The potential centres must submit their applications by the end of this month and the final decisions on awarding permits will be made at the end of October. At this point, successful teams then have six months to submit detailed business plans, including details of funding sources and enabling clinical treatments to start within the next four years.

Four previously established proton therapy initiatives are already lined up to apply for the four permits: the University Medical Center Groningen; Maastro Clinic and Maastricht University Medical Center in Maastricht; VU University Medical Center, AMC and the Antoni van Leeuwenhoek hospital in Amsterdam; and Erasmus MC, MC Leiden and TU Delft.

According to Erik Roelofs, medical physicist and program manager at the Maastro Clinic, all four permits are expected to be granted. Currently, Dutch patients who may benefit from proton therapy are referred to remote centres such as MGH in Boston, WPE in Essen, Germany or PSI in Switzerland, but this is not that common an occurrence, due to distances and waiting lists. "So for those indications that are considered proven, and those where proton therapy is very likely to be beneficial to patients (head-and-neck, breast, lung and prostate tumours), we'll be able to treat them in the Netherlands," he told medicalphysicsweb.

Proven benefits
Roelofs explained that the Netherlands takes a somewhat cautious approach when it comes to introducing new treatment regimes, as evidenced by its health insurance board, which will only reimburse proven technologies. For many years, there was concern about the lack of evidence for proton therapy, a situation that led to the instigation of ROCOCO - an ongoing international multicentric in silico trial comparing photons, protons and carbon ions.

The idea is that before a patient is referred for proton therapy, simulated plan comparisons must reveal a significant dosimetric benefit of protons. This advantageous dose distribution then needs to be translated into clinical benefits, such as reduction of side effects, using proven complication prediction models. "This general in silico concept was picked up by the health insurance board to be enough evidence to reimburse," Roelofs said. The board has stated four model-based indications: head-and-neck, lung, prostate and breast cancer, where patients are eligible for reimbursement, as well as the standard indications: intraocular tumours, chordoma/chondrosarcoma and paediatric tumours.

Another government requirement is that all patients treated with protons in the new centres are included in standardized clinical trials with uniform outcome measurement, requiring close collaboration between sites. "All four initiatives have been working on this proposal for years, so there's already good collaboration between them," Roelofs explained. "We're working together to set up one registry for protocols. When you have four centres treating in the same way, collecting the data in the same way, this will enable you to perform comparisons more easily. I think it will help create the evidence for proton therapy that is still needed."

One may question whether a relatively small country such as the Netherlands actually needs four proton therapy centres. But Roelofs is convinced of the advantages of having four small sites as opposed to one large centre. "We believe that smaller is better," he said. "You do see larger institutions around the world having a hard time keeping their head up. In Maastricht, we will start with a single gantry and when we can treat more patients, we will buy another one that will be the latest technology. A lot will happen in the coming years, with various vendors competing with technology improvements and systems getting smaller and cheaper."

A study from the Dutch Health Care Insurance Board has predicted that about 3450 patients will be eligible for proton therapy per year. To start off slowly, the government has set a maximum of 2200 patients to be treated, with 400 allocated to the Maastricht site (which covers the smallest region) and 600 to the other three. Initially, the idea was for every centre to be able to treat every indication. However, recently this was modified such that treatments of children and rare intraocular tumours will be centralized to a maximum of two centres.

Patient selection
One key issue when establishing a proton therapy programme is how to select suitable patients. "Proton therapy is not for every patient, as was recently shown in a cost-effectiveness study based on the ROCOCO head-and-neck dataset. You need a tool to pick out the right ones," said Roelofs. As such, the Maastro Clinic team has developed an online model to help with this task, available on the website predictcancer.org. "You feed in the patient parameters and the model will predict whether IMRT or IMPT is more beneficial," he explained.

One of Roelofs' roles in the Maastro Clinic proposal involves setting up a framework to collaboratively collect and share treatment data. Furthermore, he is developing a decision support system called PRODECIS, which offers a fast and reliable service for doctors that consider patient referral. "I'm currently working on combining the dosimetric comparisons from the ROCOCO trials with the online prediction models, to very quickly come up with an answer as to whether the patients are eligible for proton therapy treatment," he said.

While the system vendor has not yet been chosen, Roelofs says that the centre will definitely install a pencil-beam scanning proton gantry. "We believe that this is the state-of-the-art for compact systems for the coming years," he said. And if all goes to plan, the Maastricht proton therapy centre should begin treating patients in 2017.

Related stories

The challenges of paediatric protons
A pioneering approach to proton therapy
Future prospects for proton therapy
Framework eases multicentric in silico planning studies
RILT: what are the real risks?