Vasoactive mediators and hypotensive transfusion reactions

Abstract

It is not unusual to receive a report of severe hypotension after a transfusion. The investigation as to cause is frequently unrewarding. Some cases of transfusion-related hypotension can be from volume overload and cardiac failure whereas others can arise from acute hemolysis of mismatched blood or infusion of blood components that are contaminated with bacteria and endotoxins. After excluding these known entities by an analysis of the clinical context and laboratory investigation, further investigative options are limited. Some of the most dramatic hypotensive transfusion reactions have been associated with transfusing incompatible plasma containing immunoglobulin A (IgA) or haptoglobin to patients who are deficient and have antibodies to the corresponding protein they lack.1-8 In these patients their arterial pressure can suddenly become imperceptible and cardiac arrest can ensue or there can be associated symptoms suggesting their allergic nature such as facial, pharyngeal, and laryngeal edematous swelling and bronchospasm with or without chills, fever, or urticaria. Testing for anti-IgA is readily available from reference laboratories but it is not easily available for haptoglobin antibodies. Anaphylactic hypotensive reactions can also follow transfusion of plasma or platelets (PLTs) to patients with anti-Chido or anti-Rodgers red cell (RBC) antibodies, antibodies that cause anaphylactic symptoms rather that RBC hemolysis.9,10 They react with their corresponding soluble antigens associated with the C4d fragment of complement in the transfused plasma and on RBC. In these cases the antibodies can be detected in the RBC antibody screening tests. In other patients with posttransfusion hypotension, pulmonary symptoms can dramatically predominate and when there is noncardiac pulmonary edema an evaluation for transfusion-related acute lung injury is pursued. In many other cases, however, the diagnostic evaluation of acute posttransfusion hypotension is limited because we do not understand the pathophysiologic mechanisms involved. Few diagnostic tests are available for other nonimmune causes and the cause is frequently not found. As a result, recommendations for preventive steps are not made other than to avoid future transfusions unless absolutely necessary. In the past decade, many types of transfusion reactions, including hypotensive reactions, have been attributed to nonimmune mechanisms unrelated to donor or recipient alloantibodies or cellular and plasma protein mismatches. The pathophysiology of these non–immune-mediated reactions has not been well clarified, and routine diagnostic tests are not yet available. Some severe transfusion reactions appear to be related to passive transfer of bioactive lipids, white blood cell (WBC)- and PLT-derived cytokines and other proinflammatory molecules and vasoactive peptides generated during storage and contained in the cell-free supernatant plasma of blood components. These mediators are usually not present in clinically important concentrations at the time of donation but are generated during storage, arising from plasma or released from the cells. The accumulation of bioactive lipids, CD40 ligand, and cytokines has been implicated as possibly causing several types of reactions, particularly fevers, chills, and pulmonary symptoms.11-18 Prestorage leukoreduction prevents buildup of large amounts of WBC-derived cytokines (e.g., interleukin [IL]-6, IL-8, MCP) during storage and can lessen the risk of transfusion-related fevers and chills. PLT-derived cytokines, for example, PLT-derived growth factor, soluble CD40 ligand (CD154), and RANTES, can accumulate in leukoreduced and in nonleukoreduced PLT components during storage. In these, removing the plasma with or without washing the PLTs can prevent further reactions in some patients. Recent studies show a correlation between the amount of soluble CD40 ligand (CD154) in PLT concentrates and development of fever, rigors, and severe acute pulmonary symptoms in transfused patients.17,18 Concentrations needed to cause reactions are not known and, in one study, more than 93 percent of those receiving PLT transfusions with high levels did not experience any reaction.17 Other substances generated in blood components are vasoactive and can cause acute hypotensive transfusion reactions within minutes of starting the transfusion. In the 1990s, reports surfaced about acute hypotension during bedside-filtered PLT and plasma transfusions in patients taking angiotensin-1–converting enzyme inhibitors (ACEi) and a new mechanism for hypotensive transfusion reactions was identified.19-21 The cause of these acute hypotensive reactions was the infusion of bradykinin generated during contact activation of plasma when PLT or plasma components were infused through negatively charged leukoreduction filters at the bedside to patients who have received ACEi.22-25 This mechanism involves generating bradykinin at the time of transfusion, not during storage of the blood component. Bradykinin (BK) causes vasodilation and hypotension and is a very short-lived peptide, being generated immediately prior to transfusion during bedside filtration. In patients taking ACEi, the main degradation pathway of BK is depressed by the ACEi and protection from the harmful effects of BK depends on BK degradation by the aminopeptidase P enzymatic pathway, an alternative pathway of BK degradation that is important when ACEi is present.24,25 Aminopeptidase P (APP) activity is low in a subset of healthy persons.26 These patients are particularly susceptible to various BK-mediated ACEi-related complications including acute hypotension after receipt of filtered blood, ACEi-related angioneurotic edema, and BK-mediated hypotension during extracorporeal circulation of blood through negatively charged dialysis membranes and equipment used during therapeutic apheresis procedures.25,27,28 Patients with low APP activity do not degrade and inactivate bradykinin as well as others, allowing a prolonged half-life of BK and symptomatic vasodilation and acute hypotension. There is recessive inheritance of APP levels as well as susceptibility to ACEi-related angioneurotic edema and ACEi-related hypotensive reactions during dialysis.28,29 Families of patients experiencing ACEi-related hypotensive reactions during transfusion and during therapeutic apheresis procedures have not yet been studied to see if there is a similar genetic control of susceptibility. The underlying mechanism of hypotension after infusion of plasma derivatives containing contact activation products was observed by Alving and colleagues in the 1970s.30 They described acute hypotension during rapid infusions of plasma protein fractions and, to a lesser degree, 5 percent albumin containing large amounts of kallikreins that contaminated the products and generated bradykinin from high-molecular-weight kininogens. These kallikreins were manufacturing impurities. In contrast, in some PLT transfusion hypotensive reactions, exposure of plasma to negatively charged filters appear to be the initiator of the contact activation system. Factor XII (Hageman factor) is activated resulting in generation of kallikrein from prekallikrein (Fletcher factor), which cleaves and releases BK from HMW kininogen (Fitzgerald factor). The activation of the contact system can lead to the production of bradykinin, blood coagulation, fibrinolysis, and complement activation.31 Not all hypotensive reactions to PLT transfusions in patients taking ACEi can be attributed to bedside filtration through negatively charged filters. Acute hypotension has also occurred with positively charged filters and with PLTs that were filtered before storage.32,33 It is in this context that the study by Moreau and coworkers34 from work in the laboratory of A. Adam at the University of Montreal and published in this issue of TRANSFUSION is of special interest. Moreau and coworkers have shown small increases of BK levels in whole blood–derived PLT concentrates and in PLT-poor plasma units after prestorage filtration.34 To my surprise, 22 percent of PLT concentrates and of PLT-poor plasma units had very large increases in BK levels on Day 5 of storage at room temperature with horizontal agitation. The BK level in these 22 percent was greater than 1 ng per mL and one reached 24 ng per mL. The high BK levels declined to baseline on Day 7 of storage. In the units with high BK levels on Day 5, the kinin-forming capacity and kininogen levels were low on Day 5, indicating that the kinin-forming capacity was depleted. This observation provides a possible explanation for why BK levels declined to near baseline by Day 7; BK was no longer being generated and its degradation was unimpaired. The activity of metallopeptidases in donor plasma responsible for BK degradation did not decline during storage, so increased formation of BK rather than decreased degradation seems more likely to be the mechanism for high BK levels accumulating during storage. No BK generation or accumulation was observed in most of the units and in these the plasma kinin-forming capacity was normal on Day 5 of storage. The study by Moreau and associates suggests a new mechanism whereby patients taking ACEi medications can experience acute BK-mediated hypotension after PLT transfusions. Rather than being generated by bedside filtration, this study suggests that BK can be generated in some PLT concentrates during storage. Whether the amount of BK generated is sufficient to cause vasodilation and hypotension in transfused patients is uncertain, however. Studies are needed to document the BK concentration in the specific blood components implicated in posttransfusion hypotensive events and in patients while being hypotensive. The study by Moreau and colleagues raises the question of whether BK is generated from contact activation during storage in standard blood containers or by other mechanisms. Assuming that the blood containers are biocompatible, there could be mechanisms other than contact activation by which BK could be generated. During blood component storage there is some continuous coagulation, complement activation, and fibrinolysis and these entities could be another possible source of BK generation.35-37 A separate study by Moreau and coworkers showed that during in vitro whole-blood clotting and the simultaneous activation of coagulation and fibrinolysis, BK is generated within 5 minutes during clot formation and its vasoactive breakdown product des-Arg-bradykinin is generated during clot lysis 90 minutes later (Albert Adam, November 2006, personal communication). Molinaro and coworkers38 have also shown that BK can be generated during fibrinolysis in an in vitro system of adding tissue plasminogen activator directly to plasma. Studies with the blood component containers should be performed to evaluate the effect of the container on the contact activation system of plasma. Although the presence of BK in PLT components may be a serious risk for some recipients, there are several reasons why BK generation during PLT storage might not represent a major risk to most patients receiving PLT transfusions. Only 22 percent of the units had high BK levels during storage. Many PLT transfusions will not be transfused during the short storage period that BK levels are high. Most patients are not taking ACEi medications and only a fraction of the ACEi users will have a low level of APP, an alternative BK degradation pathway. In the patients with normal APP activity, infused BK should be degraded normally and rapidly without causing symptoms. The study of Moreau and colleagues adds evidence that changes of blood components during storage might have adverse impact on recipients. Many questions are unanswered and need further study. Further studies are needed to see if these findings can be confirmed and similar studies with apheresis PLTs should be done.