To study whether the administration of recombinant human erythropoietin increases the amount of autologous blood that can be collected before surgery, we conducted a randomized, controlled trial of erythropoietin in 47 adults scheduled for elective orthopedic procedures. The patients received either erythropoietin (600 units per kilogram of body weight) or placebo intravenously twice a week for 21 days, during which time up to 6 units of blood was collected. Patients were excluded from donation when their hematocrit values were less than 34 percent. All patients received iron sulfate (325 mg orally three times daily). The mean number of units collected per patient (+/- SE) was 5.4 +/- 0.2 for the erythropoietin group and 4.1 +/- 0.2 for the placebo group. The mean red-cell volume donated by the patients who received erythropoietin was 41 percent greater than that donated by the patients who received placebo (961 vs. 683 ml, P less than 0.05). Only 1 of the 23 patients treated with erythropoietin was unable to donate greater than or equal to 4 units (4 percent) as compared with 7 of the 24 patients who received placebo (29 percent). No adverse effects were attributed to erythropoietin. We conclude that recombinant human erythropoietin increases the ability of patients about to undergo elective surgery to donate autologous blood.

Between April 1987 and May 1989, the Centers for Disease Control investigated seven cases of transfusion-associated Yersinia enterocolitica sepsis; four were caused by organisms of serotype O:3, and one each was caused by organisms of serotype O:1,2,3; O:5,27; and O:20. All seven recipients developed septic shock after receiving units of red cells (RBCs) contaminated with Y. enterocolitica; five recipients died. The cases occurred in seven states and were unrelated. There was no evidence for contamination of the RBC units during processing. Six of the seven donors had serologic evidence of recent Y. enterocolitica infection, and it is hypothesized that these donors had asymptomatic bacteremia when they donated the implicated blood. Four of the seven donors reported gastrointestinal illness in the 4 weeks before blood donation, and one donor became ill on the day he donated blood. Y. enterocolitica grows well at 4 degrees C and in the presence of dextrose and iron. If blood is contaminated at the time of collection, storage of the RBCs at 4 degrees C provides an ideal environment for bacterial growth and endotoxin production. These cases demonstrate the need for careful evaluation of patients with transfusion reactions for possible sepsis and suggest a need to screen prospective blood donors for mild gastrointestinal illness, including those illnesses not requiring physician evaluation or medication.

Serial studies were performed in two patients with cyclic thrombocytopenia to investigate the pathogenesis of this disorder. Mean life span of autologous platelets when platelet levels were declining was subnormal (2.4 and 0.8 days), and megakaryocytes were abundant in the bone marrow during thrombocytopenia. Megakaryocyte colony- stimulating activity could not be detected in the serum of either patient at any point of their cycles. In each patient, total platelet- associated IgG varied inversely with platelet levels. Surface platelet- associated IgG was measured only in patient 2 and was significantly elevated (greater than 1,280 IgG molecules per platelet) at all stages of the cycle, even during thrombocytosis. However, the highest values were observed during thrombocytopenia. Platelet-bindable IgG in plasma declined to normal immediately before platelet levels began to rise. IgG eluted from the platelets of this patient reacted strongly with autologous and homologous platelets in contrast to a “mock eluate” prepared from platelets of a normal subject. The eluate from the patient's platelets reacted strongly with immobilized autologous and homologous glycoprotein IIb/IIIa complex and weakly with GPIb but not with isolated GPIIIa alone. In each patient the decline in platelet levels was significantly delayed following administration of intravenous gamma globulin 0.4 g/kg body weight for five days. These findings suggest that platelet-reactive autoantibodies are of pathogenic significance in some patients with cyclic thrombocytopenia.

Platelet transfusions from donors matched for cross-reactive antigens have been shown to be effective in providing hemostasis in alloimmunized thrombocytopenic patients. A significant number of these transfusions, however, fail to provide posttransfusion platelet recoveries. We investigated incompatibility in the Bw4/bw6 system as a possible explanation for these failures. The Bw4/Bw6 system is a biallelic antigen system closely associated with HLA-B. HLA-B locus antigens that are cross-reactive frequently differ in their Bw4/Bw6 specificity. Posttransfusion platelet recoveries from 21 alloimmunized thrombocytopenic patients homozygous for Bw4 or Bw6 and transfused with both Bw4/Bw6 compatible and incompatible platelets were analyzed. The mean 1-hr posttransfusion recovery was 84% following Bw4/Bw6-compatible platelets versus 52% with Bw4/Bw6-incompatible platelets (p less than 0.02). Twenty-four hours following transfusion, mean recoveries were 44% and 24%, respectively, (p less than 0.01). A subgroup of 8 patients (38%) was identified who had markedly lower responses following Bw4/Bw6- incompatible transfusions as compared to Bw4/Bw6-compatible transfusions (mean recoveries: 1 hr--compatible 100%, incompatible 27%, p less than 0.001; 24 hr--compatible 45%, incompatible 7%, p less than 0.01). These data suggest that the Bw4/Bw6 antigen system has clinical significance for some patients requiring platelet transfusion therapy and, when appropriate, should be considered in the selection of donors.

No abstract