Bone marrow toxicity is generally dose-limiting for radioimmunotherapy (RIT) with beta-emitting radionuclides. Treatment may be prescribed on the basis of administered activity or absorbed dose. An optimal definition of maximum tolerated dose will enable the clinical benefits of RIT to be maximized.
Methods: We examined data from six clinical studies of RIT with various 131-I labeled antibodies and antibody fragments that treated a total of 114 patients. We also examined a sub-set of 36 patients with minimal prior chemotherapy who were treated with 131I-labeled intact murine IgG at a single institution. For both these groups the ability of absorbed dose-based methods to predict bone marrow tolerance was compared with that of activity-based methods.
Results: Marrow toxicity was more accurately predicted by absorbed dose than by activity in the general case where a variety of different antibodies and antibody fragments were used. For the more homogeneous smaller group, well defined "dose-response" relationships were observed for both absorbed dose and administered activity. However, absorbed dose-based definitions of maximally tolerated dose yielded a better stratification of patients than activity-based definitions (including per meter squared) such that fewer patients had major toxicity when treated below "tolerance", and fewer patients had minor toxicity when treated above "tolerance".
Conclusions: Absorbed dose-based definitions of maximum tolerated dose and escalation variables are optimal for 131I-labeled antibody therapy. The ability of pre-therapy dosimetry studies to predict the behavior of therapeutic administrations must be validated for prospective clinical applications.