Artificial intelligence for predicting pathological complete response to neoadjuvant therapy in triple-negative breast cancer: A systematic review and meta-analysis.
Andrea Carlini, Flavia Foca, Marianna Sirico et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis evaluated the effectiveness of artificial intelligence (AI) models in predicting pathological complete response (pCR) to neoadjuvant therapy in early triple-negative breast cancer (eTNBC). The analysis included 20 cohorts from 19 studies, revealing a pooled AUC of 0.79 (95% CI, 0.75-0.83) for AI-based predictions, indicating high discriminative performance despite significant heterogeneity.
Key findings
- AI-based models achieved a pooled AUC of 0.79 for predicting pCR after neoadjuvant therapy.
- Substantial heterogeneity was observed among studies, with an I² of 73.1%.
- Models using radiomics data showed numerically higher AUCs compared to non-radiomics approaches.
- Funnel plot asymmetry indicated potential small-study effects or publication bias.
Why it matters
Understanding the predictive capabilities of AI in oncology can enhance treatment planning and patient outcomes. However, the observed heterogeneity and potential biases highlight the need for caution in clinical application until further validation is achieved.