Use of axillary lymph node dissection (ALND) in patients with micrometastatic breast cancer

Abstract

Background:

Sentinel lymph node (SLN) biopsy is the current prognostic tool for clinically node-negative breast cancer patients. If the SLN reveals macrometastasis, axillary lymph node dissection (ALND) is recommended. However, the use of ALND in patients with micrometastasis is debated. The objective of this study was to assess the utilization of ALND in the treatment of micrometastatic breast cancer.

Methods:

An IRB approved, retrospective study of a pooled dataset of breast cancer patients with micrometastatic disease on SLN biopsy was performed. Patients diagnosed from 1999-2016 were identified via query of a single-institution National Comprehensive Cancer Network (NCCN) breast cancer database as well as a prospective tumor board.

Results:

A total of 91 patients were diagnosed with micrometastatic nodal disease. The median age at diagnosis was 56 y (range: 31-85); median follow-up time was 47 mo (range: 0-203 mo). 42/91(46.2%) patients had ALND of which 37/42 (88.1%) were a second operation; 3/42(7.1%) patients had additional positive nodes found at ALND. 44/91 (48.4%) patients received radiation. 7/91 (7.7%) patients had a recurrence, 5/7 local, including one axillary (2.1%; patient declined ALND).

Conclusions:

Given that the risk of lymphedema after ALND ranges between 20%-53%, the morbidity of ALND may far exceed the likelihood of detecting further nodal involvement in women with micrometastatic disease: 7.1% in this series.

Introduction

Nodal involvement in breast cancer is an important predictor of survival. Currently, the standard procedure to pathologically stage clinically node-negative patients involves a sentinel lymph node (SLN) biopsy.¹ This procedure entails locating, removing, and determining the presence of cancer cells in a lymph node by injection of a nuclear dye and/or a blue dye into the breast. These dyes travel through the breast and deposit in the lymph nodes receiving the majority of lymphatic drainage from the breast, and thus theoretically are the most likely nodes to demonstrate metastatic axillary disease, hence the term “sentinel node.” The nodes picking up the blue or nuclear dye are detected intraoperatively and resected for pathologic review.² When a positive SLN is found containing micrometastases (deposits of tumor cells in lymph nodes defined as >0.2 mm and ≤2.0 mm), the chances of metastases in other nonsentinel nodes (NSLN) historically is 0%-20%.³ Several predictive models are used in the clinical setting to estimate the risk of additional nodal involvement on nonsentinel nodes, but none specifically address micrometastatic disease. In addition, no model has been able to predict a zero percent chance of additional nodal involvement beyond the SLN.^3–9

The results of this biopsy, along with the phenotypic characteristics of the tumor, help determine the course of treatment, including hormone therapy, radiation, and axillary lymph node dissection (ALND). Axillary lymph node dissection is used for predicting prognosis, improving locoregional control, and assisting in staging; however, it does not confer a survival benefit.^10–12 Patients receiving ALND could have distant metastasis even in the absence of locoregional recurrence or initial node negativity^12,13; thus, the operation of surgically clearing the nodes would have no impact on cancer-related mortality. Furthermore, the course of adjuvant therapy is determined more by the immunohistochemistry of the tumor than by the ALND.¹⁰ Currently, the use of ALND in patients with clinically node-negative lymph nodes is debated due to high rates of longterm morbidity from axillary dissection and low rates of recurrences.^1,3,13–17 Given the potential morbidities of ALND, including lymphedema, infection, peripheral neuropathy, and pain, several studies favor omitting ALND in patients with favorable tumor characteristics and low risk of further nodal involvement.^3,10,14,15 Another study suggests that SLNB alone is equivalent to ALND in terms of locoregional control and overall survival.¹⁸ However, the 2016 NCCN Breast Cancer guidelines still support the use of ALND even in patients with pathologically positive but clinically node-negative axillas.¹⁹

Given the current controversies surrounding micrometastatic axillary disease and the inconsistent use of axillary dissection, the primary goal of this study was to evaluate the utilization and results of ALND as part of a multidisciplinary treatment plan for micrometastatic breast cancer.

Methods

This is a retrospective study of a single-institution series of consecutive patients found to have micrometastatic disease after surgery for primary breast cancer. Both the institutional NCCN breast cancer database and a prospective tumor board database were queried for pathologic micrometastatic cases after definitive surgical staging. Institutional Review Board permission was obtained before data collection and analysis. All patients were diagnosed between 1999 and 2016 with primary micrometastatic disease on SLN biopsy pathology for incident, invasive breast cancer. All patients received either lumpectomy or mastectomy, in addition to axillary evaluation, at a single NCI-designated Comprehensive Cancer Center by a fellowship-trained surgical oncologist. Patients with micrometastasis associated with recurrent disease were excluded. Patients receiving neoadjuvant therapy were also excluded from this study. The institutional NCCN database collected breast cancer data from 1997 through 2012; the national data collection was closed in 2012, but a prospective programmatic Breast Cancer database continues to collect clinical and pathologic breast cancer data. Because the primary endpoint of this study is the use of axillary dissection in current breast cancer management, the time frame for the database searches was carried through to include recent cases.

Medical records were reviewed, and multiple data points were collected in addition to data collected from the NCCN and Breast Program databases. Data points collected and analyzed were as follows: date of birth, date of diagnosis, age at diagnosis, method of diagnosis (axillary ultrasound, magnetic resonance imaging [MRI], or axillary fine needle aspiration [FNA]), hormone receptor status, type of breast surgery (lumpectomy versus mastectomy), date of surgery, histology on final pathology, result of intraoperative imprint cytology (if performed), number of SLN biopsied, timing of axillary lymph node dissection (if performed), grade of cancer and largest metastasis, use and location of radiation as treatment of initial breast cancer, location and date of local recurrence, location and date of distant recurrence, last follow-up date, and number of patients deceased. Follow-up time was calculated from the last surgery date to the last documented oncology note.

Types of radiation included none, whole breast, chest wall, or nodal. If second surgery was performed, all data points were collected when available. Last follow-up date was defined as the last documented appointment with an oncologist or surgical oncologist at either an outside institution or our facility. Demographics were summarized using descriptive statistics. Wilcoxon rank-sum test, Kruskal-Wallis test, and Fisher exact test were used for data analysis.

Results

A total of 91 patients were diagnosed with micrometastatic nodal disease in the selected time frame. The median age at diagnosis was 56 y (range 31-85). Fifty patients (54.9%) received an MRI preoperatively, 11 of which showed an abnormal axilla, prompting further workup. Of these 11, eight (72.7%) had an axillary ultrasound. Thirty-four of the 91 patients (37.4%) had a diagnostic axillary ultrasound, and 13 patients had an FNA of the axilla performed for diagnosis. Of these, one patient was diagnosed with a positive axilla on percutaneous biopsy. The sensitivity of axillary ultrasound and FNA for micrometastatic axillary disease was 41.2% and 15.4%, respectively. The sensitivity of MRI for micrometastatic axillary nodes was 22%. Seven cases (7.7%) were estrogen receptor/progesterone receptor (ER/PR) negative, and five of these cases (5.5%) were ER/PR, Her2Neu negative. Thirty-eight patients received a lumpectomy, and 53 patients received a total mastectomy. The histological results on final pathology showed that the majority of patients had invasive ductal cancer (76/91, 83.5%) followed by invasive lobular cancer (11/91, 12.1%). The remaining four patients were classified as “other.” The majority of patients (50/91, 54.9%) were diagnosed with grade 2 cancer, followed by grade 3 (29/91, 31.9%), and grade 1 (11/91, 12.1%). The largest size of metastasis found during SLNB was documented, ranging from 0.2-2.0 mm (mean 0.99 mm) (Table 1).

Table 1 –

Clinical and pathologic features versus overall recurrence.

Variable	Level	n (%)			P value
		Total	NED	Any recurrence
ER status	ER+	61 (87.1%)	58 (87.9%)	3 (75%)	0.4308
	ER−	9 (12.9%)	8 (12.1%)	1 (25%)
	Total	70	66 (94.3%)	4 (5.7%)
PR status	PR+	79 (86.8%)	73 (86.9%)	6 (85.7%)	1.0000
	PR−	12 (13.2%)	11 (13.1%)	1 (14.3%)
	Total	91	84 (92.3%)	7 (7.7%)
Her2neu (FISH)	Equivocal	2 (2.2%)	2 (2.4%)	0 (0%)	0.6075
	Negative	80 (87.9%)	74 (88.1%)	6 (85.7%)
	Positive	9 (9.9%)	8 (9.5%)	1 (14.3%)
	Total	91	84 (92.3%)	7 (7.7%)
Grade	1	11 (12.2%)	9 (10.8%)	2 (28.6%)	0.3147
	2	50 (55.6%)	47 (56.6%)	3 (42.9%)
	3	29 (32.2%)	27 (32.5%)	2 (28.6%)
	Total	90	83 (92.2%)	7 (7.8%)
Histology	Ductal	76 (83.5%)	69 (82.1%)	7 (100%)	0.7029
	Lobular	11 (12.1%)	11 (13.1%)	0 (0%)
	Other	4 (4.4%)	4 (4.8%)	0 (0%)
	Total	91	84 (92.3%)	7 (7.7%)
First surgery type	Lumpectomy	38 (41.8%)	33 (39.3%)	5 (71.4%)	0.1241
	Mastectomy	53 (58.2%)	51 (60.7%)	2 (28.6%)
	Total	91	84 (92.3%)	7 (7.7%)
Second surgery type	Lumpectomy	10 (62.5%)	8 (61.5%)	2 (66.7%)	1.0000
	Mastectomy	6 (37.5%)	5 (38.5%)	1 (33.3%)
	Total	16	13 (81.3%)	3 (18.8%)
Breast radiation	No	47 (51.6%)	42 (50%)	5 (71.4%)	0.4362
	Yes	44 (48.4%)	42 (50%)	2 (28.6%)
	Total	91	84 (92.3%)	7 (7.7%)
Nodal radiation	No	71 (78%)	65 (77.4%)	6 (85.7%)	1.0000
	Yes	20 (22%)	19 (22.6%)	1 (14.3%)
	Total	91	84 (92.3%)	7 (7.7%)

NED = no evidence of disease; FISH = fluorescence in situ hybridization.

Eighty-one of 91 patients (89.0%) received in intraoperative imprint cytology evaluation of the sentinel nodes. The sensitivity of intraoperative cytology for micrometastatic disease was 4.9%. The median number of resected SLN was 2 (range: 1-8). On final pathology, 86/91 (94.5%) patients had one positive node, and 5/91 (5.49%) patients had two positive nodes for micrometastasis. Forty-two (46.1%) patients had ALND, of which 37/42 (88.1%) were performed as a second operation. Five patients had positive or atypical intraoperative cytology followed by immediate axillary dissection. Eighteen of 91 patients declined a recommended ALND. One of the 18 (5.6%) patients without axillary dissection or radiation developed a recurrence in the axilla at 88-mo follow-up. Additional positive nodes were found in 3/42 (7.1%) patients at the time of their ALND; the median number of additional positive nodes was 1.7 (range: 1-3).

Forty-four of the 91 (48.4%) patients received postoperative adjuvant radiation; 28/44 (63.6%) patients received whole breast radiation, 16/44 (36.4%) received chest wall radiation, and 20/44 (45.5%) also had directed nodal radiation in addition to either chest wall or whole breast radiation. Seven of the 91 (7.7%) patients had a documented recurrence. Four of these seven patients had local recurrences at the primary site (breast), and one had a recurrence in the axilla as previously mentioned. Two patients had distant recurrences; one recurred in the thoracic spine, and the other in liver and bone. Both of these patients had a delayed ALND. Overall, there was no difference in rate of recurrence between the 2 groups; recurrence was seen in 9.5% of patients receiving ALND and 6.1% in patients not receiving ALND (P = 0.70). The median time of follow-up was 47 mo (range: 0-203 mo); seven patients are deceased, two with a breast cancer–related death and one due to AML (Table 2). Cause of death in the other four patients is unknown.

Table 2 –

Incidence of local recurrence.

Variable	SLN only; no radiation	SLN + ALND; no radiation	SLN only and radiation	SLN + ALND and radiation	Total#
# Patients (%)	24 (26.4)	23 (25.3)	24 (26.4)	20 (22)	91
#Any recurrence (%)	3 (42.9)	2 (28.6)	0 (0)	2 (28.6)	7
#Local recurrence (%)	3 (60)	1 (20)	0 (0)	1 (20)	5

Discussion

Of recurrences after ALND, two were in the breast and two were distant metastases. There was one axillary recurrence in a patient who declined axillary dissection as well as chest radiation. This recurrence presumably represents existing axillary disease that was present but not surgically resected at the time of the initial breast and axillary operation. This seems to be a rare event, as most residual disease left in the axilla does not manifest as axillary recurrence.²⁰ Furthermore, this series demonstrates a low rate of additional nodal involvement on ALND. Although our population size is small and analysis preliminary, the results support other publications demonstrating low recurrence risk in micrometastatic patients.^{3,4,11,13,14,18,21} The fact that there was no difference in recurrence between the ALND group compared to the SLN only group supports the omission of ALND in clinically node-negative women with pathologically micrometastatic disease on sentinel node biopsy. The Z-0011 study found no significant difference in local or regional recurrence at a mean of 5.9 y follow-up among patients with 1 or 2 positive SLNs who received and did not receive ALND.^14,16 Even patients who opted out of a recommended ALND were shown to have rates of recurrence as low as (0%-3.7%).³ In contrast, the MIRROR study, a large retrospective Dutch cohort study, did demonstrate an apparent benefit of doing ALND due to an increased hazard risk of axillary recurrence for patients not receiving ALND compared to those that did, even after controlling for systemic therapy; these data was presented at the ASCO Annual Meeting in 2009 but to-date has not yet been published in manuscript format.²²

In our series, the decision to proceed to immediate ALND during surgery was based on an intraoperative SLN test. Intraoperative cytology for SLN metastasis has been shown in some series to have high false positive rates (38.6%) and extremely low sensitivity for micrometastases.^1,3,17 Although not a primary focus of this study, our analysis also demonstrates a low sensitivity of cytology for micrometastatic disease (4.9%) and suggests that the decision to perform this test is low yield in clinically node-negative patients. Intraoperative testing results in extended surgical and anesthesia times for patients; this additional time and expense are eliminated by deferring sentinel node biopsy evaluation to permanent pathology without any intraoperative evaluation in the clinically node-negative patient, particularly given the results of recent studies supporting the use of SLN only in low-volume nodal disease.^16,21

The question of quality of life is also raised when considering the necessity for ALND. Numerous studies have reported a wide variety of side effects due to ALND that can truly compromise a patient’s quality of life post recovery. Overall, postoperative complication rates are three-fold higher in groups receiving ALND compared to those that do not.²³ Trials of SLN alone versus ALND have reported clinically significant differences in patient reported cases of lymphedema; 2%-8% versus 13%-19%, although patient reporting is subject to bias and challenging to standardize.^3,24,25 Other side effects, such as parasthesias in the upper arm and axilla, arm weakness, and axillary seromas have also been reported as more prevalent in patients receiving ALND than those receiving SLN alone.^13,26 Although SLNB alone results in less cases of morbidities, it is still noted that there remains a clinically relevant risk of lymphedema, though the risk does not correlate with the number of lymph nodes removed.^27,28

Conclusions

Our study has shown a lower rate of recurrence in patients not receiving ALND than the rate of lymphedema previously described in the literature. Several of those studies have recommended omitting ALND when possible. Considering our small and highly selected single-institution database, the decision to abandon ALND in patients with a micrometastatic SLNB needs further validation, although the evidence presented in this study serve to reinforce the highly suggestive data in the current literature, and supports deferring intraoperative sentinel node testing in a clinically node-negative population.