Prediction of tumor response to experimental radioimmunotherapy with 90Y in nude mice.

Purpose: To identify those factors that predict variability in tumor response to 90Y-radioimmunotherapy based on measurement of incorporated activity and physical dimensions of individual tumors and to apply the concept of effective dose to radioimmunotherapy.

Methods And Materials: Human colon carcinoma xenografts growing in nude mice were treated with anti-CEA antibodies labeled with 90Y directly or through a bispecific antibody/labeled hapten system. Tumor response was measured as the delay in growth to eight times the treatment volume. Noninvasive activity (based on bremsstrahlung radiation) and dimension measurements were made in these animals at several times after label injection. The following parameters were compared for their ability to predict individual tumor response: (a) injected activity, (b) injected activity times a factor based on average uptake as a function of volume, (c) in vivo activity per volume measured in each animal at a single time, (d) the integral over time of in vivo activity per volume in each animal, and (e) the minimum dose for each animal in a uniformly active ellipsoid whose total activity and dimensions varied over time the same as the tumor.

Results And Conclusion: After correcting for differences in injected activity, two parameters account for much of the variability in tumor response. One of these is the general trend of larger tumors to take up less activity per volume. Additional variability can be accounted for by the in vivo activity per volume measurements. The minimum dose as introduced here is likely to be useful in estimating the biologically effective dose delivered by each treatment.