The role of magnetic resonance imaging (MRI) in external beam radiotherapy treatment planning (RTP) for cervical cancer

Background Accurate delineation of treatment volumes for radiotherapy is a crucial step in radiation treatment, but can be associated with considerable uncertainty and variability. Magnetic resonance imaging (MRI) has a proven benefit in the staging of cervical cancer and for image guided adaptive brachytherapy; however there is a paucity of data for the role of MRI in delineating external beam radiotherapy treatment volumes. Aims 1) Quantify the proportion of women with cervical cancer in whom there is evidence to support the use of MRI for radiotherapy planning (RTP), 2) Evaluate the effect of an educational intervention on MRI consistency of volume delineation for external beam target volumes, and 3) Assess volume delineation for external beam radiotherapy on MRI compared to the current standard of care, computer tomography (CT) imaging. Methods Data from an Australian model for estimating the optimal radiotherapy utilization rate for cervical cancer were used and a literature review was performed for each patient group to assess the evidence for the use of MRI for RTP. To evaluate the second aim contours from eleven clinicians were analysed before and after an educational intervention. For the third aim, twenty patients with locally advanced cervical cancer underwent a dedicated MR simulation with a 3T Skyra MRI following CT simulation, four clinicians independently contoured each CT and MRI dataset. The Dice Similarity Co-efficient (DSC) and Mean Average Surface Distance (MASD) was calculated for each structure. Results The use of an MRI for radiotherapy planning is supported in 49% of women overall with cervical cancer. The educational intervention resulted in an improvement in contouring the gross tumour, however a variable response for the other target volumes was observed. MRI contouring was associated with less variability between clinicians compared to CT contouring for most structures including the gross tumour, uterus and parametria. The overall target volume demonstrated a high level of contouring consistency on both CT and MRI, with a small improvement in the MASD seen for MRI contouring. Conclusion MRI simulation for external beam radiotherapy in cervical cancer reduces target volume variability compared to CT