Quantification of intrafraction prostate motion and its dosimetric effect on VMAT.

Abstract

Intrafraction prostate motion degrades the accuracy of radiation therapy (RT) delivery. Whilst a number of metrics in the literature have been used to quantify intrafraction prostate motion, it has not been established whether these metrics reflect the effect of motion on the RT dose delivered to the patients. In this study, prostate motion during volumetric modulated arc therapy (VMAT) treatment of 18 patients and a total of 294 fractions was quantified through novel metrics as well as those available in the literature. The impact of the motion on VMAT dosimetry was evaluated using these metrics and dose reconstructions based on a previously validated and published method. The dosimetric impact of the motion on planning target volume (PTV) and clinical target volume (CTV) coverage and organs at risk (OARs) was correlated with the motion metrics, using the coefficient of determination (R 2 ), to evaluate their utility. Action level threshold for the prostate motion metric that best described the dosimetric impact on the PTV D95% was investigated through iterative regression analysis. The average (range) of the mean motion for the patient cohort was 1.5 mm (0.3-9.9 mm). A number of motion metrics were found to be strongly correlated with PTV D95%, the range of R 2 was 0.43-0.81. For all the motion measures, correlations with CTV D99% (range of R 2 was 0.12-0.62), rectum V65% (range of R 2 was 0.33-0.58) and bladder V65% (range of R 2 was 0.51-0.69) were not as strong as for PTV D95%. The mean of the highest 50% of motion metric was one of the best indicator of dosimetric impact on PTV D95%. Action level threshold value for this metric was found to be 3.0 mm. For an individual fraction, when the metric value was greater than 3.0 mm then the PTV D95% was reduced on average by 6.2%. This study demonstrated that several motion metrics are well correlated with the dosimetric impact (PTV D95%) of individual fraction prostate motion on VMAT delivery and could be used for treatment course adaptation.