Analysis of accelerator based neutron spectra for BNCT using proton recoil spectroscopy

Boron Neutron Capture Therapy (BNCT) is a promising binary treatment modality for high-grade primary brain tumors (glioblastoma multiforme, GM) and other cancers. BNCT employs a boron-10 containing compound that preferentially accumulates in the cancer cells in the brain. Upon neutron capture by {sup 10}B energetic alpha particles and triton released at the absorption site kill the cancer cell. In order to gain penetration depth in the brain Fairchild proposed, for this purpose, the use of energetic epithermal neutrons at about 10 keV. Phase 1/2 clinical trials of BNCT for GM are underway at the Brookhaven Medical Research Reactor (BMRR) and at the MIT Reactor, using these nuclear reactors as the source for epithermal neutrons. In light of the limitations of new reactor installations, e.g. cost, safety and licensing, and limited capability for modulating the reactor based neutron beam energy spectra, alternative neutron sources are being contemplated for wider implementation of this modality in a hospital environment. For example, accelerator based neutron sources offer the possibility of tailoring the neutron beams, in terms of improved depth-dose distributions, to the individual and offer, with relative ease, the capability of modifying the neutron beam energy and port size. In previous work new concepts for compact accelerator/target configuration were published. In this work, using the Van de Graaff accelerator the authors have explored different materials for filtering and reflecting neutron beams produced by irradiating a thick Li target with 1.8 to 2.5 MeV proton beams. However, since the yield and the maximum neutron energy emerging from the Li-7(p,n)Be-7 reaction increase with increase in the proton beam energy, there is a need for optimization of the proton energy versus filter and shielding requirements to obtain the desired epithermal neutron beam. The MCNP-4A computer code was used for the initial design studies that were verified with benchmark experiments using a proton recoil spectroscopy detection system. Comparison was also made between in phantom {sup 10}BF{sub 3} readings made at the BMRR and those made at the RARAF accelerator facility