Image guided radiation therapy for bladder cancer: assessment of bladder motion using implanted fiducial markers.

Abstract

To determine bladder wall position variability during external beam radiation therapy (EBRT) for bladder cancer with intravesical fiducial markers using 2-dimensional (2D) and volumetric (3D) imaging registration. Twenty T2-4aN0-1 bladder cancer patients underwent definitive EBRT with concurrent chemotherapy between May 2001 and January 2012, and had intravesical fiducial marker placement before simulation. Computed tomographic (CT)-based treatment planning was used for an initial phase to deliver 45 Gy (1.8 Gy/fraction) to the bladder and pelvic lymphatics followed by a boost to the involved bladder wall for an additional 21.6 Gy (1.8 Gy/fraction). Orthogonal kilovoltage radiograph images (34-37 images/patient) were obtained daily, registered with digitally reconstructed radiographs from the planning scan. Translational corrections were made daily. A kilovoltage cone-beam CT (kVCBCT) was acquired weekly and its registration with the planning scan was compared with that day's 2D registration results. Of 739 treatments, 6% resulted in ≥15 mm displacement in 1 or more directions and 26% resulted in ≥10 mm displacement in 1 or more directions. Based on 2D registrations, the average millimeter difference between bony registration and fiducial marker registration (BR-FMR) in the right-left (RL) (R+), anterior-posterior (AP) (A+), and superior-inferior (SI) (I+) directions were: 0.5 ± 1.0 (range, -2.0 to +3.8), 1.7 ± 4.4 (range, -8.1 to +13.5) and -3.7 ± 5.8 (range, -16.8 to +8.3), respectively. For kVCBCT registrations, the average mm difference in the RL, AP, and SI directions were 0.3 ± 2.1 (range, -2.4 to +5.1), 3.1 ± 5.9 (range, -2.9 to +13.3) and -4.8 ± 8.0 (range, -16.4 to +9.5), respectively. Using intravesical fiducial markers, the largest difference in bladder motion based BR-FMR differences was in the superior-inferior direction. Because fiducial markers are target surrogates, setup using bony anatomy alone can lead to target displacements up to 13.5 mm anteriorly and 16.8 mm superiorly. This confirms a 1.5-1.7 cm minimum in planning target margins. These findings suggest a significant advantage in using intravesical fiducial markers to determine daily translational corrections.