First-in-human clinical translation of oxygen-enhanced MRI onto an MR Linac

Abstract

Purpose or Objective MR Linac (MRL) systems enable delivery of radiotherapy with on-line MRI. Functional imaging on the MRL permits identification, mapping and tracking of tumour sub-regions, with the potential to introduce biologically adaptive radiotherapy (Datta 2018 ClinOnc). Here we report the first application of oxygen-enhanced MRI (OE-MRI) on an MRL to identify and map tumour hypoxia. Materials and Methods A dynamic 3D inversion recovery turbo field echo (IR-TFE) OE-MRI sequence was developed on a Philips Ingenia 1.5T MR system (MRSim) in 12 healthy volunteers and 4 patients with treatment naïve head and neck (H&N) carcinoma. This protocol was modified and implemented on an Elekta-Philips 1.5T MRL with different receive coils and required modification of repetition and echo times due to hardware differences. Other hardware fitting steps included retrofitting of gas delivery in the MRL bunker; installation of gas ports, delivery tubing and MR conditional blender providing 15 l/min; and installation of a contrast agent power injector for dynamic contrast-enhanced (DCE)-MRI. Participants were recruited after providing written informed consent on a local ethics approved protocol. MRL OE-MRI was acquired in 4 healthy volunteers and a patient with H&N carcinoma. T1 measurement was carried out using an IR-TFE sequence with multiple TIs and a dynamic IR-TFE series during delivery of medical air (volumes 1-25), followed by 100% oxygen (volumes 26-70) and back to medical air (volumes 71-91). Analysis was carried out in MATLAB (Mathworks). Native T1 maps enabled conversion of signal change to ΔR1 (where ΔR1 = R1,O2 – R1,air). In all volunteers, ΔR1 was calculated in the nasal concha (NC). Patient tumour volumes were delineated on T1 post contrast images and ΔR1measurements obtained. Perfused tumour voxels (DCE signal increase p<0.05) were classified as oxygen enhancing (Oxy-E) (OE signal increase p<0.05; suggesting normoxia) or oxygen refractory (Oxy-R) (suggesting hypoxia). Results Volunteer NC ΔR1 was 0.059 ± 0.027 s-1 (p < 0.001, group ΔR1 change) and 0.065 ± 0.030 (p < 0.001) on the MRSim and MRL respectively (Figure 1a-b). There was no significant difference in NC ΔR1 between the groups on the two systems (p=0.6, unpaired t-test). Patient tumour mean ΔR1 = 0.031 ± 0.035 (p < 0.001) (n=5 patients) on the MRSim and ΔR1 = 0.035 ± 0.011 (p < 0.001) (n = 1 patient) on the MRL (Figure 1c-d). Hypoxia maps showed distinct ‘hypoxic’ and ‘normoxic’ tumour sub-regions and are presented with the tumour ΔR1 map in figure 2. Conclusion We have successfully translated OE-MRI onto an MRL for the first time. OE-MRI measurements in normal tissue (NC) and tumour are consistent between the MRSim and the MRL at 1.5 T. This novel technique facilitates introduction of hypoxia adapted radiotherapy on the MRL.