Search

The role of endothelium in the inhibition of human factor Xa was studied in a plasma environment. Human factor Xa can bind to and function on bovine aortic endothelium in a manner similar to that of bovine factor Xa. Approximately 70% of the bound factor Xa is subject to inhibition by plasma proteinase inhibitors, and the remaining 30% is irreversibly bound as part of a 125 Kd membrane-associated complex not subject to proteolytic degradation. The proportion reversibly bound and its rate of release do not alter with changes in calcium, citrate, heparin, or active proteinase inhibitor concentrations. The principal plasma proteinase inhibitor of human factor Xa was antithrombin III, which accounted for 60% to 65% of factor Xa released from endothelium, with a,-proteinase inhibitor inactivating 20% to 25% and a2-macroglobulin approximately 15%. All of the reversibly bound factor Xa was identified in complex with one of these three proteinase inhibitors. The thrombin active-site inhibitor hirudin was found to markedly accelerate the displacement of reversibly bound factor Xa from the endothelium and to associate specifically with factor Xa without a loss of activity toward chromogenic substrates, perhaps accounting for a novel mechanism of anticoagulation.

Numerous independent and interdependent factors are involved in the posttransfusion platelet response. Factors such as ABO match and platelet age are related to circumstances potentially under the control of the blood bank physician and therefore may permit circumvention by an active transfusion service. On the other hand, factors such as fever or sepsis may be unavoidable, being related more to the individual patient or clinical condition. To evaluate which factors could be circumvented, we prospectively followed the 1-hour corrected count increments (CCIs) for 962 single-donor apheresis platelet transfusions to 71 refractory hematologic oncology inpatients, with concomitant recording of implicated factors. Stepwise regression analysis allowed for determination of which concurrent and confounding clinical-, patient-, and blood bank-related factors significantly affected the CCIs. Although many implicated factors proved to be independently associated with an increased or decreased CCI, we found that no single variable consistently explained the CCI variation across the patient population. Each patient appeared sensitive to one or a few particular factors, but because of marked intraindividual variation, it was not possible to identify a priori which factors were important for a given patient. The single exception was a solid-phase red blood cell adherence assay used to cross-match platelets, but only for alloimmunized patients. We also evaluated the utility of requesting FILA-matched platelets from the local suppliers and maintained a clear distinction between platelets simply ordered as HLA matched and actually H LA-identical platelets. Accounting for the confounding clinical-, patient-, and blood bank—related factors, the cross-match assay was a better predictor of an adequate CCI than ordering platelets as FILA matched.