Prognosticating axillary lymph node metastasis in breast cancer through integrated photoacoustic imaging, ultrasound, and clinical parameters.

Purpose: To develop and validate a predictive model for axillary lymph node metastasis (ALNM) in breast cancer (BC) by integrating clinicopathological factors, ultrasound features, and photoacoustic imaging-derived SO₂ measurements, aiming to improve diagnostic accuracy and provide comprehensive clinical insights.

Methods: A total of 317 BC patients were included, with the cohort split into a training set (70%) and a testing set (30%). Univariate and multivariate logistic regression identified key predictive factors, leading to the creation of three models: ModA (clinicopathological factors only), ModB (clinicopathological and ultrasound features), and ModC (clinicopathological, ultrasound, and SO₂ measurements from photoacoustic imaging). De-Long test and ROC curve were used to evaluate and compare the diagnostic performance of the models.

Results: Multivariate analysis showed that maximum diameter, Ki67 expression, AUS report and SO₂ levels were identified as significant risk factors for ALNM. ModA achieved an AUC of 0.776 (95% CI: 0.691-0.862), ModB improved to 0.824 (95% CI: 0.738-0.909), and ModC demonstrated the highest performance with an AUC of 0.882 (95% CI: 0.815-0.950) in the testing set. The results highlight that the comprehensive model (ModC), integrating clinical, ultrasound, and photoacoustic imaging data, provides superior predictive accuracy for ALNM.

Conclusion: Integrating SO₂ measurements with traditional clinical and ultrasound data can substantially enhance the prediction of ALNM in BC patients. This combined model offers a comprehensive and reliable decision support tool for the preoperative risk assessment of axillary lymph nodes in BC.