Huijun Zhao

Doxycycline is often used as a promoter of inducible gene expression in preclinical models; however, it can also have direct effects on tumor growth and survival. This is due in part to its ability to inhibit cell invasion and regulate matrix metalloproteinase (MMP) expression. Given that doxycycline is also osteotropic, a clinical study to assess its effects on modulation of tumor progression or prevention of skeletal-related events (SRE) in patients with bone metastases from breast cancer (the Achilles trial) was undertaken. Patients received 100 mg of oral doxycycline twice daily for 12 weeks, with serum obtained at baseline and 4, 8 and 12 weeks post-initiation of doxycycline treatment. Exploratory analysis of the effects of doxycycline on circulating levels of MMP or tissue inhibitor of matrix metalloproteinase 2 (TIMP2) was performed in enrolled patients. Statistically significant associations were observed between MMP2, MMP9 and TIMP2 at baseline with significant associations maintained between absolute levels and changes in levels of MMP2 and TIMP2 at weeks 4–12 post initiation of doxycycline. Treatment with doxycycline generally resulted in decreases in MMP2 and MMP9 levels with concurrent upregulation of TIMP2 at 12 weeks post-initiation of doxycycline treatment. Despite this, we observed no association with the levels of any of these factors with either SRE-free or overall survival in this patient cohort. In summary, despite observing hypothesized effects of doxycycline administration on surrogate markers of its anti-tumor activity, measures of circulating levels of these biomarkers were not prognostic in this patient population.

Background: The microenvironment surrounding cells can exert multiple effects on their biological responses. In particular the extracellular matrix surrounding cells can profoundly influence their behavior. It has been shown that the extracellular matrix composition in tumors is vastly different than that found in normal tissue with increased amounts of certain matrices such as collagen I. It has been previously demonstrated that VEGF stimulation of endothelial cells growing on type I collagen results in the induction of bcl-2 expression and enhanced endothelial cell survival. We sought to investigate whether this increased endothelial cell survival resulted in the failure of angiostatic molecules to inhibit angiogenesis. Results: We now demonstrate that VEGF-induced survival on collagen I impairs the ability of three known angiostatic molecules, TSP-1, IP-10 and endostatin to inhibit endothelial cell proliferation. Apoptosis of endothelial cells, growing on collagen I, induced by TSP-1 and IP-10 was also inhibited following VEGF stimulation. In contrast, endostatin induced apoptosis in these same cells. Further analysis determined that endostatin did not decrease the expression of bcl-2 nor did it increase activation of caspase-3 in the presence of VEGF. Alternatively, it appeared that in the presence of VEGF, endostatin induced the activation of caspase-8 in endothelial cells grown on collagen I. Furthermore, only endostatin had the ability to inhibit VEGF-induced sprout formation in collagen I gels. Conclusion: These data suggest that TSP-1, IP-10 and endostatin inhibit endothelial cells via different mechanisms and that only endostatin is effective in inhibiting angiogenic activities in the presence of collagen I. Our results suggest that the efficacy of angiostatic treatments may be impaired depending on the context of the extracellular matrix within the tumor environment and thus could impede the efficacy of angiostatic therapies.