Therapeutic efficacy of tumor endothelium-targeted RGD-erythrocytes in a B16.F10 murine melanoma model. (A) Light microscopy image of tumor tissue showing thrombi formation in blood vessels (indicated by arrows), 4 hours after intravenous injection of RGD-modified erythrocytes. Tumors measured approximately 300 mm³. Scale bar represents 100 μm. (B) Light microscopy image of a tumor blood vessel showing local thrombus formation (indicated by asterisk) likely due to endothelial cell damage at 2 hours after intravenous injection. Scale bar represents 10 μm. (C) Light microscopy image showing a necrotic tumor core with a typical viable rim at 24 hours after intravenous injection of RGD-modified erythrocytes. Scale bar represents 200 μm. (D) Magnification of viable rim and necrotic tumor area showing a thin layer of cells forming the viable rim. Scale bar represents 50 μm. (E) Individual growth curves (n = 10) of untreated control mice. Mean tumor volume increase during the experiment: 1848 plus or minus 575 mm³ (mean ± SEM). (F) Individual growth curves (n = 8) after treatment with RAD-modified erythrocytes, arrows indicating day of treatment. Mean tumor volume increase during the experiment: 1787 plus or minus 312 mm³ (mean ± SEM). (G) Individual growth curves (n = 8) after treatment with RGD-modified erythrocytes, arrows indicating day of treatment. Mean tumor volume increase during the experiment: 636 plus or minus 149 mm³ (mean ± SEM); P = .005 compared with RAD-modified erythrocytes treated animals (Mann-Whitney test, 2 tailed).