Arteriovenous polytetrafluoroethylene (PTFE) grafts used for hemodialysis often fail as the result of myointimal hyperplasia with vascular smooth-muscle-cell (SMC) proliferation. The stenotic lesions occur primarily at the graft-vein anastomosis and less frequently at the graft-artery anastomosis. To explore the potentials of pharmacologic agents in preventing hemodialysis-graft stenosis, we first examined the susceptibility of venous and aortic SMCs to 3 antiproliferative drugs. Human aortic and saphenous-vein SMCs were cultured in a medium containing insulin, epidermal growth factor, fibroblast growth factor, and fetal bovine serum. Various concentrations of dipyridamole (0-100 microg/mL), paclitaxel (0-100 microg/mL), and tranilast (0-300 microg/mL) were added. After 72 hours, we subjected the cells to a mitochondrial enzymatic (methylthiazoletetrazolium; MTT) assay and a bromodeoxyuridine (BrdU)-incorporation assay as a means of assessing their proliferation. Dipyridamole, paclitaxel, and tranilast each inhibited the proliferation of aortic and venous SMCs in a dose-dependent manner ( P <.0001). Approximately 90% inhibition was achieved at dipyridamole concentrations of 75 microg/mL and greater in both MTT and BrdU assays; paclitaxel and tranilast were less effective. The venous SMCs were substantially more susceptible to inhibition by all 3 drugs than were the aortic SMCs in the MTT assay. The concentrations required to produce 50% inhibition (IC 50 ) in the venous cells were 5.8 microg/mL (11.5 micromol/L), 9.1 microg/mL (10.7 micromol/L), and 37.4 microg/mL (114.3 micromol/L), respectively, for dipyridamole, paclitaxel, and tranilast. These concentrations were approximately 4.2, 5.3, and 3.0 times lower, respectively, than the corresponding IC 50 values for the aortic cells. The differences in IC 50 between the aortic and venous cells for the 3 drugs were less pronounced in the BrdU assay. The results of this study suggest that strategies for the prevention of stenosis should take into account the fact that lesions at venous anastomoses of arteriovenous grafts may respond differently to drugs than do those at arterial anastomoses.