Characterisation of an accelerator-based neutron source for BNCT versus beam energy

Neutron capture in 10B produces energetic alpha particles that have a high linear energy transfer in tissue. This results in higher cell killing and a higher relative biological effectiveness compared to photons. Using suitably designed boron compounds which preferentially localize in cancerous cells instead of healthy tissues, boron neutron capture therapy (BNCT) has the potential of providing a higher tumor cure rate within minimal toxicity to normal tissues. This clinical approach requires a thermal neutron source, generally a nuclear reactor, with a fluence rate sufficient to deliver tumorcidal doses within a reasonable treatment time (minutes). Thermal neutrons do not penetrate deeply in tissue, therefore BNCT is limited to lesions which are either superficial or otherwise accessible. In this work, we investigate the feasibility of an accelerator-based thermal neutron source for the BNCT of skin melanomas. The source was designed via MCNP Monte Carlo simulations of the thermalization of a fast neutron beam, generated by 7 MeV deuterons impinging on a thick target of beryllium. The neutron field was characterized at several deuteron energies (3.0–6.5 MeV) in an experimental structure installed at the Van De Graaff accelerator of the Laboratori Nazionali di Legnaro, in Italy. Thermal and epithermal neutron fluences were measured with activation techniques and fast neutron spectra were determined with superheated drop detectors (SDD). These neutron spectrometry and dosimetry studies indicated that the fast neutron dose is unacceptably high in the current design. Modifications to the current design to overcome this problem are presented.