Evaluation of the relative biological effectiveness of a clinical epithermal neutron beam using dog brain

This investigation was designed to determine the relative biological effectiveness (RBE) of an epithermal neutron beam (FiR 1 beam) using the brains of dogs. The FiR 1 beam was developed for the treatment of patients with glioma using boron neutron capture therapy. Comparisons were made between the effects of whole-brain irradiation with epithermal neutrons and 6 MV photons. For irradiations with epithermal neutrons, three dose groups were used, 9.4 +/- 0.1, 10.2 +/- 0.1 and 11.5 +/- 0.2 Gy. These physical doses were given as a single exposure and are quoted at the 90% isodose. Four groups of five dogs were irradiated with single doses of 10, 12, 14 or 16 Gy of 6 MV photons to the 100% isodose. Different reference isodoses were used to obtain the most comparable dose distribution in the brain for the two different irradiation modalities. Sequential magnetic resonance images (MRI) were taken for 77-115 weeks after irradiation to detect changes in the brain. Dose-effect relationships were established for changes in the brain as detected either by MRI or by subsequent gross morphology and histology. The doses that caused a specified response in 50% of the animals (ED(50)) were calculated from these dose-effect curves for each end point, and these values were used to calculate the RBE values for the different end points. The RBE values for the FiR 1 beam, based on changes observed on MRI, were in the range 1.2-1.3. For microscopic and gross pathological lesions, the values were in the range 1.2-1.4. The corresponding RBE values for the MRI and pathological end points for the high-LET components (protons from nitrogen capture and recoil protons from fast neutrons) were in the ranges 3.5-4.0 and 3.4-4.4, respectively. This assumed a dose-rate reduction factor of 0.6 for the low-dose-rate gamma-ray component of this beam. Finally, a comparison was made between experimentally derived photon doses, for a specified end point, with calculated photon equivalent doses, which were obtained using the weighting factors for clinical studies on the epithermal neutron beam on the Brookhaven Medical Research Reactor (BNL) in New York. This indicated that the radiation-induced lesions seen in the present study were, on average, detected at a 12% lower photon dose than predicted by the use of the BNL clinical weighting factors. This indicates the need for caution in the extrapolation of results from one reactor-based epithermal neutron beam to another.