Onko-Gene

Critical literature watch in the molecular oncology of solid tumors

FREN

Clinical · Molecular biology · Bioinfo & AI

Development and validation of a machine learning clinicogenomic model to improve prognostic stratification in ER-positive/HER2-negative early breast cancer

Summary

A machine learning-based clinicogenomic model has been developed and validated to improve prognostic stratification in ER-positive/HER2-negative early breast cancer. This model integrates a research-grade implementation of the 21-gene recurrence score, a 14-gene immunoglobulin signature, and clinicopathologic features. Validated across four independent datasets totaling 5132 patients, it demonstrated improved discrimination and identified larger low-risk groups compared to the standalone 21-gene recurrence score. Explainability analyses revealed time-dependent feature contributions, highlighting the potential of this integrative approach for more accurate risk assessment.

Critical appraisal

Clinical. This clinicogenomic model represents a significant step forward for prognostic stratification in ER+/HER2- early breast cancer. By identifying larger low-risk groups (relative increases of 33.8 to 111.8%) while maintaining or improving survival estimates, it could potentially reduce overtreatment in some patients. The improved discrimination (absolute increase in concordance index of 0.035 to 0.058) suggests a better ability to distinguish at-risk patients. Although the study is retrospective, these findings warrant prospective clinical trials to validate its impact on therapeutic decision-making and potentially change clinical practice within 3 to 5 years.

Molecular biology. The model integrates a research-grade implementation of the 21-gene recurrence score, a 14-gene immunoglobulin signature, and clinicopathologic features. The use of microarray and RNA-sequencing data from multiple cohorts (n=5132) strengthens the robustness of the approach. The integration of immune markers, whose contribution is more stable over time according to explainability analyses, adds a relevant biological dimension. This molecular discovery, while based on existing signatures, is an advance in their combination for better prediction, paving the way for more comprehensive clinical tests.

Bioinfo & AI. The development of a random survival forest (RSF) model is an innovative application of machine learning for integrating heterogeneous data (genomic, immune, clinical). External validation on three independent datasets is a key strength, demonstrating the model’s generalizability. The use of time-dependent explainability analyses is crucial for understanding the contributions of different features, improving transparency and trust in the model. If deployed clinically, this model could optimize oncology decision workflows by providing more nuanced risk stratification than current tools.

The summary and the appraisal were written by a language model from the title and abstract of the original article. A curator decided to publish it. Methodology