31%)a
Microcalcifications with or without spicular formation	104 (66%)	54 (34%)b

^a,bP < 0.01.

Table 5. Relationship between histopathological type and axillary lymph node positivity

Histopathological type	Axillary lymph node metastasis
	Negative	Positive
Invasive ductal	524 (73%)	190 (27%)a
Invasive lubular	12 (80%)	3 (20%)
Mucinous	46 (88%)	6 (12%)b
Medullary	17 (100%)	0 (0%)a
Tubular	7 (88%)	1 (12%)b
Others	8 (80%)	2 (20%)

^a,bP < 0.01.

Table 6. Relationship between histological grade and axillary lymph node metastasis

Histological grade	Axillary lymph node metastasis
	Negative	Positive
1	34 (87%)	5 (13%)
2	182 (75%)	60 (25%)
3	217 (68%)	100 (32%)

P < 0.01.

DISCUSSION

The detection of positive nodes alters the postoperative management of invasive breast cancer. Indeed, the presence of histopathological axillary lymph node metastasis is the most useful prognostic factor. The NSABP B-04 trial suggested that axillary dissection does not affect survival (5-7). However, that study has been heavily criticized (8,9) and there is a possibility that inadequate local treatment can adversely affect patient outcome. In accordance with this, some researchers have advocated the usefulness of axillary sentinel lymph node biopsy (10). However, we feel that the procedure is a little complicated for general use in a community hospital setting, because sentinel node biopsy needs injection of labeled colloids followed by scintigraphic scans and a pathologist for frozen section examinations of the dissected lymph nodes.

In the present study, we re-evaluated the need for axillary lymph node dissection in all T1N0M0 tumors. Nodal positivity is known to be higher in larger tumors, even those in theT1N0 category (11). Malignant microcalcification on mammography was a significant predictor for lymph node metastases. Cody (12) also reported that mammography could detect the spread of the carcinoma when compared with manual palpation. In addition, histological grade 3 invasive carcinomas also show a high rate of positive axillary lymph node metastases. Port et al. (8) reported that the presence of lymphovascular invasion was closely associated with lymph node positivity. However, lymphovascular invasion by tumor cells could not be analyzed in our patients.

It was difficult to characterize a subgroup that had an acceptably low risk of nodal involvement, since the impact of each factor on the reduction of lymph node positivity was relatively small. However, our study has demonstrated a significant effect of tumor size, histopathological grade and mammographic findings on the rate of axillary lymph node involvement in T1N0 invasive breast cancer. The age of the patients and menopausal status also had a small but significant impact on lymph node positivity. Since adjuvant chemotherapy has been reported to be more beneficial for younger patients (<50 years of age) with positive nodes, axillary lymph node dissection may be clinically more important in such patients.

It may be possible to avoid axillary lymph node dissection in postmenopausal patients (50 years or older) where the histology is favorable when the tumor diameter is [le]1.0 cm and there is no microcalcification or spiculation on mammography. Since the frequency of patients who satisfied this criterion was low (6%), further work is necessary to establish more useful factors which can predict axillary lymph node involvement preoperatively. It is difficult to make a correct diagnosis of lymph node metastases by using imaging diagnostic procedures such as ultrasound and computed tomography, and this subject is under investigation (13,14).

In conclusion, although axillary lymph node dissection may not be necessary for every patient with a T1N0 tumor, universal abandonment of axillary lymph node dissection cannot be recommended at present. Further studies of new prognostic/predictive factors which are closely linked to axillary lymph node metastases are required.