When a particle beam interacts with materials in the beamline, or even within the patient, secondary neutrons are created. These neutrons are extremely efficient at ionization and far more likely to cause cell death than particles such as X-rays or protons. Thus if these neutrons reach the patient, the risk of a second cancer increases. How serious a problem this is, and how to address it, was the subject of a panel session at this week's PTCOG 47 meeting in Jacksonville, Florida.

Unfortunately, the exact risk of exposure to secondary neutrons is hard to quantify, as their relative biological effectiveness (RBE) is somewhat of an unknown quantity. "We don't know what the RBE is for neutrons produced in a proton machine," Eric Hall, director of the Center for Radiological Research at Columbia University (New York, NY), told delegates. "But by and large, neutrons are bad news for inducing carcinogenesis."

According to Hall, the major secondary-dose contributions come from neutrons produced in the proton treatment nozzle, with internally-produced neutrons (asides those created in the target itself) negligible in comparison. "It's this last part of the beam that produces most of the neutrons that give the total body dose," he explained.

Harald Paganetti, associate professor of radiation oncology at Massachusetts General Hospital (Boston, MA), concurred: "Most of the risk is due to treatment-head neutrons, not patient neutrons, which is good news because we can't do anything about patient neutrons".

Externally produced neutrons can, however, be reduced. For starters, the material used to make the collimator can make a big difference, as neutron production increases with the atomic number of this material. Collimators are typically made from brass, which is not ideal in this respect. "What we need is a material with a low mass number," explained Hall.

Switching from brass to polyethylene, for example, results in a threefold reduction in secondary neutron production - although the thickness requirements for such a collimator may be a hindrance. Hall cited another option that may do the job better: namely a boron-containing polyethylene, which has demonstrated a significant reduction in neutron dose using a 7.5 inch thick collimator.

One way to avoid the problem of collimator-induced protons altogether is to use active beam scanning. This is, however, a far more complex method, thus most proton treatment systems still employ passive scattering. Having said that, if talk at the PTCOG meeting is anything to go by, it seems likely that beam scanning will become a lot more prevalent over the next year or so.

"Having spent $125 million on a proton facility with the object of sparing normal tissue, does it really make sense to scatter through brass that sprays the body with neutrons when you don’t know what the RBE is?" commented Hall.

Ultimately, the question that needs answering is how the risk of developing second cancer from proton radiotherapy compares to other treatment regimes? According to Paganetti, the actual lifetime risk is, on average, in the same range as that posed by intensity-modulated radiation therapy (IMRT). The neutron dose from proton therapy is, however, widely variable between machines, and he cautions that the uncertainties are larger when estimating neutron doses in proton therapy.

Rounding up the session was Thilo Elsaesser from GSI, the heavy-ion research centre in Darmstadt, Germany. Elsaesser spoke about the issue of secondary neutrons in carbon-ion therapy. He explained that although more neutrons are produced per primary carbon ion than per proton, less carbon ions are needed to achieve the same dose in a patient. Thus the actual dose equivalents turn out to be roughly similar. "Secondary neutrons are not a particular concern in carbon-ion therapy," he concluded.