Dosimetric Effects of Intrafraction Motion Detected By Triggered Imaging during Prostate SBRT

Abstract

Triggered imaging is a tool used to automatically detect intrafraction motion of implanted fiducial markers. kV images are taken with a defined frequency and the beam is gated if the fiducials move out of the set tolerance; if the beam is held, additional images are taken to determine if the patient needs to be shifted. The objective of this study was to determine the dosimetric effects of intrafraction shifts detected by triggered imaging during prostate SBRT, as well as investigating the effect of PTV margin reduction. 9 prostate SBRT patients were treated with 1-2 VMAT arcs, each receiving 3625cGy/5fx. A planar kV image was acquired every 15 seconds. If the fiducial positions were more than 3mm (our CTV to PTV margin) from their expected positions, the software automatically gated the beam and either a kV pair (8/9 patients) or a CBCT was done to reposition the patient. These IGRT shifts were reviewed retrospectively and each patients’ largest shifts in a single day were recorded; 3D vector shifts ranged from 0.9mm to 12.6mm (rotations were ignored). The shifts were used to move isocenter in the planning software and recalculate dose for 2 scenarios: 1. Assuming only 1 fraction was affected and 4 were treated accurately (1:4 plans), and 2. Assuming all 5 fractions were affected by this worst-case magnitude shift (5fx plans). Next, PTV margins were reduced from 3mm (1 patient had 5mm initially) to 2mm and 1mm and similar evaluations were done with the shifted isocenters. Finally, robustness was tested by normalizing the original plans such that the CTV in the shifted plans received at least the original plan’s coverage and evaluating the effect on normal tissues. The mean V100 PTV was significantly reduced for 2 and 3mm margins for the 1:4 plans (95.9% vs 89.4 and 89.1% with p=0.020 and p=0.004 for 2 and 3mm margins, respectively). For the 5 fx plans, the mean V100 PTV and the D100 CTV were significantly reduced from the original plan, regardless of the margin used (p=0.004). As the margins decreased, the dosimetric impact of the shifts on V100 PTV for 1:4 plans decreased; there was no trend with the targets coverage for margin reduction otherwise. Interestingly, the normal tissue effects were not as dramatic: only the urethra in the 5 fx plans showed a significant difference in maximum dose. For the robustness plans, only in the 5 fx original margin plans were the urethra max dose and the rectal volume receiving 1810cGy significantly different. Triggered imaging is a useful technology for ensuring accurate prostate SBRT treatments. Intrafraction monitoring with triggered imaging can prevent loss of target coverage and unnecessary additional dose to normal tissues and may allow for margin reductions for SBRT prostate plans. In addition, it may be possible to preemptively increase the robustness of plans without adversely affecting normal tissues even in the presence of intrafraction motion.