This study addresses the question of whether a complete pathological response observed in the primary tumour after NACT can predict a similar response in the axillary lymph nodes and therefore allow for less invasive surgical approach for axillary staging.
The rate of pCR of the breast in our study was 21%, which is in keeping with the 22% from the meta-analysis by Cortazar and Geyer [8] and the 24% of the Dutch multicentre trial [19].
Patients who achieved pCR in both the breast and axilla comprised 16.9% of all patient population, which is similar to results from Schipper et al.’s 17.5% [21] and Fayanju et al.’s 19% [22] studies that included 20,000 patients.
In our study, pCR of the primary tumour was associated with change in the stage of the axillary lymph nodes after NACT. Patients who had presented initially with positive axillary lymph nodes and showed a pCR in the primary tumour mostly showed similar response pattern in the axilla after clearance.
Morgan et al. [23] showed similar results in their study which included 83 patients with pathological axillary lymph nodes who received NACT. Twenty-one patients had pCR in the breast. Of those, 17 (80%) had also a complete response in the axilla which is the same percentage as our study.
Another study by Tadros et al. [24] included the same number of patients as our study yet had a higher rate of pCR in both the breast and axilla (89.6%). This difference can be attributed to the fact that the study only included HER2-positive and triple-negative subtypes. The multicentre study by Samiei et al. [19] reports a 45% of patients (245/544) achieving pCR in both breast and axilla.
The pCR in the primary corresponded to a similar response in the axillary lymph nodes in most of the study cohort.
There seem to be a variation among different groups of patients treated with NACT in terms of the expected response in the axilla following pCR of the primary.
The presence of pCR in both the breast and axilla depended largely on the molecular subtype of the tumour. The commonest subtype was HER2 enriched (91.7%) followed by Luminal B (83.3%), triple-negative breast cancer (78.2%) and finally Luminal A (66.7%). These results are in concordance to large database study by Barron et al. [18] that included 30,821 patients who received NACT. Among the 3128 patients who achieved pCR in both the breast and axilla, the most common molecular subtypes were HER2 enriched (88.7%), Luminal B (86.7%), TNBC (85.9%) and Luminal A (69.5%).
A similar good response was observed in grade 3 tumours as well, which was also witnessed in the multicentre study by Samiei et al. (p = 0.045) [19].
The response was inversely proportional to the degree of HR positivity, with HR-positive tumours less likely to show pCR in the axillary lymph nodes following complete response in the primary. This same observation which was described in the meta-analysis by Houssami et al. [25] and the pooled analysis by Cortazar et al. [26].
Patients with lobular cancer were less likely to show similar pattern.
There have been previous published literature aiming at predicting patients with axillary pathological complete response following NACT.
Schipper et al. [21] proposed a model that included patient age, clinical tumour stage, histological subtype, hormone receptor and HER2 status, as well as receiving trastuzumab and taxanes. Their model had a specificity of 43% and PPV of 65%.
Kim et al. [27] proposed a prognostic nomogram to predict who achieve axillary pCR following NACT. The nomogram items included patient age, initial clinical T stage, clinical nodal stage, tumour grade, ER, HER2, Ki67 expression, clinical tumour and nodal response and regimen of NACT.
The factors included in that nomogram seem to correlate to the factors that we have identified as influencing the pCR in both breast and axilla.
To the best of our knowledge, this is one of the few studies looking specifically at the group of patients who showed pCR in the primary tumour to establish correlation with the axillary response. This group is identified as a potential group where axillary staging could be less invasive. It is possible that this group of patients can be offered a less invasive axillary staging procedure such as sentinel node biopsy or target axillary dissection.
We recognize many limitations to our study. A main limitation is the fact that the axillary lymph node positivity was established pre-chemotherapy with ultrasound scan and without a biopsy. Although the signs of a malignant lymph node on ultrasound are highly reliable [28], most centres depend on a core biopsy to establish this diagnosis. At the time of studying this cohort at our centre, the radiological diagnosis was considered adequate to establish nodal involvement. It is now clear that most centres require core biopsy to confirm axillary nodes involvement.
Another main limitation is the lack of availability of target axillary dissection approach (TAD) [29]. In our centre, this approach has not been adopted yet. Selected patients with low disease burden in the axillary lymph nodes initially could have been offered this approach in an attempt to spare them axillary clearance. However, most of the patients included in this study had presented with heavy nodal involvement initially (at least 3 pathological lymph nodes) and therefore would not have been suitable for TAD according to the current guidelines [17] even if it were available.
Another limitation is the retrospective nature of the study. This is compensated by the high-quality data accurately obtained from pathological reports to minimize a recall bias. The small number of patients included in the study reflects the real-life percentage of patients with initial nodal metastases who develop pCR (13–23% [19, 22]) in the primary.
Based on the findings in our study, a less invasive axillary staging approach may be safely recommended in a selected group of node-positive patients after NACT when the primary tumour shows pCR. This group may include HER2-positive, HR-negative and grade 3 tumours. Less response is expected to be observed in the axillary lymph nodes in HR-positive and lobular carcinoma even with pCR in the primary tumour.
However, more body of evidence to support this hypothesis is needed. A target axillary dissection (TAD) approach may be considered in early small nodal disease [29] but should be discussed carefully within the MDT.
Further studies are needed to attempt at selecting groups of node-positive patients who can be safely offered less invasive axillary staging surgery after NACT. A selected group of patients among those who show pCR in the primary tumour are one of the possible groups.