The purpose of this study was to test the hypothesis that combined thromboxane A2 synthetase inhibition and receptor blockade is superior to either action alone in preventing cyclic flow variations in stenosed and endothelially injured canine coronary arteries. Forty-five dogs developed coronary cyclic flow variations after a plastic constrictor was placed around the left anterior descending coronary artery at the site where the endothelium was injured and received different interventions. In Group I, 17 dogs were treated with SQ 29,548, a thromboxane A2-prostaglandin H2 receptor antagonist. In Group II, 11 dogs received dazoxiben, a thromboxane A2 synthetase inhibitor. In Group III, R 68,070, a dual thromboxane A2 synthetase inhibitor and thromboxane A2-prostaglandin H2 receptor antagonist, was administered to 11 dogs. Group IV comprised six dogs that received aspirin before receiving R 68,070. Complete abolition of cyclic flow variations was achieved in 71% of dogs in Group I, 82% in Group II, 100% in Group III (p = 0.06 compared with Group I) and 50% in Group IV (p = 0.03 compared with Group III). Epinephrine was infused into dogs with abolished cyclic flow variations: all dogs in Group I had cyclic flow variations restored, 44% in Group II (p = 0.01 compared with Group I) and 64% in Group III (p = 0.04 compared with Group I). The plasma epinephrine levels required to restore cyclic flow variations were 2.2 +/- 0.5 ng/ml (control 0.04 +/- 0.01) in Group I, 8.7 +/- 4.5 ng/ml (control 0.05 +/- 0.02) in Group II and 7.4 +/- 2.6 ng/ml (control 0.07 +/- 0.02) in Group III.(ABSTRACT TRUNCATED AT 250 WORDS)