Session Item

Auto Beam Hold (ABH) intrafraction monitoring (Varian Medical Systems, Palo Alto, USA) automatically detects radio-opaque markers on a series of kV portal images during treatment. If a marker moves outside a Search Region (SR) of user-defined diameter, the beam is automatically stopped. This study evaluated the performance of ABH in detecting prostate motion in patients with a hydrogel spacer.

We analysed 583 fractions for 25 patients treated using ABH. All patients had fiducial markers and a hydrogel spacer (Space OAR, Augmenix Inc. Bedford, USA) inserted approximately two weeks before planning CT. Planning target volume (PTV) margins were 0.5 cm to 1.0 cm. Prescriptions were 60 Gy in 20 fractions and 74-78 Gy in 37-39 fractions. Treatment plans were produced with two full arcs using Eclipse. Marker migration prior to treatment was quantified using a “Day 0” CT scan matched to prostate on the planning CT. Patient-specific SRs were defined to prevent gross errors as: SR = 2*(0.5*PTV margin + largest marker migration at day 0). Markers were excluded from ABH if their migration was more than half the PTV margin, meaning the radius of the SR circle would not exceed the PTV margin. Migration was also monitored throughout treatment using cone beam CT (CBCT) images. Patients had daily orthogonal kV imaging, matched to markers. CBCT was performed on the first three fractions, then weekly, when soft tissue matching was first performed followed by kV pair marker matching. ABH images were triggered every 72o of gantry angle.

Average marker migration at Day 0 was 0.16 cm (range 0 – 0.47 cm) and average SR diameter was 1.1 cm (range 0.7 – 1.9 cm). Migration increased to 0.22 cm after one week (p<0.05) but subsequently stabilised at 0.22 cm after 3 weeks. Beam-hold was triggered in 75/583 fractions, triggered twice in 6/75 fractions, and three times in 2/75. Treatment interruptions were patient dependent, with no beam-holds in 7/25 cases, while 5/25 patients had ≥ 8 beam-holds during their treatment course. Fractions were grouped by SR for analysis of the rate of beam-hold events (hit rate). Hit rates were 20% for SR <1.0 cm (N=260), 8% for 1 ≥ SR < 1.5 (N=221), and 6% for SR ≥ 1.5cm (N=102). For all patients, the average vector shift applied after re-imaging was 0.47 ± 0.25 cm. On 12 occasions, patient position remained unacceptable and the patient was re-setup. On seven occasions, beam-hold was exerted but upon review of new images no shifts were deemed necessary. Fractions with beam-hold events were longer on CBCT days (16.9 ± 5.2 min vs 10.8 ± 1.9 min) and non-CBCT days (10.3 ± 3.9 min vs 5.2 ± 1.2 min). The majority of this time was spent on acquiring and matching new orthogonal images.

ABH is effective in detecting and preventing excessive prostate motion during radiotherapy in patients with hydrogel spacer. Increased treatment time was generally manageable within a standard 15 min session.