To increase the quality of dose delivery, Schippers recommended that "in future, everyone should move to pencil-beam scanning". Currently, this technique – in which a narrow pencil beam is magnetically deflected to paint the dose throughout the target – is only used by PSI and a small number of other sites. Meanwhile, every other proton therapy centre in the world still employs passive scattering.

Pencil-beam scanning is more efficient than beam delivery via scattering, and also offers the highest flexibility for shaping the dose distribution. However, it is inherently more sensitive to organ motion during treatment. Interplay effects between the motion of the target and the proton beam can lead to hot and cold spots in the target volume. So how can the accuracy of proton dose delivery be improved?

Schippers detailed several potential approaches: gating, in which the beam is only applied at certain points in the breathing cycle; adaptive scanning, in which the pencil beam is moved to track the organ motion (although this is only in the research stage and not used yet); and fast rescanning, where the target volume is painted multiple times to average out motion effects. This latter approach requires high scanning speeds, which can be achieved by PSI's state-of-the-art Gantry 2. PSI is also investigating fast 3D scanning, in which the beam intensity is also rapidly modulated during the beam sweep.

One other option for increasing the beam delivery accuracy is integrating MRI guidance with proton therapy – such as is being developed for photon-based treatments. "I think that this is one of the things that we should go for in the future," Schippers said.


The third item in Schippers' wish-list was high flexibility - both in the treatment dynamics and in the equipment itself, which represents a huge investment that must be future-proof and upgradeable. "A cyclotron is the ideal accelerator for maximum flexibility," he told delegates, citing benefits including a continuous beam, high reliability and rapidly adjustable beam intensity. The 250 MeV cyclotron at PSI, for example, can modify beam intensity with 3% accuracy in just 50 μs.

The disadvantage of the cyclotron is that, in contrast to a synchrotron, it produces a beam at a single energy. Altering the beam energy requires external regulation by a degrader. The PSI system can be adjusted between 238–70 MeV, with 1% field changes (or 5 mm change in penetration depth) in 50–80 ms. Increasing the beam intensity, meanwhile, to 1–1.5 µA, would enable splitting of the beam between multiple treatment rooms. This would allow more than one gantry to be used at one time, greatly increasing patient throughput.

Schippers went on to discuss the issue of higher proton energy, and why one would actually need this. One key application is proton radiography, as protons with an energy of 350 MeV will travel straight through the patient. "The best way to measure the range of protons in a patient is by measuring the energy loss of protons in a patient," he explained.

Increasing the proton energy will also sharpen the edge of the dose distribution, as the beam spreads less, which could prove beneficial in the treatment of very small lesions. PSI is currently working to develop such a high-energy system, by adding a linac based on a design of the TERA Foundation (Italy) to the existing beam transport system in order to boost proton energy from 250 to 350 MeV (the ImPulse project).

The one low

Finally, Schippers took a look at proton therapy's inescapable need to lower costs. Ultimately, this will be achieved via the development of smaller accelerators that can fit into a single treatment room.

The size of a cyclotron can be reduced by increasing the magnetic field. However, at very strong fields, the field weakens towards the cyclotron's outside edge. To mitigate this effect, synchrocyclotron systems in which the frequency of the driving field is adapted with radius are being investigated. This arrangement is exploited in Mevion's S250 system and IBA's Proteus ONE, both of whom announced first installations of their systems towards the end of last year. Around this time, installation also commenced of ProTom's Radiance 330, a compact synchrotron system.

Looking further ahead, there's the dielectric-wall accelerator, which could be small enough to be mounted on a rotating gantry. And at the very end of Schippers' usability time scale of "now, up until mañana", sits the fixed-field alternating gradient accelerator (FFAG), the laser-driven accelerator and the plasma wakefield accelerator.

Schippers ended his presentation with a note of caution. "Smaller is better; but can we achieve the same quality as we can with the current bigger system?" he asked. "I'm not saying don't do it, but just be very careful."