Abstract

We investigated the hypoxia-dependent cytotoxicity of AQ4N (banoxantrone) using a panel of 13 cancer cell lines and studied its relationship to the expression of the quinone reductase DT-diaphorase (NQO1), which is widely found in cancer cells. We also investigated pharmacologic treatments that increase tumor hypoxia in vivo and their impact on AQ4N chemosensitivity in a solid tumor xenograft model. AQ4N showed ≥8-fold higher cytotoxicity under hypoxia than normoxia in cultures of 9L rat gliosarcoma and H460 human non-small-cell lung carcinoma cells but not for 11 other human cancer cell lines. DT-diaphorase protein levels and AQ4N chemosensitivity were poorly correlated across the cancer cell line panel, and AQ4N chemosensitivity was not affected by DT-diaphorase inhibitors. The vasodilator hydralazine decreased tumor perfusion and increased tumor hypoxia in 9L tumor xenografts, and to a lesser extent in H460 tumor xenografts. However, hydralazine did not increase AQ4N-dependent antitumor activity. Combination of AQ4N with the angiogenesis inhibitor axitinib, which increases 9L tumor hypoxia, transiently increased antitumor activity but with an increase in host toxicity. These findings indicate that the capacity to bioactivate AQ4N is not dependent on DT-diaphorase and is not widespread in cultured cancer cell lines. Moreover, the activation of AQ4N cytotoxicity in vivo requires tumor hypoxia that is more extensive or prolonged than can readily be achieved by vasodilation or by antiangiogenic drug treatment.