The histological status of the axillary lymph nodes is the most important prognostic factor in patients with breast cancer.1 The sentinel lymph node biopsy (SLNB) has proved to be a reliable alternative to the traditional axillary lymph node dissection (ALND) with regard to predicting the histological status of the remaining axillary lymph nodes in clinical T1/2 N0 breast cancer.2–6 The SLNB has the advantage of reduced postoperative morbidity compared with ALND.7 In case of a positive sentinel lymph node (SLN), a complementary ALND is recommended to maximize regional control and complete axillary staging.

Several validation studies of SLN biopsies followed by ALND in breast cancer patients have been published. All these studies report the risk of false-negative sampling, with rates varying from 0% to 22%.2,6,8–12 A meta-analysis of 13 studies including 912 patients reported a false-negative rate of 5.1%.13 Once the validation phase is completed, an unknown number of patients with undetected tumor-positive nodes at SLNB do not undergo an ALND. Undetected tumor-positive nodes of clinical importance are those that lead to axillary recurrence.14

Several questions arise considering axillary relapse. In the setting of a negative SLNB, it would be interesting to identify prognostic factors for the incidence of axillary relapse, especially regarding prevention. The clinical consequences for the patient are unclear, and the nature of subsequent therapy is still open for debate.

The aim of this study was to identify the extent of this problem in current practice. The clinical consequences for patients with recurrent axillary disease were clarified. Furthermore, a systemic review of the literature was performed to determine incidence, patient and tumor characteristics, and subsequent therapy.

METHODS

Between January 1998 and December 2003, 696 consecutive patients had an SLNB for clinical T1/2 N0 breast cancer in a regional teaching hospital. After a validation phase, in which 20 patients underwent SLNB with an ALND in the same procedure, patients with a tumor-free SLN did not undergo an ALND. In any case of tumor involvement of the SLN, an ALND was performed. The follow-up consisted of a physical examination every 3 months during the first 2 years and subsequently every 6 months. All data were collected in a prospective database. The median age of these patients was 57 years. The median tumor size was 16 mm. The primary tumor was Tis in 22 patients (3%), T1 in 390 patients (56%), and T2 in 243 patients (35%). The histological tumor type was invasive ductal cancer in 73% and invasive lobular cancer in 13% of the patients (Table 1).

TABLE 1 Patient and tumor characteristics of 696SLNB patients
Full size table

Lymphatic Mapping and Operative Procedures

The SLN procedure was performed with 60 MBq of 99mTc nanocolloid as a radioactive tracer and 2 mL of blue dye (Bleu Patente V; Guerbet, Aulnay-sous-Bois, France) for lymphatic mapping. The SLN was identified and harvested during surgery guided by lymphoscintigraphy, the blue lymphatic vessels, and detection of radioactivity by the gamma probe.

Pathologic Examination of the SLN

The SLN was bisected, after which both halves were embedded in paraffin. Each part was step-sectioned at 500-μm intervals at three levels and stained with hematoxylin and eosin. Immunohistochemical staining was performed with Cam 5.2 (Becton Dickinson, San Jose, CA).

Review of the Literature

To determine the axillary relapse rate after a negative SLNB for breast cancer, a systematic review of the literature was performed. PubMed and the Cochrane library were searched with the use of the Medical Subject Heading terms “breast neoplasms” and “sentinel lymph node biopsy.” This pair was linked to the terms “neoplasm recurrence,” “treatment outcome,” and “diagnostic errors.” This search strategy resulted in 221 titles. Only 11 studies dealt with follow-up of a cohort of SLN-negative patients or a case report on axillary relapse after negative SLNB in breast cancer patients. Two other studies were found through links and references.

RESULTS

At least 1 SLN could be identified in 676 (97.1%) of 696 patients. The median number of harvested SLNs was 2 (range, 0–9). In 237 (35%) of the 676 patients, the SLN contained metastatic disease. In 86 patients, this concerned micrometastases. In 6 of these 86 patients with micrometastatic disease, an ALND was omitted.

After a median follow-up of 26 months (range, 1–90 months), an axillary recurrence was detected in 2 patients out of 439 with a negative SLNB. The incidence of axillary relapse after tumor-negative SLNB was therefore .46%.

In one patient, physical examination revealed axillary lymph node recurrence 4 months after the SLNB. The ALND specimen contained two tumorous lymph nodes. The patient received an aromatase inhibitor. In a second patient, axillary relapse was detected by routine physical examination 27 months after the SLNB. She underwent an ALND and ovariectomy and received tamoxifen. The SLNs of these two patients were re-examined but did not reveal any metastasis. Patient and tumor characteristics concerning these two patients are listed in Table 2.

TABLE 2 Patient and tumor characteristics of all patients in the literature with recurrent axillary disease
Full size table

In a third patient with a .2-mm micrometastasis in the SLN, ALND was omitted. Axillary recurrence was resected 22 months after the SLNB. Tumor was found in the axillary fat and was not related to any preexisting lymph node structure. No technical problems were met during the ALND.

PubMed and the Cochrane library search resulted in 10 studies concerning the follow-up of a cohort of SLN-negative patients with breast cancer and in 3 case reports on axillary recurrence. The results of a total of 3184 patients (including the present series) with a median follow-up of 25 months (range, 16–46 months) were pooled. In eight patients, an axillary relapse was diagnosed. This resulted in an axillary recurrence rate of .25% (Table 3). Axillary relapse after negative SLNB of all 11 published cases occurred after a median of 21 months. The data concerning these patients are listed in Table 2.

TABLE 3 Incidence of axillary recurrence after negative sentinel lymph node biopsy in breast cancer patients
Full size table

DISCUSSION

Axillary recurrence after a negative SLNB in breast cancer patients is, at .46%, rare in this group. The mean axillary relapse rate in comparable studies is equally low at .25%. These rates are far lower than would be expected if compared with the false-negative rates of the SLNB in the validation phase; a meta-analysis reported a false-negative rate of 5.1%.13

These results are supported by follow-up studies of clinically node-negative breast cancer patients in whom surgical axillary staging was omitted. A population-based study showed that 34% of the axillary lymph nodes of clinical stage I breast cancer patients contain metastases.15 In contrast with these findings, Greco et al.16 and Fisher et al.17 demonstrated that only 6.7% to 17.8% of the patients without ALND developed axillary recurrence after a follow-up period of 5 to 10 years. Axillary relapses were detected after a median period of 14.7 to 31 months. Hence, substantially fewer clinical recurrences were observed than would be expected on the basis of data reported in literature.

Several factors can explain the difference between the false-negative rate of the SLNB in the validation phase and the axillary relapse rates, as well as the lower than expected axillary recurrence rate after omitting ALND. According to the studies by Greco et al.16 and Fisher et al.,17 axillary relapse is to be expected, if it occurs, after a median of 14.7 to 31 months at a follow-up of 63 to 126 months. The follow-up period of the studies in the series in Table 3 amounted to a median length of only 16 to 46 months and might therefore be too short to lead to comparable results.

In contrast to earlier series, most patients currently receive adjuvant systemic treatment because of tumor and patient characteristics. Adjuvant chemotherapy has proved to destroy metastases in tumor-bearing axillary nodes and therefore can be expected to decrease axillary relapse rates.18

Another cause for the low relapse rate might be the decreasing incidence of failure to identify the SLN after the learning phase. A study with a longer validation phase showed an increase in identifying the SLN from 67% with 18 patients to 96% with 177 patients.19 The false-negative rates from the published studies always represent the validation phase. The studies reporting on the follow-up of SLN-negative patients have always passed this phase.

The young age of the patients with axillary recurrence is remarkable; almost all patients in literature are younger (median, 46 years) than the median age in this series (median, 57 years). This corresponds with the median age of 48 years of patients with axillary relapse after ALND.20

The clinical consequences for patients with axillary relapse after a negative SLNB are yet unclear, but similarities to patients with axillary recurrence after ALND are hard to overlook. In both groups, ap-proximately 30% of the patients with axillary recurrence present with simultaneous locoregional or systemic failure. Approximately 50% of the patients with axillary relapse after ALND develop distant metastatic disease. This suggests an ominous prognosis for patients with axillary relapse after a negative SLNB.20

It is tempting to consider axillary relapse as a presentation of formal locoregional recurrence. A patient with an axillary recurrence should therefore receive therapy for locoregional failure.

In conclusion, axillary recurrences after negative SLNB occur, but at a much lower rate than would be expected on the basis of historical figures and false-negative SLN findings. Considering the similarities to axillary relapse, subsequent therapy should be aimed at locoregional and systemic control.