In vivo endorectal dosimetry of prostate tomotherapy using dual MOSkin detectors

Verification of dose to the anterior rectal wall in helical tomotherapy to the prostate is important due to the close proximity of the rectal wall to the treatment field. The steep dose gradient makes these measurements challenging. A phantom-based study was completed, aimed at developing a system for measurement of anterior rectal wall doses during hypofractionated prostate stereotactic body radiotherapy (SBRT) utilizing tomotherapy delivery. An array of four dual MOSkinTM dosimeters, spaced 1 cm apart, was placed on a replica Rectafix® immobilization spacer device. This Perspex probe is a more rigid alternative to rectal balloons, to improve geometric reproducibility. The doses at each point were measured in real time and compared to doses calculated by the treatment planning system (TPS). Additionally, distance-to-agreement (DTA) measurements were acquired to assist in the comparison of measured and predicted doses. All dual MOSkin detectors measured dose to within ± 5% of the TPS at the anterior rectal wall. Whilst several points were outside of experimental error, the largest deviation from the TPS predicted dose represented a DTA of only 1.3 mm, within the acceptable DTA tolerance of 3 mm. Larger deviations of up to -11.9% were observed for the posterior and side walls; however, if acceptable DTA measurements are accounted for, then an agreement of 75% was observed. Although larger differences were observed at the other rectal wall locations, the overall effect of dose at these points was not as significant, given the lower doses. Despite the very high-dose gradient region, real-time measurements of the anterior rectal wall doses were within acceptable limits of TPS-predicted doses. The differences between measured and planned data were due to difficulties in precisely locating each detector on the TPS dose grid, which presented large variations in dose between CT voxels in regions of steep dose gradients. The dual MOSkin system would, therefore, be a useful device for detecting errors in real time, such as patient shifts or incorrect setup, during tomotherapy of the prostate.