The cytotoxic activity of Adriamycin was examined in the MGH-U1 human bladder carcinoma line, grown as monolayer culture, as spheroids, and as xenografts in immune-deprived mice. The MGH-U1 cells grown as spheroids were much more resistant to Adriamycin (concentration of drug resulting in 37% cell survival, 4.5 µg/ml) than when treated as monolayer cultures (concentration of drug resulting in 37% cell survival, 0.9 µg/ml). Adriamycin fluorescence was demonstrated only in the outer two layers of cells forming the spheroids, suggesting that limited drug penetration is an important factor in the resistance of spheroids to Adriamycin. Sequential trypsinization of spheroids 750 µm in diameter allowed us to determine the cytotoxic effects of Adriamycin in MGH-U1 cells derived from different depths of the spheroid. We found that cells near the surface of the spheroid had a survival similar to those of exponentially growing monolayer cells treated with Adriamycin. Cells located in the middle of the viable rim were more resistant to Adriamycin, and those found near the necrotic center were most resistant to Adriamycin. The effects of Adriamycin treatment on spheroid growth delay were determined also. In spite of a small cytotoxic effect on the clonogenic fraction of cells in MGH-U1 spheroids, the growth delay effect of Adriamycin in intact spheroids was marked. This observation is consistent with Adriamycin killing primarily the cells in the outer layers of the spheroid, where most of the proliferation in the spheroid occurs. In vivo treatment of MGH-U1 xenografts with Adriamycin followed by assessment of cell survival in vitro showed minimal evidence of cytotoxicity, consistent with the poor drug penetration observed in the spheroid model.
These studies suggest that: (a) Adriamycin penetrates poorly into solid tissues; (b) in vitro clonogenic survival following Adriamycin exposure of a cell suspension may predict falsely for drug sensitivity to chemotherapy; (c) a small decrease in clonogenic survival can be translated into a long growth delay but, ultimately, the tumor regrows because some clonogenic cells are spared; and (d) for Adriamycin, the spheroid model more closely parallels the in vivo effects than does monolayer culture. The use of the spheroid model for the study of Adriamycin cytotoxicity gives further insight into the action of this drug in solid tumors.
This work was supported in part by a grant from the National Cancer Institute of Canada.