Tumor-Shed Antigen Affects Antibody Tumor Targeting: Comparison of Two 89Zr-Labeled Antibodies Directed against Shed or Nonshed Antigens

We investigated the effect of shed antigen mesothelin on the tumor uptake of amatuximab, a therapeutic anti-mesothelin mAb clinically tested in mesothelioma patients. The B3 mAb targeting a nonshed antigen was also analyzed for comparison. The mouse model implanted with A431/H9 tumor, which expresses both shed mesothelin and nonshed Lewis-Y antigen, provided an ideal system to compare the biodistribution and PET imaging profles of the two mAbs. Our study demonstrated that the tumor and organ uptakes of 89Zr-B3 were dose-independent when 3 doses, 2, 15, and 60 Ug B3, were compared at 24 h after injection. In contrast, tumor and organ uptakes of 89Zr-amatuximab were dose-dependent, whereby a high dose (60 Ug) was needed to achieve tumor targeting comparable to the low dose (2 Ug) of 89Zr-B3, suggesting that shed mesothelin may afect amatuximab tumor targeting as well as serum half-life. The autoradiography analysis showed that the distribution of 89Zr-B3 was nonuniform with the radioactivity primarily localized at the tumor periphery independent of the B3 dose. However, the autoradiography analysis for 89Zr-amatuximab showed dose-dependent distribution profles of the radiolabel; at 10 Ug dose, the radiolabel penetrated toward the tumor core with its activity comparable to that at the tumor periphery, whereas at 60 Ug dose, the distribution profle became similar to those of 89Zr-B3. These results suggest that shed antigen in blood may act as a decoy requiring higher doses of mAb to improve serum half-life as well as tumor targeting. Systemic mAb concentration should be at a severalfold molar excess to the shed Ag in blood to overcome the hepatic processing of mAb-Ag complexes. On the other hand, mAb concentration should remain lower than the shed Ag concentration in the tumor ECS to maximize tumor penetration by passing binding site barriers.