CAN IN-VIVO DOSIMETRY ENHANCE CURRENT PRACTICE IN VMAT HEAD AND NECK TREATMENT? A retrospective service evaluation assessing the use of in-vivo dosimetry as a tool to assist in the implementation of an interventional rescan.

Abstract

Background Patient positioning during a course of external beam radiotherapy (EBRT) can be verified and corrected on a daily basis using image guided radiotherapy (IGRT) which can account for patient misalignment but not for dose discrepancies due to anatomical changes. The use of adaptive radiotherapy (ART) where the plan is adapted to the anatomical changes during the course of treatment due to these changes could be enhanced by utilising computed tomography (CT) geometrical data and dosimetric data. This retrospective service evaluation aims to address this by assessing the use of in-vivo dosimetry as a tool to implement ART. Methods The methodological approach to this small scale service evaluation was quantitative. Random sampling technique was applied as the subjects were chosen from a population who received treatment to head and neck using VMAT technique. Retrospective in-vivo dosimetry analysis was carried out on patients who underwent an interventional re-scan during treatment to their head and neck. Deformable image registration was applied to the weekly cone beam computed tomography (CBCT's) in order to assess the dose to the 'actual' patient anatomy. The dose was calculated as a percentage dose difference (%DD) between planned and actual anatomy for comparative purposes in this study to assess the evaluation outcome. Results The %DD between the planned anatomies as measured by the in-vivo software versus the dose calculated using the treatment planning system at all reference points for each of the eight patients were calculated. The resultant means of patients 1-5 (rescanned and not re-planned) versus patients 6-8 (re-planned) were analysed for significant difference at all reference points. The resultant p-value at reference points C1 and C3 was 0.08 and 0.07 respectively. C5 position compared similarly to the isocentre position with a value of 0.25 whereas C7 was poorest with a value of 0.62, due to the poor accuracy of DC at this point. Conclusion Statistical results proved a weak relationship between the %DD of the deformed (changed) and planned (presumed unchanged) anatomies meaning ART could not be implemented based on these results. Whether in-vivo dosimetry can be used to aid the decision to implement a rescan was proved to be ineffective. There was no significant difference detected between the rescanned and not re-planned patients and the re-planned patient.