Distinct Biological Response Patterns During Neoadjuvant Radiation Therapy in Soft Tissue Sarcoma: A Multiparametric MRI Clustering Analysis

Motivation: Despite standardized neoadjuvant radiation therapy (NRT) protocols in soft tissue sarcomas (STS), treatment response varies significantly between patients, yet lacks reliable non-invasive monitoring methods during treatment. Goal(s): To identify distinct radiation therapy response patterns using multiparametric MRI for early response monitoring. Approach: Prospective analysis of 20 STS patients using standardized quantitative MRI protocol at three timepoints during NRT. Results: ADC was the only parameter showing significant temporal evolution during treatment (p<0.001). Statistical clustering analysis identified three distinct, highly stable response patterns with significant inter-cluster differences in ADC (p=0.001) and Ktrans (p=0.003). Impact: This analysis reveals distinct biological response phenotypes during radiation therapy that can be identified through non-invasive imaging. These findings could enable early response prediction and therapy adaptation, potentially improving patient outcomes through personalized treatment approaches.