Historic randomized study data have failed to demonstrate a survival benefit for axillary lymph node dissection for patients with clinically positive nodes8–10, yet ALND has remained the standard of care for these patients. Recent practice-changing clinical trials assessing de-escalation of axillary surgery have excluded patients with clinically positive nodes1,2, resulting in a significant dearth of contemporary data to guide surgical management of the axilla in this patient cohort. More recently, nodal downstaging with NAC has become an acceptable strategy to avoid ALND for patients with clinically positive nodes who respond clinically to treatment11–13, but nodal response rates vary considerably by tumor subtype with the highest rates seen in patients with Her2-overexpressing (Her2+) or triple negative breast cancer (TNBC)14–16. For patients with ER-positive disease, nodal downstaging rates rarely exceed 20%, and since residual disease is not prognostic of outcome for this cancer subtype15,17,18, patients are typically recommended for surgery first, and often ALND.
The concern for de-escalation of axillary surgery for patients with clinically positive nodes appears to stem from data that demonstrate that these patients have less favorable tumor characteristics than patients with SLN+ disease3–6. Our data of primarily ER-positive patients support these findings and demonstrate that progression of tumor characteristics from lower to higher risk mirrors the progression of clinical nodal disease from SLN+ to cNUS to cNpalp, with statistically significant differences existing at both ends of the nodal spectrum. Patients with clinically palpable nodes have lower PR-positivity, higher tumor grade, higher Ki67 levels, and higher rates of LVI than patients with SLN+ disease, but these higher risk features do not predict for higher nodal stage. When we limit our study group to ER+ patients only, we note fewer clinicopathologic group differences, which may reflect that the differences in risk features are driven by the ER- breast cancers and that there is more homogeneity among ER+ breast cancers despite nodal presentation. However, we are cautious to interpret this lack of significant difference as evidence for the absence of a difference, especially when the effect fell out of significance after a reduction in sample size. Regardless, in both the overall cohort and the ER+ subgroup, only tumor size and lobular histology appear associated with higher pathologic nodal burden among patients with node positive disease.
Previous studies support our findings that size of the primary tumor and non-ductal histology are the pertinent factors associated with higher pathologic nodal stage in patients with both SLN+ and cNUS breast cancer5,19−22. Caudle et al5 compared over 700 patients who had either SLN+ or cNUS-detected disease but excluded patients with palpable adenopathy, and similarly identified that age, grade, and PR-positivity differed between the two groups, but these higher risk features did not predict for higher nodal stage. Only tumor size and lobular histology were related to extent of nodal disease, and this did not differ between SLN+ or cNUS patients. Unlike the present study, the authors demonstrated that the method of nodal presentation, i.e. having metastases identified by US (odds ratio [OR] 3.80, [95% CI 2.65 – 5.43]; p < .0001) was correlative of ≥ 3 positive nodes. However, it is unclear whether this variable would be predictive of pathologic nodal upstaging (> pN1 disease). In addition, only 77.9% of patients in the cNUS group were ER-positive, which was statistically different than the SLN+ group, and 14% were Her2+. In a study by Mamtani et al20, in which patients with SLN+ disease were evaluated, only tumor size and lobular histology were independently associated with additional positive nodes on ALND when adjusting for ER and Her2 status. Grade and presence of LVI did not appear to influence nodal positivity.
Other studies have demonstrated that certain high-risk features do predict for higher nodal stage, which led authors to conclude that patients with clinically positive nodes should not be considered for omission of ALND3–6. However, a disproportionately high number of patients in these studies had ER-negative (23.9%-29.7%) or Her2-positive (18.7%-31.2%) disease. Modern day treatment paradigms would likely include neoadjuvant chemotherapy for this patient cohort with node-positive, ER-negative or Her2-positive disease, which make these datasets less reliable for guiding surgical management recommendations for the primarily ER-positive breast cancer patients who proceed straight to surgery.
Few studies have evaluated nodal burden in patients with palpable adenopathy and ER-positive, Her2-negative breast cancer, in whom downstaging with NAC is less effective and alternatives to ALND are needed. Angarita et al.7 performed a subgroup analysis of this specific patient population and demonstrated that 44.6% of patients had pN1 disease, and tumor size and lobular histology were significant predictors of >pN1 disease. Crown et al22 similarly demonstrated that 57% of patients with palpable adenopathy had pathologically low nodal burden (pN1). More than 97% of patients in that study had ER-positive disease, and only tumor size and non-ductal histology were again associated with higher nodal stage. Higher risk features such as palpable adenopathy, LVI and high tumor grade did not predict for higher nodal stage.
Certainly, it is understood that higher risk features carry worsened prognoses, but this finding should not obligate patients to extensive lymphatic axillary surgery, particularly since randomized trial data have not demonstrated a survival benefit to ALND8–10. With the risk of disabling lymphedema and functional morbidity accompanying this surgical procedure, the oncologic benefit to ALND needs to be continually assessed for all patients. Verheuvel et al.4 demonstrated a worsened overall (hazard ratio [HR] 2.67, [95% CI 1.48 – 4.84]) and disease-free survival (HR 2.71, [95% CI 1.49 – 4.92]) in patients with cNUS positive disease compared to SLN+ patients (despite both groups having been treated with ALND). However, the cNUS group had a significantly higher proportion of TNBC patients compared with the SLN+ group in addition to having higher risk features. Recognizing that TNBC patients have higher rates of distant relapse, the disparity in histologic subtypes between the two groups likely contributed to the survival differences noted. As recommendations for adjuvant systemic therapy are being tailored to tumor biology, many patients with higher risk features will be recommended for adjuvant chemotherapy regardless of nodal status.
The first interim analysis of the RxPONDER trial demonstrated that worsened tumor biology reflected in a high Oncotype Recurrence Score predicts the need for adjuvant chemotherapy in post-menopausal women with ER-positive disease, not pN1 nodal status23. The Early Breast Cancer Trialist Cooperative Group (EBCTCG) meta-analysis also highlighted the benefit of radiotherapy on locoregional recurrence and survival in patients with 1-3 positive nodes24, expanding the indication for nodal irradiation in patients with minimal nodal disease. As more patients with low nodal burden are treated with nodal irradiation and the decision for systemic chemotherapy is made based more on tumor biology than nodal status, the benefit of surgical clearance of the axilla must be re-evaluated. Additionally, for patients with ER-positive disease, options for adjuvant systemic therapies continue to expand with studies evaluating the benefit of CDK 4/6 inhibitor therapy and fulvestrant in the adjuvant setting. Clearly, there is opportunity to consider de-escalation of axillary surgery in patients who present with clinically positive nodes for whom effective adjuvant therapies exist.
Randomized trials assessing the safety of omission of ALND in patients with clinically positive nodes who are not effectively downstaged with NAC are needed. The TAXIS study, a multicenter, randomized, phase III trial is enrolling patients with clinically positive nodes to tailored axillary surgery (TAS) followed by ALND and regional nodal irradiation (RNI) or TAS followed by RNI25. The TAS procedure, as described by the investigators, involves a standard SLNB with selective removal of all palpable disease and documented removal of the biopsy-proven, clipped node. Data from our study and others confirm that a large minority of patients with clinically positive nodes have minimal nodal disease and therefore support this trial design.
We acknowledge the limitations of our study, including its retrospective design and small sample size. Many patients with clinically positive lymph nodes during this study period were treated with NAC at our institution and were excluded, thus limiting the sample size of the cN+ patient cohort. Many of the SLN+ patients in our database who underwent ALND were considered ineligible for Z0011 based on type of surgery (76.6% underwent mastectomy) rather than number of nodes involved or extranodal extension, which may result in more favorable features attributed to the SLN+ group. Furthermore, while some studies have found significant associations between presence of LVI and nodal burden26, we did not see this same effect, which may be a result of using the core needle biopsy rather than the surgical specimen to assess LVI. We purposefully chose this approach to determine whether any features identified preoperatively could be used to predict nodal disease and therefore affect nodal surgery. The absence of a direct association in our dataset could be due to the inherent risk for a false-negative result on a core biopsy and/or the small sample size. However, others have also demonstrated no correlation with LVI and nodal burden19,22.