2191 - Real-Time Monitoring and Appropriate Margin Selection in Linac-Based Prostate SBRT: Analysis of Two Different Intrafraction Motion Monitoring Systems

Purpose/Objective(s): To quantify intrafraction prostate motion using two continuous monitoring systems—system A (electromagnetic transmitter-based) and system B (4D transperineal ultrasound-based)—and assess their impact on motion patterns.

Materials/Methods: Data were collected from 92 treatment fractions using system A (47 patients) and 100 fractions using system B (30 patients). All patients received SBRT regimen (1–5 fractions) and underwent a strict bowel and bladder preparation regimen. Prostate motion was recorded at 1 Hz for system A and 2 Hz for system B from the start of CBCT to the treatment completion, without any zeroing due to patient repositioning. A mixed-effects regression model analyzed motion trends over time, and Kaplan-Meier failure analysis estimated time to exceed motion thresholds of 2 mm, 3 mm, 4 mm, and 5 mm in each direction and in any direction, providing insight into motion progression and margin considerations.

Results: Median monitoring time was 13 minutes for system A (range: 5.9–35.6) and 6.4 minutes for system B (range: 4.9–29.3). Both systems showed predominant posterior motion, with consistent lateral and vertical patterns over time. However, in the longitudinal direction, the two systems exhibited a clear divergence (system × time coefficient = -0.07 vs. 0.05, P < 0.001). The mixed-effects regression model revealed a significant relationship between motion and time across all directions, with the most pronounced increase observed in the posterior direction (time coefficient = 0.20, P < 0.001). Kaplan-Meier failure curves showed comparable results between the two systems for all the analyzed thresholds, except for a significant difference in the longitudinal direction at the 2 mm threshold (log-rank test, P=0.005). The failure odds were 7.3% vs. 4.2% at 3 minutes, 14.3% vs. 7.0% at 5 minutes, and 36.7% vs. 8.6% at 8 minutes. A significant percentage of fractions (>5%) exceeded the 2 mm motion threshold at 1.8 minutes, the 3 mm threshold at 3.8 minutes, the 4 mm threshold at 6.6 minutes, and the 5 mm threshold at 8.1 minutes.

Conclusion: Significant differences between the two systems were observed in the longitudinal direction, with system A and system B exhibiting opposite motion patterns. Furthermore, system B provided partial immobilization in the longitudinal axis due to probe pressure against the perineum. These findings suggest that for fast linac-based SBRT treatments, intrafraction prostate motion can be mitigated with 4-5 mm margins without the need for continuous monitoring. However, if 2-3 mm margins (especially in the posterior direction) are used, intrafraction monitoring is essential to prevent target missing and overdosing to organs at risk.