In vivo prompt gamma neutron activation analysis for the screening of boron-10 distribution in a rabbit knee: a simulation study.

Boron neutron capture synovectomy (BNCS) is under development as a potential treatment modality for rheumatoid arthritis (RA). RA is characterized by the inflammation of the synovium (the membrane lining articular joints), which leads to pain and a restricted range of motion. BNCS is a two-part procedure involving the injection of a boronated compound directly into the diseased joint followed by irradiation with a low-energy neutron beam. The neutron capture reactions taking place in the synovium deliver a local, high-linear energy transfer (LET) dose aimed at destroying the inflamed synovial membrane. For successful treatment via BNCS, a boron-labeled compound exhibiting both high synovial uptake and long retention time is necessary. Currently, the in vivo uptake behavior of potentially useful boronated compounds is evaluated in the knee joints of rabbits in which arthritis has been induced. This strategy involves the sacrifice and dissection of a large number of animals. An in vivo (10)B screening approach is therefore under investigation with the goal of significantly reducing the number of animals needed for compound evaluation via dissection studies. The 'in vivo prompt gamma neutron activation analysis' (IVPGNAA) approach uses a narrow neutron beam to irradiate the knee from several angular positions following the intra-articular injection of a boronated compound whose uptake characteristics are unknown. A high-purity germanium detector collects the 478 keV gamma photons produced by the (10)B capture reactions. The (10)B distribution in the knee is then reconstructed by solving a system of simultaneous equations using a weighted least squares algorithm. To study the practical feasibility of IVPGNAA, simulation data were generated with the Monte Carlo N-particle transport code. The boron-containing region of a rabbit knee was partitioned into 8 compartments, and the (10)B prompt gamma signals were tallied from 16 angular positions. Results demonstrate that for this level of spatial resolution, an estimate of (10)B distribution inside the joint can be obtained to within 10% uncertainty, under ideal conditions. Variations of the anatomic dimensions among individual rabbit knees and potential knee positioning errors will result in an uncertainty of over 20%. IVPGNAA thus provides sufficient resolution and quantification regarding the in vivo uptake characteristics of boronated pharmaceuticals to serve as a useful means of screening new compounds of potential use in BNCS.