Online Correction of Intrafraction Motion during Volumetric Modulated Arc Therapy for Prostate Radiotherapy: A Cohort Study Quantifying the Frequency of Shifts and Analysis of Men at Highest Risk.

Abstract

We have previously described our early experience using fiducial markers to correct for intrafraction motion during radiation therapy (RT) to the prostate using the TrueBeam Advanced imaging package. We sought to further characterize the utility of this approach in a larger cohort, and analyze factors associated with intrafraction motion. A total of 132 men with fiducial markers treated with RT for intact prostate cancer at a single center were treated with a VMAT technique using 2-3 arcs. All patients underwent planning CT after a rectal enema and same day placement of 3 fiducial markers (Gold Anchor). Triggered kV images were acquired every 10 seconds using an onboard imaging system. Intrafraction motion correction was considered if any two fiducial markers were observed beyond a 3 mm tolerance margin. A manual 2D-3D match was performed using the fiducial markers from the single triggered kV image to obtain a couch shift. Shift data were extracted from the record and verify system and expressed as a single 3-dimensional translation. Shift percent was defined as the number of instances of a >3 mm intrafraction correction divided by the total number of fractions for a given patient. Clinical variables were evaluated, including body mass index, hormone therapy (ADT), prostate, rectal bladder volumes, and rectal width (transverse dimension of rectum at the mid-gland of prostate on simulation CT). Across 2659 fractions, intrafraction motion correction was performed 582 times, in 463 fractions (17%). 101/132 patients (77%) had at least one shift during their treatment course, and 48/132 patients (36%) had shifts with an average magnitude of at least 5 mm. The median shift was 3.6 mm (range, 0-2.4 cm; IQR, 1.5-5.4 mm). 25% of men had a shift percent >20% (SP>20%). Univariate analysis revealed that only larger rectal volume or width, smaller prostate size, and use of ADT were associated with SP>20% (p<0.05). Men with rectal width in the top quartile (>3.6 cm) were more likely to have intrafraction motion corrected with SP>20% (47% vs 18%, p = 0.0016), and similarly men with rectal volume in the top quartile (>112 cc) were more likely to have SP>20% (44% vs 19%, p = 0.0067). On multivariate analysis, only rectal parameters (e.g., top quartile rectal width, HR 3.9, p = 0.0024) were correlated with a higher frequency of intrafraction motion. Intrafraction motion occurs in a significant percentage of men undergoing prostate RT with VMAT, and is correctable utilizing a common feature. On multivariate analysis, rectal volume and width were associated with larger shift percent. Treatment approaches which do not account for intrafraction motion should consider including methods of immobilization, or larger PTV margins in order to avoid marginal miss of the prostate.