Mark Hill from the University of Oxford discussed the development of a gated treatment head assembly for Xstrahl's SARRP small-animal irradiator, by a multidisciplinary team from the University's Department of Oncology. "During an animal's breathing cycle there is significant movement," he said. "The amount of movement depends on the position within the animal, and can be up to the order of 5 mm."

Hill and colleagues also incorporated MR imaging into the preclinical treatment chain. MRI complements the SARRP's existing cone-beam CT (CBCT) capability and enables visualization of abdominal targets that are not easily located using CBCT. The new gating system, Hill explained, can be used during both imaging and treatment of the animal.

To perform breathing-gated, MR-guided preclinical radiotherapy, a mouse is first imaged using MRI, and then transferred to the SAARP where CBCT is performed. The two images are co-registered (which takes 5–10 min) and a treatment plan generated. The team created a dedicated holder to transfer the mouse from the MRI to the SARRP. Hill noted that moving the mouse back and forth between the two systems only caused minimal movement in its internal organs.

Hill's colleague Boris Vojnovic presented the details of the new gating system. The animal's breathing motion, he explained, is detected by illuminating its body with modulated near-infrared (850 nm) light delivered via an optical fibre, and measuring changes in diffuse reflectance via a second fibre. Tests of this breathing monitoring system revealed that it could sense surface movement of less than 0.1 mm.

The recorded respiratory signal is processed and used to determine a trigger level defining the "no motion" periods when imaging or irradiation can be performed. Pre-and post-inhalation margins can be programmed to minimize movement, while not over-extending irradiation time. The gating system itself comprises a rotary shutter that opens and closes by rotating through 90°. The high-speed shutter exhibits an open–close latency of less than 20 ms and is silent.

Vojnovic noted that the system uses information from the previous breathing cycle to define the next gating period (which is possible in animals due to their fairly regular breathing pattern). "This approach allows us to adapt the gating to any decrease or increase in the animal's breathing rate," he explained.

Lastly, as the gating period is defined by the individual animal's breathing pattern, the delivered dose must be monitored. To do this, the researchers use an integrated transmission ionization chamber. Vojnovic says that this chamber hardens the treatment beam slightly, but only by about the same amount as an aluminium imaging filter.

Demonstrating gating

To demonstrate the effectiveness of the new treatment head, Hill displayed a movie showing X-ray images of an anaesthetised mouse, recorded at 6 frames/s. Without gating to the animal's breathing, movement of the diaphragm could clearly be seen. When gating was applied, and X-ray images only recorded in periods of minimal movement, the resulting movie appeared as a stationary image. "This shows that with X-ray gating, we are only irradiating when the animal's breathing is in the rest phase," said Hill.

To test the gating system further, the researchers employed a fast X-ray camera to image a mouse with an implanted BB fiducial. They examined a heavily anaesthetised mouse, breathing at about 40 bpm, and a lightly anaesthetised mouse, breathing at about 100 bpm. In both cases, the system could track the displacement of the BB. "Applying gating enabled us to completely remove the movement of BB," noted Hill.

Finally, to check the system's ability to perform gated beam delivery, Hill and colleagues placed EBT3 radiochromic film on the animal's chest and irradiated the film with a collimated X-ray beam. Dose was delivered to achieve a specified beam-on time. The gated irradiation created a clear image on the film, while the non-gated beam caused a blurred image. "The dose profile from the ungated field showed a lower target dose and increased dose to surrounding tissues," Hill explained. "You need to be careful that you deliver the full dose that you think you are giving."

"Breathing-gated irradiation using the new treatment head facilitates irradiation in moving targets, and is an important new tool for preclinical radiotherapy studies," Vojnovic concluded. "The use of MRI combined with CBCT enables identification of targets not readily visible with CT alone."

An alternative approach

Speaking later in the conference session, Anne-Marie Frelin-Labalme from ARCHADE (Advanced Resource Center for Hadrontherapy in Europe) in France discussed her work on gated irradiation of free-breathing animals.

Using PXi's X-RAD 225Cx preclinical IGRT research system, Frelin-Labalme and colleagues implemented gating by moving a shutter in and out of the X-ray beam. The shutter can switch between open and closed in about 120 ms. To monitor the animal's breathing, they used a commercial pressure-balloon-based system, the signal from which drives the opening and closing of the shutter.

To study the accuracy of this approach, the researchers developed a dynamic phantom comprising a PMMA probe with holes for water or contrast agents and space to include dosimetric film. The probe is driven by a motor to perform motion of variable amplitude, frequency and waveform.

The team first evaluated the phantom using the 2D fluoroscopy mode of the X-RAD 225Cx. Measuring the phantom position over time – for sinusoidal motion at frequencies of 0.5, 1 and 2 Hz, and motion amplitudes from 0.3 to 10.4 mm (representative of mouse and rat respiration) – revealed good agreement between the waveform and the phantom's motion. "The phantom could also reproduce more realistic motion, such as exponential waveforms," said Frelin-Labalme.

Next, the researchers investigated motion-compensated microPET imaging, by inserting PET tracer solutions into the phantom. PET scans recorded with and without gating, for sinusoidal phantom motion of various amplitudes and frequencies, showed that without gating, the width of the imaged distribution increased with motion amplitude. Gating eliminated this effect, with reconstruction using eight bins per respiratory cycle performing better than that with four bins per cycle.

The researchers also evaluated gated radiation delivery with their proposed system. Using a 10 mm diameter collimator, they irradiated the phantom with EBT3 film placed on top, whilst immobile and undergoing sinusoidal (1 Hz, 4.2 mm) motion. They also performed gated irradiation, with gating widths of 580, 370 and 270 ms.

With motion and no gating, the image on the film was blurred compared to the stationary case. Dose profiles also showed that the ungated beam delivered less dose to the 9 mm diameter target and slightly more dose to the surrounding volume. Gating successfully compensated for the sinusoidal motion and minimized the image blur. "With gating, we could almost reproduce the nominal dose distributions," said Frelin-Labalme.

She concluded that implementation of respiratory gating with the X-RAD 225Cx precision irradiator is feasible, and that the dynamic phantom provides a useful tool for evaluating gating, for both imaging and irradiation. Next, the team plans to integrate PET scanning into treatment planning, optimize the gating protocols using more realistic waveforms and perform in vivo evaluation of this approach.

Related stories

• Radiotherapy needs preclinical research
• Small-animal irradiation: easing translation into clinical studies
• Progress in micro-irradiation technology
• SmART symposium launches in Maastricht