Although autologous blood transfusion is safe, the procedure requires caution [1, 2]. We describe a 65-year-old man who presented with abrupt-onset hemolysis and acute renal failure after autologous blood transfusion during a total hip replacement for coxarthrosis. One month before presentation, when blood had been taken for autologous transfusion, the transfusion center found a low titer of a cryoagglutinin antibody that was at the limit of significance. No reaction was observed during infusion of autologous blood. Two days after surgery, the patient developed hemolysis and acute renal failure. The patient was referred to our intensive care unit for hemodialysis.

Other causes of renal failure were excluded, and an inquiry ruled out transfusion error and blood conservation accidents. The result of a direct Coombs test was positive, confirming the presence of cryoagglutinin antibodies. The patient had no history of cold agglutinin disease and no acrocyanosis or hemoglobinuria during exposure to cold. Other causes of hemolytic anemia were excluded. The immunologic analysis of the cryoagglutinin antibodies showed a titer of 1000. The thermal amplitude showed activity at a temperature greater than 22 °C. The antibody was agglutinating at temperatures greater than 4 °C and was hemolysing at temperatures greater than 22 °C. The antibody was a monoclonal IgM of complement type with no anti-I or anti-i specificity. No evidence of hematologic disorders was found.

Hemolysis has never been described in autotransfusion except when frozen, inadequately deglycerolized, or infected blood was used [3-5]. In cold agglutinin disease, hemolysis is caused by the fixation of the complement by the cryoagglutinin antibodies. A cryoagglutinin antibody is required for the fixation of the complement to the erythrocytes. When the sites on erythrocytes are already saturated by molecules of C3 and C4, they have no free sites for the binding of the complex cryoagglutinin antibody complement and thus are at minor risk for being lysed. This mechanism explains why, in chronic cold agglutinin disease, erythrocytes resist hemolysis by the cryoagglutinin antibodies.

Our patient had no history of anemia. When blood was originally obtained, results of the direct Coombs test were inconclusive. At that time, erythrocyte surfaces were probably not saturated by complement. Three weeks later, when the patient had surgery, the cryoagglutinin antibody concentration had increased. The transfusion of autologous red blood units, drawn at the time of hypothetical onset of the disease and reinfused without reheating when the cryoagglutinin antibody titer was high, provoked threatening hemolysis and renal failure. The particularly large thermal amplitude of this case of cold agglutinin disease might also have favored hemolysis.

Thus, we believe that the presence of high titers of cryoagglutinin antibodies in patients not known to have any history of chronic disease is a contraindication to transfusion, even of autologous blood units.