Investigating DWI changes in white matter of meningioma patients treated with proton therapy

Purpose: To evaluate changes in diffusion and perfusion-related properties of white matter (WM) induced by proton therapy, which is capable of a greater dose sparing to organs at risk with respect to conventional X-ray radiotherapy, and to eventually expose early manifestations of delayed neuro-toxicities. Methods: Apparent diffusion coefficient (ADC) and IVIM parameters (D, D* and f) were estimated from diffusion-weighted MRI (DWI) in 46 patients affected by meningioma and treated with proton therapy. The impact on changes in diffusion and perfusion-related WM properties of dose and time, as well as the influence of demographic and pre-treatment clinical information, were investigated through linear mixed-effects models. Results: Decreasing trends in ADC and D were found for WM regions hit by medium-high (30-40 Gy(RBE)) and high (> 40 Gy(RBE)) doses, which are compatible with diffusion restriction due to radiation-induced cellular injury. Significant influence of dose and time on median ADC changes were observed. Also, D* showed a significant dependency on dose, whereas f consistently showed no dependency on dose and time. Age, gender and surgery extent were also found to affect changes in ADC. Conclusions: These results overall agree with those from studies conducted on cohorts of mixed proton and X-ray radiotherapy patients. Future work should focus on relating our findings with clinical information of co-morbidities and thus exploiting such or more advanced imaging data to build normal tissue complication probability models to better integrate clinical and dose information.

Automated summary

This study evaluated changes in diffusion and perfusion-related properties of white matter (WM) induced by proton therapy, showing significant influence of dose and time on these properties and observing the impact of age, gender, and surgery extent on changes in ADC. The results suggest the potential for building normal tissue complication probability models using advanced imaging data to better integrate clinical and dose information.