Development and validation of a radiogenomics model to predict axillary lymph node metastasis in breast cancer integrating MRI with transcriptome data: A multicohort study

Abstract

IntroductionTo develop and validate a radiogenomics model for predicting axillary lymph node metastasis (ALNM) in breast cancer compared to a genomics and radiomics model.MethodsThis retrospective study integrated transcriptomic data from The Cancer Genome Atlas with matched MRI data from The Cancer Imaging Archive for the same set of 111 patients with breast cancer, which were used as the training and testing groups. Fifteen patients from one hospital were enrolled as the external validation group. Radiomics features were extracted from dynamic contrast-enhanced (DCE)-MRI of breast cancer, and genomics features were derived from differentially expressed gene analysis of transcriptome data. Boruta was used for genomics and radiomics data dimension reduction and feature selection. Logistic regression was applied to develop genomics, radiomics, and radiogenomics models to predict ALNM. The performance of the three models was assessed by receiver operating characteristic curves and compared by the Delong test.ResultsThe genomics model was established by nine genomics features, and the radiomics model was established by three radiomics features. The two models showed good discrimination performance in predicting ALNM in breast cancer, with areas under the curves (AUCs) of 0.80, 0.67, and 0.52 for the genomics model and 0.72, 0.68, and 0.71 for the radiomics model in the training, testing and external validation groups, respectively. The radiogenomics model integrated with five genomics features and three radiomics features had a better performance, with AUCs of 0.84, 0.75, and 0.82 in the three groups, respectively, which was higher than the AUC of the radiomics model in the training group and the genomics model in the external validation group (both P < 0.05).ConclusionThe radiogenomics model combining radiomics features and genomics features improved the performance to predict ALNM in breast cancer.

The Role of AI in Breast Cancer Lymph Node Classification: A Comprehensive Review

“…Combining different data modalities usually leads to better performance. Therefore, such an approach should be more extensively researched in future studies to provide optimal models for breast cancer lymph node classification [ 9 , 50 ].…”

“…An interesting study by Chen et al combined DCE-MRI radiomics data with transcriptomic data from The Cancer Genome Atlas for a set of 111 patients with breast cancer [ 50 ]. Another fifteen patients were enrolled for the external validation group.…”

“…When evaluated in the validation group, the radio-genomics model showed increased performances compared to both the radiomics-only and genomics-only models (AUC 0.82 vs. 0.71 and 0.52, respectively) ( Table 5 ). The authors used logistic regression as the model of choice and performed a five-fold inner CV to optimize the model hyperparameters, whereas no dispersion metrics were reported for the results [ 50 ]. It remains to be seen whether a model with a more complex architecture would better capture the intricacies of multi-modal data.…”

Radiomic Models Predict Tumor Microenvironment Using Artificial Intelligence—the Novel Biomarkers in Breast Cancer Immune Microenvironment

“…For instance, in the study by Fan et al 49 in 2022, a radiomic model was built to predict prognosis of breast cancer using the Oncotype DX Recurrence Score, and another study by Woodard et al 50 showed the relationship between BI-RADS-based features and the Oncotype DX Recurrence Score. Additionally, another study by Chen et al 51 in 2022 showed that models combining radiomic and genomic features performed better than radiomic models or genomic models alone in predicting ALNM of breast cancer. Radiogenomics has been applied to predict the TME in several types of solid tumors but not breast cancer.…”

Lymph node metastasis prediction and biological pathway associations underlying DCE-MRI deep learning radiomics in invasive breast cancer

“…Multitype studies based on radiomics and/or deep learning features have been conducted to evaluate LNM in breast cancer. For example, in one multi-center study, a radiogenomics model combining radiomics features and genomics features improved the performance of predicting LNM in breast cancer [ 8 ]; in another based on 110 MRI images from a single center, potential pathways for the genetic mechanism of deep learning imaging phenotypes were identified [ 9 ]. The biological significance of the radiomics feature phenotype had been analyzed in multiple studies, but deep learning feature phenotypes were only identified during the image recognition process, and the quantification of specific features was not mentioned.…”