Reduction of blood loss by infusion of human platelets in a rabbit kidney injury model

BACKGROUND: As a first step toward testing the efficacy of stored platelets or platelet substitutes in vivo, a kidney injury model was developed to assess the hemostatic properties of human platelets in normal and thrombocytopenic rabbits. STUDY DESIGN AND METHODS: New Zealand white rabbits were made thrombocytopenic by two consecutive injections of busulfan. Two weeks later, human platelets were transfused to animals whose reticuloendothelial systems were inhibited by the administration of ethyl palmitate. The left kidney was exposed and a slice excised from the anterior pole. The blood was contained in a parafilm boat and absorbed by preweighed gauze to assess blood loss. The percentage of human platelets transfused to the rabbit was determined by flow cytometry on blood collected from the cut site using anti-CD42a (marker for human platelets). The degree of activation of the human platelets was determined using anti-CDB2a (marker specific for human p-selectin). RESULTS: Blood loss was similar in normal animals treated with saline alone (35.4 +/- 5.8 g; n = 4); ethyl palmitate and saline (42.5 +/- 5.7 g; n = 6, p = 0.4); or ethyl palmitate and fresh human platelets (45.7 +/- 7.9 g; n = 6, p = 0.3). Bleeding in thrombocytopenic rabbits infused with saline was increased (75.6 +/- 3.9 g; n = 7) as compared with nonthrombocytopenic animals. A significant reduction in blood loss was noted in thrombocytopenic rabbits given fresh human platelets (51.6 +/- 4.5 g; n = 6, p = 0.0023). Transfusion of human platelets to rabbits did not cause activation of the platelets. Furthermore, transfusion of thrombin-activated platelets (60-98% activated) to thrombocytopenic rabbits reduced blood loss (54 +/- 7.3 g; n = 7) to the same extent as fresh platelets. CONCLUSIONS: This is the first report describing a kidney injury model developed to assess the efficacy of fresh and activated human platelets in reducing blood loss in thrombocytopenic rabbits. This model could monitor the efficacy of human platelets prepared by various preservation protocols in suppressing bleeding in rabbits.