MRI radiomics and biological correlations for predicting axillary lymph node burden in early-stage breast cancer

Abstract

Background and aims: Preoperative prediction of axillary lymph node (ALN) burden in patients with early-stage breast cancer is pivotal for individualised treatment. This study aimed to develop a MRI radiomics model for evaluating the ALN burden in early-stage breast cancer and to provide biological interpretability to predictions by integrating radiogenomic data.

Methods: This study retrospectively analyzed 1211 patients with early-stage breast cancer from four centers, supplemented by data from The Cancer Imaging Archive (TCIA) and Duke University (DUKE). MRI radiomic features were extracted from dynamic contrast-enhanced MRI images and an ALN burden-related radscore was constructed by the backpropagation neural network algorithm. Clinical and combined models were developed, integrating ALN-related clinical variables and radscore. The Kaplan-Meier curve and log-rank test were used to assess the prognostic differences between the predicted high- and low-ALN burden groups in both Center I and DUKE cohorts. Gene set enrichment and immune infiltration analyses based on transcriptomic TCIA and TCIA Breast Cancer dataset were used to investigate the biological significance of the ALN-related radscore.

Results: The MRI radiomics model demonstrated an area under the curve of 0.781-0.809 in three validation cohorts. The predicted high-risk population demonstrated a poorer prognosis (log-rank P < .05 in both cohorts). Radiogenomic analysis revealed migration pathway upregulation and cell differentiation pathway downregulation in the high radscore groups. Immune infiltration analysis confirmed the ability of radiological features to reflect the heterogeneity of the tumor microenvironment.

Conclusions: The MRI radiomics model effectively predicted the ALN burden and prognosis of early-stage breast cancer. Moreover, radiogenomic analysis revealed key cellular and immune patterns associated with the radscore.

Keywords: Axillary lymph node; Breast cancer; Genomics; Magnetic resonance imaging; Radiomic.

Fig. 1

Flowchart of patient selection and study design. SLNB: sentinel lymph node biopsy; ALND: axillary lymph node dissection; ALN: axillary lymph node burden

Fig. 2

Model performance analysis. Receiver operating characteristic (ROC) curves for various models are presented for centers I (a), II (b), III (c), and IV (d), along with the AUC and 95% confidence intervals shown in the bottom right. Kaplan–Meier curves illustrate the survival probabilities of patients in the predicted high- and low-ALN burden groups for in Center I (e) and DUKE (f)

Fig. 3

Three typical cases demonstrated the clinical application of radiomics models. The ALN burden was correctly classified according to the radscore, and MRI image-based decision support was provided. Pathological images are 20×Microscope field

Fig. 4

Transcriptomic and immunological analysis related to radscore. (a) Volcano plot of differentially expressed genes, with upregulation indicated in red, downregulation in blue, and nonsignificant genes in gray. (b) Clustered heatmap of pathway enrichment analysis. (c) Bubble chart of pathway enrichment based on gene ratio, with bubble size representing gene count. (d) Box plots of immune cell scores across two radscore groups

Fig. 5

Spearman correlation heatmap between radiomic features and immune marks. Blue indicates a negative correlation, while red signifies a positive correlation. The deeper the color, the stronger the correlation. CLP, common lymphocyte precursors