Multiparametric magnetic resonance imaging for radiotherapy response evaluation in high-risk soft tissue sarcoma: A pilot study

Background and purpose: Soft tissue sarcomas (STS) are rare mesenchymal tumors for which no clinically validated quantitative magnetic resonance imaging (qMRI) parameters exist yet. This study explores repeatability and association with histopathology of qMRI parameters during and after neo-adjuvant angiogenesis inhibition (oral pazopanib) and radiotherapy for localized, high-risk STS. Materials and methods: For fifteen patients, qMRI parameters, including apparent diffusion coefficient (ADC), volume transfer constant (K^trans) and T₂ relaxation times were acquired twice at baseline (B1 and B2), twice during neo-adjuvant treatment and pre-surgery. For all three parameters, the mean was determined per tumor. Subsequently, repeatability coefficient (RC or %RC) was assessed from B1 and B2 mean values. Mixed models were estimated to study the association between percentage viable cells from histopathology and absolute change from baseline (ΔqMRI) for ADC mean and percentage change from baseline (%ΔqMRI) for T₂ and K^trans at each time point. Results: RC was 0.17 × 10^-3 mm²/s for ADC and %RC was, 5 % and 65 % for T₂ and K^trans, respectively. The changes in mean ADC and T₂ showed both increases and decreases at each timepoint, whereas mean K^trans predominantly showed decreases. ΔqMRI for ADC mean, %ΔqMRI for T₂ mean and %ΔqMRI for K^trans mean showed no statistically significant association with % viable cells. Conclusion: This pilot study reported relatively low repeatability coefficients for ADC and T₂ and a higher repeatability coefficient for K^trans and showed heterogeneous changes in qMRI parameters in fifteen STS patients, however with no association between these parameters and percentage viable cells.