Rheumatoid arthritis is one of the most common epidemic diseases in the world. For some patients, the treatment with steroids or nonsteroidal antiinflammatory drugs is not effective, thus necessitating physical removal of the inflamed synovium. Alternative approaches other than surgery will provide appropriate disease control and improve the patient's quality of life. In this research, we evaluated the feasibility of conducting boron neutron capture synovectomy (BNCS) with the Tsing Hua open-pool reactor (THOR) as a neutron source. Monte Carlo simulations were performed with arthritic joint models and uncertainties were within 5%. The collimator, reflector and boron concentration were optimized to reduce the treatment time and normal tissue doses. For the knee joint, polyethylene with 40%-enriched Li 2 CO 3 was used as the collimator material, and a rear reflector of 15 cm thick graphite and side reflector of 10 cm thick graphite were chosen. The optimized treatment time was 5.4 min for the parallel-opposed irradiation. For the finger joint, polymethyl methacrylate was used as the reflector material. The treatment time can be reduced to 3.1 min, while skin and bone doses can be effectively reduced by approximately 9% compared with treatment using the graphite reflector. We conclude that using THOR as a treatment modality for BNCS could be a feasible alternative in clinical practice.

Purpose: To evaluate the feasibility of Boron Neutron Capture Synovectomy by Tsing Hua Opening‐pool Reactor (THOR) in Taiwan and to determine the optimal treatment parameters with epithermal neutron beam. Method and Materials: MCNP5 was used to model the THOR epithermal neutron beam interactions with knee and finger phantom. The phantom was established according to the structure of human joints with different boron concentration. The treatment parameters were used to model the optimum treatment assembly, such as different thickness of reflectors and beam orientations. The Figure of merits (FOMs) such as total treatment time, total maximum skin dose and synovium to bone treatment ratio were used to evaluate the effect of the treatment parameters. Results: Monte Carlo calculations predict a total therapy time of BNCS between 5 and 15 min for the human knee by optimum THOR beam assembly. The treatment parameters of BNCS vary with joint sizes. The optimum treatment condition for different joint size can be achieved by using the opposed parallel beam, placing the inflamed joint near the source, and adding 10cm side and rear graphite reflectors. To compare with BNCS using the neutron beam produced by accelerator, the THOR epithermal beam will reduce the total skin dose from 205 RBEcGy to 130.24 RBEcGy and increase the TRbone from 72 to 74.28. Conclusion: This study predicts the optimum THOR beam assembly for BNCS. The result shows the quality and overall clinical efficacy of THOR epithermal neutron beam for BNCS is more suitable than the beam produced by accelerator. It provides the potential application of BNCS by epithermal neutron beam.