In a parallel study, Browder et al.(9) grew cyclophosphamide-resistant tumors (Lewis lung carcinoma and EMT-6 breast carcinoma) in immunocompetent mice to focus more specifically on the effects of a cytotoxic drug on tumor endothelium under different dosing schedule. Cyclophosphamide treatment in a conventional MTD regimen only modestly delayed growth of both tumor types in mice. In contrast, when cyclophosphamide was instead supplied regularly (once every 6 days), tumor growth was significantly impaired, although the tumors eventually prevailed. Provocatively, the relapse of drug-resistant Lewis Lung tumors could be prevented by a combination therapy, this time involving similar metronomic dosing with the experimental angiogenesis inhibitor TNP-470. TNP-470 had previously been shown to impair but not regress subcutaneous tumor growth in mice (10) and to enhance high-dose episodic chemotherapy (11). In combination, cyclophosphamide and TNP-470 eradicated aggressive drug-resistant tumors in 32/38 tumor-bearing mice. Other trials assessed drug-sensitive Lewis Lung carcinomas and L1210 leukemias, both of which could be regressed without relapse using metronomic dosing of cyclophosphamide alone. The dose of cyclophosphamide used in this study was relatively high and resulted in significant toxicity; one wonders whether lower, nontoxic levels of cyclophosphamide would have sufficed, in combination with TNP-470, to induce the regression of these tumors. As predicted, metronomic cytotoxic dosing elicited repeated waves of apoptosis of tumor endothelial cells. Using an angiogenesis bioassay in normal mice, Browder et al.(9) confirmed that metronomic dosing of cyclophosphamide, as well as of a number of other cytotoxic drugs (including 5-fluorouracil [5-FU]), was antiangiogenic.

Collectively, these two studies (5, 9) clearly establish that metronomic regimens of cytotoxic drugs can be antiangiogenic, repositioning cytotoxic therapies as bi-or multifunctional against distinct heterotypic cell types in tumors (Figure 1).[The concept of antiangiogenic dosing was originated by Browder et al.(9), as noted by Klement et al.(5).] Both studies further demonstrated the value of combining modified chemotherapeutic regimens with experimental angiogenesis inhibitors. While the data presented are compelling, direct extrapolation to the clinical setting presents several immediate challenges related to the choice of drug, dose, and schedule for maximum antiangiogenic activity. In de-emphasizing the tumor cell as a target, this strategy requires a fundamental change in our approach to therapy, one that potentially includes retreatment of refractory cancers with agents that have previously failed, or the use of agents traditionally deemed inactive or ineffective in a particular cancer type. Second, identification of a MTD by standard toxicity criteria is relatively straightforward, whereas selecting the optimum antiangiogenic dose that is nontoxic yet efficacious may be difficult; surrogate markers of response and/or accurate preclinical models will be important. Novel imaging modalities designed to monitor angiogenesis may prove instrumental in this regard.