A PRELIMINARY CELL-KINETICS MODEL OF THROMBOCYTOPENIA AFTER RADIOIMMUNOTHERAPY

Thrombocytopenia is often the dose-limiting toxicity for radionuclide therapy. Prediction of platelet counts after therapy is important for treatment planning. Simple prediction methods based on linear correlat ion between radiation dose and blood count nadir have been insufficien t because they have not considered time, because of the complicated hi erarchical structure of the hematopoietic system in which platelets ar e not directly injured by low dose rate radiation and because of chang ing radiation dose rates to marrow with time, This study addresses the se problems using a cell kinetics model. Methods: The model consists o f compartments for progenitor cells, megakaryocytes, platelets and str omal cells. A linear quadratic formula was used for progenitor cell su rvival. Stromal cells were described by a model based on a maximum lik elihood estimate for cellular damage, repair and proliferation. Report ed values for murine cellular turnover rates and radiosensitivity of p rogenitor cells were used in the model calculations. Experimental mice received 4 Gy of external beam radiation for tumor implantation and 1 2.4-23.3 MBq Cu-67-2-iminothiolane-BAT-Lym-1 (BAT = 6-[p-(bromoacetami do) etra-azacyclotetradecane-N,N',N'',N'''-tetraacetic acid) 19-30 day s later. Blood counts were measured three times each week. Results: Th e model predicted the severity of thrombocytopenia, and the time of th e nadir corresponded to measured values in mice. For a dose of 14.2 MB q Cu-67-2-iminothiolane-BAT-Lym-1 that induced a platelet nadir of 20% of baseline (Grade II), the model predicted that at least 20 days wer e needed before a second 14.2-MBq injection if a subsequent nadir of < 10% of baseline (Grade IV) was to be avoided. Conclusion: The nadir an d duration of thrombocytopenia predicted by the model were similar to those observed in the mice. Predicted information could be useful for planning the dose and timing of fractionated radionuclide therapy. Thi s model provides a stepping stone for future development of a predicti ve model for patients.