Hepatic radiation dose during I-131 MIBG therapy for refractory neuroblastoma: Tracer prediction and therapy-imaging estimation

1818 Objectives: The use of tracer measurements to predict the hepatic absorbed radiation dose during I-131 metaiodobenzylguanidine(MIBG) therapy of neuroblastoma may guard against resultant hepatic toxicity. We compared the intra-therapy radiation dose predicted from tracer imaging to that with estimation aided by intra-therapy imaging. Methods: Planar, conjugate-view images of the liver were obtained for 7 patients at 0.5, 2, 4, 24, 72, 144 and 188hrs post tracer infusion of 2mCi of I-131 MIBG, and after therapy infusion of 249 to 764mCi (mean 458mCi), at 72, 120 and 168hrs. At each time, a reference source of known I-131 activity was placed on the skin to allow a proportionality between count and activity from reference and liver (a method immune to global camera deadtime). Liver weight was obtained from a fit to published average values versus patient weight, and then an S value from a fit to calculated values versus liver weight. Tracer data were well fit by a sum of two or three decaying exponentials, the cumulated activity was calculated from the fit, and the tracer dose estimated. Scaling that dose 1) by the therapy/tracer administered-activity ratio yielded the predicted therapy dose, and 2) by the average ratio of therapy-measured over tracer-fit liver activity, at the three post-therapy imaging times, yielded the therapy-imaging estimated dose. Results: The tracer-predicted hepatic radiation dose during therapy was linearly dependent on the administered activity(r=0.73). The mean hepatic radiation dose predicted from tracer-imaging was 1517±142cGy(mean±SEM) compared to 1220±203cGy with therapy-imaging estimation. The tracer prediction was higher than the post-therapy-imaging estimate in 5 of the 7 patients. Clinically, no hepatic toxicity was observed in any patient. The therapy-imaging liver radiation dose correlated less well with administered activity(r=0.23). The correlation between the two radiation-dose estimates, tracer-predicted versus therapy-imaging, had an r value of 0.18 (the slope was 0.13). All correlations were considerably stronger for the subset of 5 patients whose weight fell between 20 and 32kg (administered activity between 249 and 570mCi). Conclusions: The hepatic radiation dose during therapy that was predicted by scaling pre-therapy radiation dose by the ratio of administered activities was higher than that estimated using post-therapy activity measurements. Correlations may have been weakened because of the inability to measure individual-patient hepatic mass, and liver activity at time points earlier than three days post therapy.