MRI-guided radiotherapy for renal tumours

Abstract

The current surgical treatment for renal cell carcinoma (RCC) experiences complications due to the invasiveness of the treatment. Minimally invasive alternative treatments, such as radiofrequency and cryogen ablation, have fewer complications but a lower local control. Radiotherapy is an established minimally- to non-invasive treatment and might improve morbidity or outcome for RCC patients. A precise radiation technique will be necessary since the kidney moves during respiration and is very radiosensitive. The feasibility of two possible treatments, MRI-guided brachytherapy and an MRI-linac treatment, was studied. Brachytherapy was simulated on the MR-images of seven selected patients with increasing lesion size. This showed that for all patients a catheter could be drawn in a position where it could be inserted during treatment. As the treatment plan (1x20 Gy to the Gross Tumour Volume (GTV)) for the largest lesion (32 mm diameter) violated the dose constraint of the kidney and the plans for the other lesions (11-27 mm diameter) obeyed all dose constraints, MRI-guided radiotherapy seems dosimetrically feasible for lesions smaller than approximately 30 mm diameter. An MRI-linac (MRL) is combination of a 6MV linac and a 1.5T MRI that is under development at the UMC Utrecht. Three major aspects of future MRL-treatments for renal tumours (fast imaging, kidney motion and dose distributions), were investigated. The use of pencil-beam navigators, a fast 1D MRI sequence, for kidney imaging was shown. The accuracy of tracking water-air, water-fat and fat-air boundaries was smaller than 0.5 mm in a phantom, with a time lag of approximately 30 ms. By correlating 2D kidney images and pencil-beam navigators the possibility of complete organ tracking by 1D navigators was shown. Breath-hold monitoring based on navigator guidance was demonstrated in seven patients with an accuracy smaller than 1.5 mm. To determine the efficiency of gated radiotherapy, kidney motion during 2 minutes of free breathing and 10 consecutive breath holds was studied in 15 patients with a renal lesion, using 2D cine MRI. The variation in kidney expiration position during free breathing was smaller than 2 mm for 80% of the patients and for all patients smaller than for breath hold. No relevant time trends were detected during free breathing. Gating on free breathing is, for gating windows of 1 to 5 mm, also more efficient than uncorrected breath hold, which can be improved by a base-line correction. Since free breathing seems more comfortable for patients, it is preferred during MRL-kidney-treatments. An MR-based virtual treatment (1x25 Gy) for twenty patients with a renal lesion showed that for 10 patients (50%) a static plan within the dose constraints could be made. The other plans showed that when bowel is inside the PTV, the maximum bowel dose can be too high. If the tumour is large (>30mm) the healthy- kidney-constraint can be violated, specifically for mid-pole-tumours. In all patients where the static plan constraints were obeyed, the GTV-dose and the kidney-dose remained within limits for a gating window up to 12 mm. These results indicate that an MRL-treatment for small renal masses is feasible