Therapeutic Efficacy of Antiglioma Mesenchymal Extracellular Matrix ¹³¹I-Radiolabeled Murine Monoclonal Antibody in a Human Glioma Xenograft Model¹

The development of Mabs, particularly those reactive with primary brain tumors but not with normal brain, provides a potential means of delivering therapeutic agents selectively to human malignant gliomas. Mab 81C6, an IgG2b immunoglobulin, which defines an epitope of the glioma-associated extracellular matrix protein tenascin, has been shown to bind to human glioma cell lines, glioma xenografts in nude mice, and primary human gliomas, but not to normal adult or fetal brain. To test the therapeutic potential of this Mab for targeted delivery of isotopes, nude mice bearing progressively growing s.c. xenografts of D-54 MG, a human glioma cell line, were given injections via the tail vein of either buffer, unlabeled 81C6, ¹³¹I-labeled 81C6, or ¹³¹I-labeled 45.6, a nonspecific control Mab of the same isotype. Specific activities of the Mab range from 6.0 to 15.5 mCi/mg with protein doses from 7.6 to 167 µg. The doses given by injection per animal for labeled 81C6 were 50, 250, 500, and 1000 µCi and 500 and 1000 µCi for 45.6. Tumor response was measured by growth delay in reaching 1000 or 5000 mm³ tumor volumes using the Wilcoxon rank sum test, and by comparing the proportion of tumors that had regression in volume after treatment using the Fisher exact test. Statistically significant growth delays at 1000 mm³ were noted in 1 of 3 experiments with 500 µCi 81C6 (P < 0.001) and 2 of 3 for 1000 µCi 81C6 (P = 0.001 and <0.001). At 5000 mm³, statistically significant growth delays were seen with radiolabeled 81C6 in 2 of 2 experiments at 250 µCi (P = 0.01 and 0.02), 4 of 4 at 500 µCi (P = 0.03-<0.001), and 2 of 2 at 1000 µCi (P = ≤ 0.001) and with radiolabeled 45.6 in 1 of 1 at 1000 µCi (P = 0.01). The percentage of animals with tumor regression progressively increased with increasing doses of isotope. For radiolabeled 45.6, there were 0 of 10 regressors at 500 and 1 of 10 at 1000 µCi. For radiolabeled 81C6, there were 0 of 6 regressors at 50 µCi, 1 of 16 (6%) at 250 µCi, 7 of 38 (18%) at 500, and 15 of 28 (54%) at 1000 µCi. Statistically significant tumor regression was seen only at doses of 500 and 1000 µCi of ¹³¹I-81C6. The initial tumor size for those regressing was significantly smaller than those not regressing (P = 0.01 for 500 µCi and 0.0009 for 1000 µCi). The estimated dose to tumor was 9719 cGy for 1000 µCi 81C6 and 2346 cGy for 1000 µCi 45.6. Doses to other organs for 81C6 and 45.6 were equivalent ranging from 135 cGy for brain to 2415 cGy for lung. Whole body dose determined by total body measurement with dose calibrator and direct individual tissue counting with a gamma counter were equivalent. Comparative dosimetry calculations were made based upon data extrapolated from prior trace-labeled localization studies (5 µCi/5 µg/animal). The estimated radiation dose to tumor from these studies in which no therapeutic tumor response was seen underestimated the dose observed in a directly measured therapeutic trial by 35–52%. In this xenograft model, a radiolabeled antiglioma Mab against the extracellular matrix protein tenascin demonstrated therapeutic efficacy. The promising results obtained in this animal model suggest a potential value for this form of therapy against human malignant gliomas.