Optimal detection of the neutron capture therapy agent borocaptate sodium (BSH): a comparison between 1H and 10B NMR

Boron Neutron Capture Therapy (BNCT), an experimental binary cancer treatment modality, requires selective targeting of 10B containing compounds to tumors. One of the compounds under evaluation in an EORTC phase I trial, and used in Japan for patient treatments for many years, is borocaptate sodium (BSH, also known as sulfhydril boron hydride). To optimize the clinical applications, a noninvasive method is needed to monitor the distribution of the boron compound, and NMR may offer such a possibility. A comparison between the relative sensitivities for detecting BSH by 10B or 1H NMR was conducted at two magnetic field strengths: 2 and 4.7 T. At each field strength, similar-sized radio frequency (rf) coils were used for both nuclei. Theoretical predictions for the intrinsic signal to noise (S/N) advantage of 1H over 10B detection vary between a factor of 5.4 and a factor of 28.9, depending on whether the effective resistance is dominated by coil losses or sample losses. Our tests, conducted on relatively small aqueous samples, which loaded the coils less than expected for animal or human subjects, resulted nevertheless in advantage factors close to the lower limit of this range. The measured S/N detection advantage factors for 1H were about 5.2 at 4.7 T, using a dedicated 1H coil, and 7.7 at 2 T, where the measurements were conducted with a double-tuned coil. However, when predicting the expected performance for in vivo MRS or MRI, one should bear in mind that proton detection has to be conducted by spectral-editing pulse sequences with an inherent S/N loss by at least a factor of 2, and that the T1 relaxation time for 10B in BSH is about 30 times shorter than the 1HT1 value. In view of these considerations, direct 10B detection could well be the preferred strategy for MRI/MRS of BSH in vivo.