Management of the Axilla after Neoadjuvant Systemic Therapy

Introduction

Neoadjuvant systemic therapy (NAST) for the treatment of breast cancer has been shown to have multiple benefits including increasing the breast conservation rate, allowing the assessment of responsiveness to systemic therapy, and providing prognostic information based on pathologic response [1-9]. The perceived benefits of NAST have led to its increased use in the USA. According to Murphy et al., examination of the National Cancer Database from 2010 to 2015 demonstrated increased use of NAST from 15.7 to 26% [10]. This increase in use was most prominent in high pathologic responder subtypes such as ER and PR−/Her2neu− and Her2neu + cohorts [10].

The increased use of NAST poses important questions for the optimal management of the axilla. Lymph node metastasis remains an important prognostic marker for patients with breast cancer [11-14]. In some cases, extent of disease in the lymph nodes can help guide selection of the type of cytotoxic chemotherapy to recommend and inform decision-making for adjuvant treatments, including additional systemic therapy and radiation fields [15••]. As such, proper staging and management of the axilla is critical. The increasing use of NAST creates challenges for the multidisciplinary treatment team in determining the optimal timing of axillary surgery as well as what procedure to perform. The objective of this article is to review the current literature on axillary management in patients with breast cancer who are treated with NAST.

Clinical staging of the axilla prior to treatment planning

Thorough clinical staging of the axilla of patients newly diagnosed with breast cancer is the first step in the treatment pathway. For patients who are being considered for NAST, the National Comprehensive Cancer Network (NCCN) advises imaging evaluation of the axillary basin by either ultrasound or MRI in addition to clinical exam, followed by biopsy of any radiographically or clinically suspicious lymph nodes [15••]. The preferred biopsy technique is core needle [16] as it is less invasive than excision biopsy and provides the necessary information for staging and treatment decision-making. Core needle biopsy is also safe with a recent systematic review describing a complication rate of < 2% (hematomas, bleeding, infection, vasovagal reaction) [17]. Fine needle aspiration is also an acceptable technique with similar specificity to core needle biopsy (98% versus 96%). However, the overall sensitivity is lower with FNA compared with core needle biopsy (74% versus 87%), supporting the use of a core needle when perceived to be a safe option [18]. The results of this pretreatment clinical work-up are critical for defining subsequent axillary surgery.

Clinically node negative

Patients who are clinically node negative at the time of presentation should undergo a sentinel lymph node (SLN) biopsy as the first step in their axillary staging. SLN biopsy has been shown in a randomized controlled trial to be accurate, with an identification rate of > 97% and a false negative rate (FNR) of 9.8% [19]. Although the use of SLN in this clinical context is widely agreed upon, some controversy exists regarding whether to perform the SLN biopsy prior to versus after NAST. Advocates for SLN biopsy prior to NAST argue that this timing provides the most accurate assessment of the axillary nodes at the time of presentation, which has implications on prognosis, selection of chemotherapy agents, and planning of radiation therapy fields [20]. Further, advocates cite higher accuracy rates of pre-NAST lymphoscintigraphy when compared with post-NAST [21]. The higher accuracy of pre-NAST SLN can be partially explained by alterations of the lymphatic drainage pathways due to treatment-induced fibrosis in the lymphatic channels [21]. However, SLN biopsy prior to NAST requires an additional surgical procedure and can delay initiation of systemic therapy.

Most current clinical practice guidelines recommend performing SLN biopsy after NAST in patients who are clinically node negative [15••, 22]. Although the accuracy of post-NAST SLN is slightly lower compared with pre-NAST SLN, post-NAST SLN biopsy has an acceptable FNR, especially when dual dye technique is utilized (Table 1). Two meta-analyses have summarized the available literature regarding post-NAST SLN biopsy. A 2016 meta-analysis of 16 studies included 1456 patients who had operable breast cancer and underwent a SLN biopsy after neoadjuvant chemotherapy followed by an axillary lymph node dissection (ALND) [29•]. Identification rates were high (97% with dual dye, 96% using only blue dye or radiocolloid). The pooled false negative rate was 11% if hematoxylin and eosin (H&E) staining alone was employed and 4% if immunohistochemistry (IHC) was included. These findings are slightly better from that of an earlier meta-analysis by Xing et al. which reported a pooled identification rate of 91% with a false negative rate of 12% [26]. The improvements observed may reflect surgeons increased comfort and familiarity with the SLN procedure in the setting of NAST.

Table 1.

Select studies assessing the use of SLN biopsy before versus after NAST for clinically node-negative breast cancer

	SLN biopsy before NAST			SLN biopsy after NAST
Author, year	Grube, 2008 [23]	Schrenk, 2008 [24]	Papa, 2008 [25]	Xing, 2006 [26]	Hunt, 2009 [20]	Classe, 2009 [27]	Pecha, 2011 [28]	Geng, 2016 [29•]
# of patients	55	45	58	1273	575	130	172	1456
Identification rate	100%	100%	98.8%	91%	97.4%	94.6%	89.5%	96%
Overall FNR	0%π	0%	0%	12%	5.9%†	9.4%	16.3%	6%
Method of SLN mapping	100%
radionuclide;	71.1% dual	100% dual	NA*	82.1% dual	100% dual	100%
select cases also used blue dye	dye	dye		dye	dye
radionuclide; select cases also used blue dye	NA*
Study

methodologySingle institution, RetrospectiveSingle institution, ProspectiveSingle institution, ProspectiveMeta-analysisSingle institution, Retrospective Multicenter, ProspectiveSingle institution, RetrospectiveMeta-analysis

^πNo ALND performed; median follow-up of 16.3 months for clinical FNR of 0%

^†Only 84 patients underwent ALND and were used to calculate FNR

^*Varied methods of SLN mapping used in the different studies included in the meta-analysis

In addition to the comparability of identification and false negative rates, a study by Hunt et al. examined outcomes in SLN biopsy following NAST compared with those undergoing SLN biopsy prior to chemotherapy. Enrolled patients had clinically node-negative T1–T3 breast cancer. This study reported no difference in outcomes such as loco-regional recurrence and disease-free or overall survival between those who underwent SLN before or after NAST [20].

There are pragmatic reasons why performing a SLN biopsy after NAST may be preferred. Post-NAST SLN biopsy spares patients an extra surgery which may delay initiation of NAST. Additionally, the timing of SLN around NAST may impact what axillary surgery patients ultimately receive. For clinically node-negative patients found to have a positive SLN prior to NAST, an ALND would be recommended as the standard of care. This recommendation does not change, even if patients go on to have a robust response to the NAST. Patients who undergo a SLN biopsy in the post-NAST setting and are found to have persistent metastatic disease in the SLN would also be recommended an ALND as standard of care as patients with persistent metastatic disease in the SLN after NAST have a high chance of having disease in additional nodes (as high as 50% in one study) [30]. Although a number of RCTs have examined the utility of deescalating the extent of axillary surgery for patients with SLN biopsypositive disease, none of these studies included patients undergoing NAST [31-33].

However, for some patients, NAST may convert a clinically negative but pathologically positive axilla to a pathologically negative one. If patients are found to have a negative SLN post-NAST, these patients would not be recommended additional axillary surgery and would be potentially spared the morbidity of an ALND (paresthesias, pain, reduced shoulder range of motion, wound infections, seromas, and lymphedema [34-36]). Performing the SLN biopsy post-NAST may decrease the extent of axillary surgery required, especially for those patients with a robust response to NAST.

Clinically node positive

Traditionally, all patients who presented with clinically positive lymph nodes were recommended a full ALND. An ALND clears the nodal basin of any lymph node disease, minimizing the risk of local recurrence and providing full pathologic nodal staging. SLN biopsy, intended to be a diagnostic test to identify metastatic disease, had no role for patients who already had axillary disease confirmed. ALND remains an acceptable, and sometimes preferred, option for many patients with axillary metastases identified at the time of initial presentation.

However, for patients who receive NAST, there is an increasing body of evidence supporting more conservative axillary surgery for patients who have a good clinical response to treatment. With modern-era systemic therapy, a substantial proportion of patients will have a pCR to NAST. According to data from the ACOSOGZ1071 trial, approximately 65% of patients with Her2neu + disease, 50% of patients with ER and PR−/Her2neu− disease, and 21% of patients with ER and/or PR+/Her2neu− disease had a pCR in the axilla [37]. Patients who have had a pCR in the nodal basin derive no oncologic benefit from ALND. However, these patients are at risk for substantial morbidity associated with the procedure, including paresthesias, pain, reduced shoulder range of motion, wound infections, seromas, and lymphedema [34-36]. Performing SLN following NAST is one proposed way of tailoring the extent of axillary surgery and attempting to differentiate those patients with residual axillary disease who would benefit from a full ALND from those patients with a nodal pCR.

Although deescalating the extent of axillary surgery for patients who have had a good response to NAST is a worthy objective, there are significant challenges associated with this approach. Theoretical concerns exist about the accuracy of the SLN biopsy following NAST for patients known to have nodal metastases at the time of diagnosis. Robust responses to chemotherapy may lead to fibrosis and scarring of lymphatic channels, both of which may alter lymphatic flow following chemotherapy. In this scenario, lymphatic mapping maybe less successful (decreased SLN identification rate). Further, the lymphatic channels that drain the area of the breast with the cancer prior to NAST may drain via different pathways following NAST. In this scenario, a SLN may be identified through the lymphatic mapping, but it may not be the same SLN that the breast drained to prior to NAST. Alternate drainage such as this could result in a false negative result. A false negative with regard to pathologic axillary staging has high clinical relevance in this scenario where 100% of patients were known to have cancer in the lymph nodes at diagnosis [20, 26, 38••].

Several clinical trials have examined the use of SLN biopsy for patients with positive axillary lymph nodes treated with NAST (Table 2). The ACOSOG Z1071 study was a phase II study specifically designed to test the accuracy of SLN biopsy following NAST, reporting both identification rate and false negative rate. This trial enrolled 756 patients with T0–4, N1–2, M0 breast cancer with biopsy-proven lymph node metastasis who received NAST. All patients underwent a SLN biopsy after NAST, followed by an immediate ALND. For this trial, a false negative rate of 10% was deemed a priori to be the threshold where performing a SLN would be clinically acceptable [19]. The overall false negative rate for patients in ACOSOG Z1071 was 12.6% if at least 2 SLNs were identified. As this was higher than the pre-determined threshold of 10%, the authors concluded that “one cannot reliably detect all axillary lymph node metastases in cN1 breast cancer patients following chemotherapy by SLN procedures” [39••].

Table 2.

Trials assessing the use of SLN biopsy following NAST for clinically node-positive breast cancer

Study	ACOSOG Z1071 [39••]	SENTINA [38••]	SN FNAC [40••]
# of patients	637	592	135
Identification rate	92.7%	80.1%	87.6%
Overall FNR	12.6%†	14.2%	8.4%*
≥ 3 SLNs removedδ	9.1%	9.6%	4.9%*
Dual dye used	10.8%	8.6%	5.2%*
≥ 2 SLNs removed and removal of clipped node	6.8%	NA	NA
Study methodology	Multicenter, prospective	Multicenter, prospective	Multicenter, prospective

^†With ≥ 2 SLN removed

^*Considered IHC disease and ITC’s to be positive

^δ≥ 2 SLNs removed for SN FNAC

The SENTINA trial (Table 2) was another prospective study assessing optimal timing of SLN biopsy in the setting of NAST. This study enrolled 1737 patients into one of four arms at 103 institutions in Germany and Austria. Arm C (N = 592) of this trial was comparable with the ACOSOG Z1071 study design, consisting of patients who were clinically node positive and converted to clinically node negative following neoadjuvant chemotherapy. This arm underwent a SLN biopsy followed by an immediate ALND. In Arm C of SENTINA, a SLN was identified 80.1% of the time. The authors also reported a false negative rate of 14.2% [38••].

Finally, the SN FNAC trial (Table 2) analyzed the use of SLN biopsy followed by immediate ALND after neoadjuvant chemotherapy for patients with biopsy-proven lymph node-positive breast cancer. This study enrolled 153 patients. The study design was unique in that it mandated the use of IHC after H&E staining to confirm SLN negativity. Further, the study classified isolated tumor cells (ypN0[i +1], ≤ 0.2 mm) in the SLN as node positive. With these criteria, the SLN identification rate in SN FNAC trial was 87.6% with a false negative rate of 8.4% [40••].

All three of these clinical trials assessing the role of SLN biopsy after NAST for patients with clinical N1 disease reported lower identification rates and higher false negative rates than has historically been observed for the SLN biopsy in the setting of clinically negative lymph nodes [19]. However, secondary analyses of these studies have been able to identify subsets of patients for whom the SLN was more accurate. Use of a dual dye technique with both vital blue dye and radiolabeled colloid resulted in improved identification rates; this was reported in SENTINA to be 87.8% versus 77.4% with single dye techniques. Dual dye also improved the false negative rate in all three trials, dropping to 10.8% for ACOSOG Z1071, 8.6% in SENTINA, and 5.2% in SN FNAC. One additional factor observed to improve the false negative rate was sampling of a greater number of lymph nodes [38••, 39••, 40••]. As the collection of more SLNs (at least 2 or 3) and a dual dye technique improved the false negative rates to below the clinically acceptable threshold of < 10% set in ACOSOG Z1071 and SN FNAC [39••, 40••], many oncologists argue that SLN biopsy can be considered a safe alternative for women with cN1 who undergo NAST. However, other surgeons argue that a FNR near 10% is still too high for a cohort of women known to have metastatic disease in their axillary nodes and would present too great of a risk of leaving residual disease in the axilla.

One additional concept that has emerged in an attempt to further decrease the false negative rate of SLN after NAST for patients with known axillary metastases is removal of the “clipped node.” In ACOSOG Z1071, a subset of patients (203 of the 637 patients) had a clip placed in their axillary lymph node when the core biopsy was performed. The majority of patients (~ 75%) had this clipped node removed as one of the SLN’s. For these patients for whom the clipped node was removed as a SLN, the false negative rate of SLN was 6.8% [41•]. The improved SLN accuracy associated with removing the clipped node makes conceptual sense. This lymph node was known to have metastatic disease prior to NAST. Removal of this lymph node as a SLN provides some indication that the lymphatic channels drain via similar pathways post-NAST as they did pre-NAST. Further, removing this lymph node which biopsy confirmed to have metastatic disease at presentation and determining that disease has been fully eradicated by NAST also increases surgeon and patient comfort that there is no residual disease hiding elsewhere in the axillary basin. The NCCN guidelines currently encourage marking of biopsied lymph nodes (clips, tattoo, etc.) and documenting their removal as a way to improve the false negative rate with SLN [15••]. This technique combined with use of dual tracer and removal of at least 2 lymph nodes makes SLN a reasonable option for deescalating the extent of axillary surgery for women with axillary metastases but a strong response to NAST.

However, 24.1% of patients in the ACOSOG Z1071 trial who had their axillary nodes clipped did not have the clipped node removed as one of the SLN’s. For these cases where the clipped node was instead identified in the ALND specimen, the false negative rate of SLN biopsy was 19% [41•]. Recognizing the importance of removing and pathologically evaluating the clipped node, alternate methods of localization have been successfully described. Caudle et al. described a procedure called targeted axillary dissection (TAD), which involves use of iodine-125 seeds to preoperatively localize the clipped node in combination with performing a SLN biopsy with dual mapping. In their experience, removal of the SLN alone resulted in a false negative rate of 10.1%, comparable with that observed in the ACOSOG Z1071 trial. However, combining removal of the preoperatively localized clipped node in addition to the SLN (the TAD procedure) decreased the false negative rate to 1.4% [42••].

The key component of the TAD procedure is removal of the clipped lymph node along with the SLN. Iodine-125 seeds are one option, familiar to surgeons who already use the seeds for breast location. However, other techniques, including tattooing of biopsied lymph nodes at the time of biopsy [43] and wire location of the clipped node on the day of surgery, [44] have also been described. Other surgeons have opted to not localize preoperatively, as the clipped node is removed as part of the SLN specimen 75% of the time. If the clipped node is not removed, these surgeons perform selective sampling of lymph nodes in the axilla, guided by preoperative cross-sectional imaging to find and remove the clipped node. Regardless of the technique used, the critical concept is that the clipped node should be removed and evaluated. Combined, these experiences suggest that there is a role of SLN following NAST for patients with known axillary disease, if both SLN mapping with dual technique and removal of the clipped node are performed.

The benefits to avoiding unnecessary ALND for women with a strong response to NAST are clear. However, it is worth considering the consequence of false negative findings if a more conservative axillary surgery is employed. Current NCCN guidelines urge clinicians to “strongly consider radiation to the chest wall, infraclavicular region, supraclavicular area, internal mammary nodes, and any part of the axillary bed at risk” in clinical N1, ypN0 disease as well as for “any positive axillary node after chemotherapy” [15••]. Based on current guidelines, all women with a positive lymph node at the time of presentation are strongly considered to have comprehensive nodal radiation. Given this, the consequences of leaving microscopic axillary disease as a result of a false negative axillary surgery are likely low.

It is important to note that several ongoing studies are assessing the utility of radiation in patients with node-positive disease who undergo neoadjuvant chemotherapy, specifically asking whether the extent of radiation could be lessened for those patients with a good clinical response to NAST. The Alliance for Clinical Trials in Oncology A011202 trial is designed to assess the effectiveness of axillary radiation compared with ALND followed by axillary radiation for patients with node-positive breast cancer treated with neoadjuvant chemotherapy [NCT01901094]. The NSABP B-51 trial is also currently enrolling to assess the effectiveness of additional chest wall and regional nodal radiation therapy following mastectomy or breast conserving surgery and SLN biopsy ± ALND in patients with positive axillary nodes prior to neoadjuvant chemotherapy who convert to pathologically negative axillary nodes [NCT01872975]. These studies may yield additional information on the optimal management of patients with node-positive breast cancer at presentation.

Summary

With the increasing use of NAST and improving rates of pCR for patients diagnosed with breast cancer [10], optimal management of the axilla in this population is becoming an increasing challenging but important. For patients presenting with clinically negative axillary nodes, a consensus among professional organizations exists that performing a SLN biopsy following NAST is the preferred approach [15••, 22]. If a positive SLN is identified in this setting, an ALND remains the standard of care.

For patients with clinically node-positive disease, ALND remains an acceptable option. However, post-NAST SLN biopsy is an option for patients with robust responses to NAST. Landmark studies including ACOSOG 1071, SENTINA, and SN FNAC have demonstrated acceptable identification and false negative rates when a dual dye technique for SLN mapping is used as well as ensuring the removal of at least 2 lymph nodes [38••, 39••, 40••]. Additionally, utilizing IHC [40••] and removal of clipped nodes is recommended to further improve accuracy [41•]. Ongoing studies assessing the effectiveness of axillary radiation for node-positive breast cancers are currently underway and may provide for alternative treatments to axillary lymph node dissection (Alliance A011202 [NCT01901094] and NSABP B-51 [NCT01872975]).

Opinion statement.

As the use of neoadjuvant systemic therapy (NAST) increases, the optimal management of the axilla has become increasingly complex. Consensus among professional organizations is that those patients with clinically negative axillary nodes who are being considered for NAST should undergo a sentinel lymph node (SLN) biopsy following NAST. If a positive SLN is subsequently identified, an axillary lymph node dissection (ALND) is the current standard of care. For patients with clinically node-positive disease, SLN biopsy is a reasonable option for those with a good response to NAST. Patients should undergo SLN mapping with a dual dye technique. Additionally, at least 2 lymph nodes should be removed, including the previously biopsied and marked lymph node with cancer. In this setting, the identification and false negative rates are acceptable. Patients found to have a negative SLN at this time may be spared the morbidity associated with ALND. Patients found to have persistently positive lymph nodes following NAST, either clinically or pathologically, should undergo a complete ALND.