The clinical study was performed at OncoRay in Dresden, using the first prototype of a knife-edge slit camera being developed by IBA. The camera works by directing the prompt gamma rays emitted during irradiation through a tungsten slit collimator, placed perpendicular to the beam. The prompt gammas are incident upon a segmented detector, creating a one-dimensional depth profile of the beam path. If the proton range changes, the prompt gammas pass through the slit at a slightly different angle and hit different detector segments, revealing any range shift.

"Our goal was to evaluate the [prompt gamma imaging] system under clinical conditions, with the final goal of reducing range uncertainties and irradiation of normal tissues," Richter told the conference delegates.

Richter explained that the slit camera is ideally suited to pencil-beam scanning (PBS) proton therapy, where information is available for every spot and absolute range analysis is possible. Measurements during passively double-scattered (DS) proton therapy are more complex, as the additional material present in the beam line generates an increased neutron background. In addition, only one prompt gamma signal is recorded for the entire field. However, at the time, the centre only had DS proton therapy available for clinical use, so that is what they used.

To determine whether prompt gamma imaging was possible for DS proton therapy, the researchers first performed a series of phantom experiments. To eliminate the background signal, they recorded data with the slit closed to acquire a background signal. Subtracting this background from a signal recorded with the slit open resulted in a clear prompt gamma profile. "This was quite successful," noted Richter. Tests in water and head phantoms demonstrated that the slit camera could detect 2 mm global range shifts.

Finally, the team examined whether it is possible to gain more localized information from the prompt gamma signal. Using a rotating modulator, they recorded prompt gamma profiles for each rotation step. Time resolved analysis of these measurements enabled separation of profiles from different iso-energy layers, providing additional information.

The clinical study

In August of last year, Richter and colleagues used the prototype slit camera to measure the prompt gamma ray depth distribution during treatment of a patient with head-and-neck cancer. The patient received seven fractions of DS proton therapy, using three fields. Richter emphasized that the measurements, which only added minimal extra time, were performed to evaluate the method and were not used to influence the patient's treatment.

The team performed prompt gamma imaging on one proton field during six treatment fractions, with the slit closed to take a background reading during the other fraction. They evaluated inter-fractional variations in the time-integrated prompt gamma profiles, as well as profiles corresponding to different steps of the modulator wheel (the iso-energy layers). In-room CT was also performed for three fractions to allow dose reconstruction.

Prompt gammas were successfully detected during the six fractions with the open slit. After smoothing and application of auto-detection, the time-integrated profiles revealed inter-fractional global range variations of within ±2 mm. "The prompt gamma imaging measurements were in agreement with the control CT-based dose recalculations," said Richter. He noted that the iso-energy layers could also be measured, and that their profiles were consistent with the time-integrated profiles.

Richter concluded that the team successfully performed the world's first clinical prompt-gamma-based proton range verification, noting that though the technique is more challenging in DS proton therapy, it is feasible. "We showed that a global range shift of a few millimetres is detectable and that iso-energy layers can be resolved," he said. Details of the clinical treatment are being published in the Green Journal (Radiother. Oncol. doi: 10.1016/j.radonc.2016.01.004).

The Dresden site will get PBS capability in the next couple of months, and the researchers will then continue the clinical study using both DS and PBS. They also plan to perform phantom studies comparing the slit camera capability for the two delivery modes.

Related stories

• Ranging scheme may improve proton therapy
• Resolved gamma lines verify proton range
• Can prompt gammas monitor in real time?
• Proton therapy: how imaging can help