Nomogram for Predicting Axillary Lymph Node Pathological Response in Node-Positive Breast Cancer Patients After Neoadjuvant Chemotherapy

0
0
0

Nomogram for Predicting Axillary Lymph Node Pathological Response in Node-Positive Breast Cancer Patients After Neoadjuvant Chemotherapy

Introduction
Neoadjuvant chemotherapy (NAC) has become a cornerstone in managing locally advanced or node-positive breast cancer, offering tumor downstaging and improving surgical outcomes. A critical prognostic factor in these patients is the pathological complete response (pCR) of axillary lymph nodes (ALNs), defined as the absence of residual invasive carcinoma in lymph nodes after treatment. Achieving ALN pCR allows for potential de-escalation of axillary surgery, reducing complications such as lymphedema. However, current methods for assessing axillary response, including sentinel lymph node biopsy (SLNB), carry limitations in accuracy. This study aimed to develop and validate a clinically practical nomogram to predict ALN pCR in biopsy-proven node-positive breast cancer patients after NAC, enabling personalized surgical decision-making.

Study Design and Patient Cohorts
The research involved retrospective and prospective cohorts from two Chinese centers. The retrospective cohort included 467 patients with cytologically confirmed ALN-positive breast cancer treated at the National Cancer Center/Cancer Hospital of the Chinese Academy of Medical Sciences (CHCAMS) between 2007–2014. Key inclusion criteria were: histologically confirmed invasive breast cancer, ALN metastases confirmed by fine-needle aspiration (FNA), completion of NAC, and subsequent axillary lymph node dissection (ALND). Exclusion criteria included distant metastases, incomplete clinical data, or negative ALN status pre-NAC.

For external validation, two prospective cohorts were analyzed:

  1. CHCAMS cohort (2016–2018): 167 patients.
  2. Beijing Tiantan Hospital (BTH) cohort (2018–2020): 114 patients.
    These cohorts followed similar inclusion criteria but excluded patients undergoing SLNB post-NAC to focus on ALND outcomes.

Treatment Protocols and Pathological Assessment
Patients received anthracycline- and taxane-based NAC regimens, with trastuzumab added for HER2-positive cases. Pertuzumab was introduced for HER2-positive patients after 2020. Clinical tumor response was evaluated using ultrasound and RECIST criteria, categorizing responses as complete (CR), partial (PR), stable (SD), or progressive disease (PD).

Pathological assessment of surgical specimens used the Miller-Payne grading system. ALN pCR was defined as no residual invasive carcinoma in excised lymph nodes. Hormone receptor (HR) status (ER/PR) and HER2 expression were determined via immunohistochemistry, with HER2 positivity defined per ASCO/CAP guidelines.

Statistical Analysis and Nomogram Development
Univariable and multivariable logistic regression identified predictors of ALN pCR. Variables analyzed included age, clinical T stage, tumor histology, histological grade, primary tumor response, ER/PR status, and HER2 expression. A stepwise selection process (backward elimination) refined the model, with odds ratios (ORs) and 95% confidence intervals (CIs) calculated.

The nomogram was internally validated using a 50/50 split of the retrospective cohort into training and test sets. Discrimination was assessed via area under the receiver operating characteristic curve (AUC). Calibration plots evaluated model accuracy. External validation used the prospective cohorts to test generalizability.

Key Findings

  1. Retrospective Cohort Analysis

    • ALN pCR Rate: 24.6% (115/467) achieved ALN pCR.
    • Predictors of ALN pCR:
      • Clinical T4 Stage: Reduced likelihood of pCR (OR: 0.321, 95% CI: 0.121–0.856; P=0.023).
      • Primary Tumor CR: Strongest predictor (OR: 0.189, 95% CI: 0.123–0.292; P<0.001).
      • ER Positivity: Lower pCR rates (OR: 0.530, 95% CI: 0.304–0.925; P=0.025).
        HER2 status and PR did not independently predict ALN pCR.

  2. Nomogram Performance

    • Internal Validation:
      • Training set AUC: 0.719 (95% CI: 0.638–0.771).
      • Test set AUC: 0.753 (95% CI: 0.704–0.791).
    • External Validation:
      • CHCAMS prospective cohort AUC: 0.720.
      • BTH cohort AUC: 0.720.
        Calibration plots demonstrated strong agreement between predicted and observed pCR probabilities.

  3. Prospective Cohort Insights

    • Higher pCR Rates: External cohorts showed improved ALN pCR rates (37.1% in CHCAMS, 40.3% in BTH) compared to the retrospective group (24.6%), likely reflecting advances in NAC regimens and targeted therapies.
    • HER2 Discrepancy: While HER2-positive status did not predict pCR in the retrospective analysis, its prevalence increased in external cohorts (40.1% vs. 28.7%), suggesting evolving diagnostic criteria and treatment efficacy over time.

Clinical Implications
The nomogram integrates three readily available clinical variables—clinical T stage, primary tumor response, and ER status—to estimate individual probabilities of ALN pCR. For example:

  • A patient with T2 stage, primary tumor CR, and ER-negative status would score highly, indicating a >50% likelihood of ALN pCR.
  • Conversely, T4 stage, partial tumor response, and ER positivity would predict low pCR probability (<20%).

This tool aids surgeons in deciding whether to pursue ALND or consider SLNB in patients likely to achieve axillary pCR. It also facilitates patient counseling by quantifying the benefits of NAC and personalized surgical planning.

Limitations and Future Directions

  • Retrospective Design: Potential selection bias and missing data (e.g., Ki-67 index, tumor-infiltrating lymphocytes) limited model refinement.
  • Temporal Heterogeneity: Evolving HER2 testing guidelines and NAC regimens (e.g., pertuzumab introduction) may affect reproducibility.
  • External Validation: While prospective cohorts validated the model, broader multicenter studies are needed to confirm generalizability across diverse populations.

Future research should incorporate molecular markers (e.g., genomic signatures) and imaging modalities (e.g., MRI radiomics) to enhance predictive accuracy. Additionally, integrating the nomogram with SLNB protocols could further reduce unnecessary ALNDs.

Conclusion
This study presents a robust, clinically applicable nomogram for predicting ALN pCR in node-positive breast cancer patients post-NAC. By leveraging routine clinicopathological data, the model facilitates risk-stratified axillary management, aligning with trends toward personalized oncology. Its successful validation in prospective cohorts underscores its potential to improve surgical outcomes and quality of life for patients undergoing NAC.

doi.org/10.1097/CM9.0000000000001876

Was this helpful?

0 / 0