JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Deep learning assisted non-invasive lymph node burden evaluation and CDK4/6i administration in luminal breast cancer

  • 0Department of Breast Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China.
Iscience +

|

July 18, 2025

|

July 18, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

A new model, the lymph node prediction network (LNPN), accurately assesses lymph node burden in luminal breast cancer patients. This tool aids in optimizing CDK4/6 inhibitor therapy and reducing unnecessary surgeries.

Area Of Science

  • Oncology
  • Medical Imaging
  • Biostatistics

Background

  • Accurate lymph node staging is crucial for luminal breast cancer treatment, especially with decreasing axillary surgery rates.
  • Current methods may be insufficient for precise recurrence risk evaluation in the context of de-escalated axillary surgery.
  • Optimizing CDK4/6 inhibitor therapy requires precise assessment of lymph node involvement.

Purpose Of The Study

  • To develop and validate a multi-modal model, the lymph node prediction network (LNPN), for lymph node burden assessment.
  • To evaluate LNPN's performance in differentiating lymph node metastasis in luminal breast cancer patients.
  • To explore LNPN's utility in guiding treatment decisions, including CDK4/6 inhibitor therapy and axillary lymph node dissection.

Main Methods

  • Development of the LNPN, a multi-modal model integrating clinicopathological parameters and ultrasonographic features.
  • Multicenter cohort study involving 411 luminal breast cancer patients.
  • Performance evaluation using Area Under the Curve (AUC) for binary and ternary lymph node burden classification.

Main Results

  • LNPN achieved an AUC of 0.92 for binary classification (N0 vs. N+) and 0.82 for ternary classification (N0/N1-3/N ≥ 4).
  • In patients with 1-2 metastatic lymph nodes after sentinel lymph node biopsy (SLNB), LNPN predicted high-burden metastases (N ≥ 4) with an AUC of 0.77.
  • The model demonstrated robust performance in a diverse patient cohort.

Conclusions

  • LNPN offers a non-invasive method for assessing lymph node metastasis and recurrence risk in luminal breast cancer.
  • This tool has the potential to reduce unnecessary axillary lymph node dissection (ALND).
  • LNPN can aid in decision-making for CDK4/6 inhibitor therapy in luminal breast cancer patients.

Keywords: