Fifty-seventh annual meeting of the American association of physicists in medicine
TH-CD-204-05: Developing a Non-Invasive MRI Prognostic Biomarker to Predict Response to Hypofractionated Radiotherapy
To investigate the usefulness of Oxygen-Enhanced Magnetic Resonance Imaging (OE-MRI), the so-called semi-quantitative blood oxygen level dependent (BOLD) and tissue oxygen level dependent (TOLD) responses, as well as quantitative apparent transverse relaxation rate (R2*) and longitudinal relaxation rate (R1) responses, as a prognostic biomarker of radiation response.
Twenty-two (22) subcutaneous Dunning R3327-AT1 syngeneic rat prostate tumors were evaluated. When tumors reached (0.7–2 cm3), OE-MRI was performed at 4.7 T. The following day tumors were irradiated with 1 fraction of 15 Gy to the gross tumor volume (GTV) while inhaling air or oxygen. MRI and irradiation were repeated 1 week later. Radiation response was assessed by tumor growth delay measurements as time for tumor to reach 4 times the initial irradiated volume (VQT). Survival estimates from VQT were assessed using Kaplan-Meier.
Inhaling oxygen during hypofractionation significantly improved response to treatment based on survival probability estimates (P < 0.0001, log-rank test). Significant correlations were determined between VQT and pretherapy or between fraction OE-MRI baseline values and parameter changes before the first and second fraction of dose. Immunohistochemistry confirmed tumor reoxygenation between dose fractions consistent with OE-MRI.
Inhaling oxygen during irradiation enhanced response in this tumor type, which was reported to exhibit substantial hypoxia. OE-MRI response to inhaling oxygen was related to response of individual tumors, suggesting utility as prognostic biomarkers. Investigations will be required in other tumor types to validate these observations, but they show that OE-MRI parameters are related to tumor response to hypofractionation, suggesting the potential for patient stratification for personalized radiotherapy treatment planning.
This study was supported in part by funds from the National Cancer Institute 5R01CA139043, NCI Cancer Center Support Grant 1P30 CA142543. MRI experiments were performed in the Advanced Imaging Research Center, supported by NIH National Institute of Biomedical Imaging and Bioengineering Resource grant EB015908.