DeLaney began by summarizing the clinical advantages of protons: they offer a superior dose distribution compared to photons and, because less dose is delivered to normal tissue, result in lower toxicity. "From that perspective, particles are very appealing," he said. "In particular, larger lesions benefit most from treatment with protons and charged particles, because they have a larger volume of tissue that's spared exit dose".

As proton-therapy technology continues to advance, features such as rotational gantries, higher energies and larger fields have expanded the treatment's application to potentially any site in the body. However, this larger remit has also resulted in the emergence of particular problems caused by target motion and tissue heterogeneity.

Proton therapy also brings other, more practical, challenges for the clinician to deal with. Proton-beam delivery systems are generally less reliable than linacs, have more downtime, no back-up facility if they fail, and require longer maintenance cycles.

And then there's that other big sticking point: cost. "If the cost of proton therapy was the same as photon treatment, a lot of the debate about where to use photons would really be a moot point," DeLaney stated. The cost of setting up a facility to perform intensity-modulated proton therapy (IMPT), for example, has been calculated at about 2.5 times that of establishing an intensity-modulated radiation therapy system.

But it's not always just about the set-up cost. DeLaney cited one example in which the costs of treating medulloblastoma in a five-year old child were calculated. "If you can reduce the risk of late radiation-associated complications over the life of the child, you will reduce the overall health-related cost issues," he explained.

In addition, DeLaney reckons that, with time, "some combination of technical and efficiency improvements reduces the cost of most modalities." Development of a lower-cost cyclotron could also deal with downtime issues, by enabling the possibility of a back-up system being deployed. Where appropriate, hypofractionated therapy can also reduce treatment costs, although DeLaney noted that it's not practical for all cases.

Any answers?
Some of these challenges were addressed in the follow-up presentation, given by Anthony Lomax of the Paul Scherrer Institut (PSI) in Switzerland. Lomax gave PTCOG delegates the "physicist's perspective" on advances and challenges in proton therapy.

He began by explaining that proton therapy is currently predominantly delivered via passive scattering. Switching to spot scanning - in which a narrow pencil beam is magnetically deflected to a series of positions throughout the target - could, however, vastly improve the sparing to proximal normal tissue.

The use of spot-scanning-based proton therapy is still relatively rare. It's currently only employed by PSI itself, very recently by MD Anderson Cancer Center (Houston, TX) and, for carbon ions, by GSI in Darmstadt, Germany. "This technique uses protons more efficiently than passive scattering, it really only delivers the Bragg peaks in the target itself," Lomax explained.

Following on from a shift from passive scattering to spot scanning, the next key advance may well be IMPT: the simultaneous optimization of all the Bragg peaks from all of the incident beams. This would increase the dose conformality even further, said Lomax, citing IMPT as the "third stage" in the progression of proton therapy.

Lomax admitted that many challenges still need to be addressed. "The advantage of protons is that they stop," he declared. "The disadvantage is that we don't know where they stop." And this so-called range uncertainty - caused by issues including tumour shrinkage, patients changing weight and CT artefacts in scans of patients with metal implants - can result in big treatment inaccuracies.

Range uncertainties need to be tackled by improvement in tumour imaging. "Imaging has had such a big impact on conventional radiotherapy, and there's going to be a big role for imaging in proton therapy," said Lomax. Advanced techniques such as dual-energy imaging could help reduce CT artefacts, while PET has been used to provide a post-therapy estimate of where the dose was delivered.

And, of course, there's the other big obstacle faced by any highly-conformal treatment - organ motion, which poses a particular challenge for spot scanning. "Depending on when I apply the scanned beam in relation to the motion, I may or may not hit the target," Lomax noted.

One solution could be rescanning: repainting the beam many times so that statistics dictate the coverage and homogeneity of the dose to the target. He did, however, caution to the "synchronicity" effect, in which the period of the scan is in phase with the breathing period. Tumour tracking is another possibility. Ultimately, a combination of rescanning and tracking - tumour retracking - could provide the optimal solution.

With this in mind, researchers at PSI are working to develop a new proton therapy gantry that will enable this type of ultrafast scanning. They hope to treat the first patient with this in the summer of next year.