Quantitative imaging and microlocalization of boron-10 in brain tumors and infiltrating tumor cells by SIMS ion microscopy: relevance to neutron capture therapy.

Boron neutron capture therapy (BNCT) is dependent on the selective accumulation of boron-10 in tumor cells relative to the contiguous normal cells. Ion microscopy was used to evaluate the microdistribution of boron-10 from p-boronophenylalanine (BPA) in the 9L rat gliosarcoma and the F98 rat glioma brain tumor models. Four routes of BPA administration were used: i.p. injection, intracarotid (i.c.) injection [with and without blood-brain barrier disruption (BBB-D)], and continuous timed i.v. infusions. i.p. injection of BPA in the 9L gliosarcoma resulted in a tumor-to-brain (T:Br) boron-10 concentration ratio of 3.7:1 when measured at the tumor-normal brain interface. In the F98 glioma, i.c injection of BPA resulted in a T:Br ratio of 2.9:1, and this increased to 5.4:1 when BBB-D was performed. The increased tumor boron uptake would potentially enhance the therapeutic ratio of BNCT by >25%. At present, ion microscopy is the only technique to provide a direct measurement of the T:Br boron-10 concentration ratio for tumor cells infiltrating normal brain. In the 9L gliosarcoma, this ratio was 2.9:1 after i.p. administration. In the F98 glioma, i.c injection resulted in a ratio of 2.2:1, and this increased to 3.0:1 after BBB-D. Ion microscopy revealed a consistent pattern of boron-10 microdistribution for both rat brain tumor models. The boron-10 concentration in the main tumor mass (MTM) was approximately twice that of the infiltrating tumor cells. One hour after a 2-h i.v. infusion of BPA in rats with the 9L gliosarcoma, tumor boron-10 concentrations were 2.7 times higher than that of infiltrating tumor cells [83 +/- 23 microg/g tissue versus 31 +/- 12 microg/g tissue (mean +/- SD)]. Continuous 3- and 6-h i.v. infusions of BPA in the 9L gliosarcoma resulted in similar high boron-10 concentrations in the MTM. The boron-10 concentration in infiltrating tumor cells was two times lower than the MTM after a 3-h infusion. After 6 h, the boron-10 concentration in infiltrating tumor cells had increased nearly 90% relative to the 2- and 3-h infusions. A 24-h i.v. infusion resulted in similar boron-10 levels between the MTM and the infiltrating tumor cells. Boron concentrations in the normal brain were similar for all four infusion times (approximately 20 microg/g tissue). These results are important for BNCT, because clinical protocols using a 2-h infusion have been performed with the assumption that infiltrating tumor cells contain equivalent amounts of boron-10 as the MTM. The results reported here suggest that this is not the case and that a 6-h or longer infusion of BPA may be necessary to raise boron-10 levels in infiltrating tumor cells to that in the MTM.