2262 - Clinical Implementation of Triggered Imaging for Prostate SBRT Motion Management: An Initial Experience

H. Dong, J. P. Neylon, S. E. Tenn, J. M. Lamb, L. Valle, M. L. Steinberg, A. U. Kishan, and M. Cao; Department of Radiation Oncology, University of California, Los Angeles, Los Angeles, CA

Purpose/Objective(s): Intrafraction motion during prostate stereotactic body radiotherapy (SBRT) can result in suboptimal tumor control and increased normal tissue toxicity. While MR-guided radiotherapy with real-time gating provides excellent motion management, its accessibility is limited. The purpose of this study is to report an alternative motion tracking strategy using triggered kV imaging with automatic beam hold. Materials/

Methods: This prospective registry consecutively enrolled 15 patients receiving prostate SBRT with 40Gy in 5 fractions treated with volumetric arc therapy. Cone-beam CT (CBCT) was performed for implanted gold fiducial alignment and anatomy verification. An additional kV/MV planar imaging was acquired prior to the first treatment field to correct for any potential motion during CBCT review. During beam delivery, kV image acquisition was triggered every 20 degrees of gantry rotation. Beam holds were automatically activated when one or more fiducials moved outside a predefined search radius of 4 mm or after two consecutive triggered images on which the fiducials were deemed undetectable. Following any beam interruption, kV/MV imaging was acquired to correct the intrafraction motion before resuming treatment. Associations between beam interruption and patient age and body-mass index (BMI) were evaluated.
Results: The mean treatment time with triggered imaging was 15.8±5.8 min with a median 14.5 min (IQR, 12.4-17.2 min, N=15 patients), compared with 15.6±8.1 min when only orthogonal planar imaging correction was applied per field (N=205 patients). In 68% of the fractions (50 of 75 fractions) kV/MV correction was required immediately after the CBCT, suggesting the necessity to manage patient motion during CBCT review. Furthermore, the magnitude of the correction positively correlated to the CBCT review time (Spearman ?=.23, P=.046), indicating that longer CBCT setup likely leads to increased patient motion. Moreover, the correction also positively correlated to the number of beam interruptions (Spearman ?=.28, P=.015), suggesting that if a patient tends to move during the CBCT review, they are also likely to move during treatment delivery. The average number of beam interruptions was 0.66 per fraction and 3.27 per patient. Intriguingly, 42% of the interruptions occurred during the first third of the beam on time. The mean vertical, longitudinal and lateral intrafraction motion was 2.03, 4.71 and 1.94 mm per correction, respectively. The number of beam interruptions is associated with patient characteristics. Univariate regression analysis associated patient BMI but not age with beam interruptions (P=.021, N=15 patients). An increase of 5.5 kg/m² in BMI translates to an increase of 3 beam holds during the treatment course.
Conclusion: This study reports our institutional experience of a triggered imaging-based intrafraction motion management strategy for prostate SBRT, which could potentially be employed for personalizing treatment margin and image guidance.