What are the immediate and delayed adverse effects of blood transfusion?

What are the immediate and delayed adverse effects of blood transfusion?

Immediate Reactions

Immediate transfusion reactions are those that occur during or within 24 hours after transfusion whereas delayed reactions occur several days to years after the trans-fusion. The immediate adverse effects of blood transfusion are summarized in Table 49.1.

Acute Hemolytic Reaction Acute hemolytic transfusion reactions are the most frequent cause of fatal transfusion reactions reported to the FDA. These reactions result from either failure to detect blood group incompatibility or inadvertent transfusion of a blood product to the wrong patient. The risk of death from acute hemolytic reactions has been estimated to be 1:587,000–1:630,000 using data reported to the FDA. Data from the French Haemovigilance system reported approximately 1 death per 2 million units transfused. Most acute hemolytic reactions are caused by transfusion errors. Data from the United Kingdom Serious Hazards of Transfusion (SHOT) initiative indicated that 52% of events investigated involved incorrect transfusion of blood or blood compo-nent. An analysis of transfusion errors in New York State that resulted in transfusion of a blood product to other than the intended recipient indicated that erroneous transfusion occurred in 1 of 19,000 red cell units trans-fused. The frequency of fatal reactions was 1:1,800,000. Approximately half the errors occurred at the bedside and involved administration to the wrong recipient (38%) or phlebotomy errors (13%). Blood bank errors, including testing the wrong sample, transcription errors, and issuing the wrong unit, accounted for 29% of errors. Fifteen percent of events involved multiple errors.

The severity of the reaction is in general related to the amount of blood transfused. Mortality appears to be directly related to the volume of incompatible cells transfused. A review of 41 cases of hemolytic transfusion reactions causing renal failure demonstrated 25% mortality when 500–1,000 mL of blood was transfused and 44% mortality when more than 1000 mL was transfused. However, it is important to remember that fatalities have occurred when as little as 30 mL was transfused. Transfusion guidelines recommend that patients be monitored carefully during the first 15–30 minutes of a transfusion, so that there will be prompt recognition of a serious reaction.

It is important that clinicians recognize that blood group incompatibility, while common, is not the only possible cause of what appears to be a hemolytic transfu-sion reaction. The patient may have an intercurrent illness causing in vivo destruction of red cells. Autoimmune hemolytic anemias, congenital hemolytic anemias, drug-induced hemolysis, microangiopathic hemolytic anemias, and infections such as malaria or babesiosis where red cells are parasitized or clostridial infections where toxins hemolyze red cells, are among the disorders that may mimic an acute hemolytic transfusion reaction. Ventricular assist devices, membrane oxygenators, and artificial heart valves may also cause hemolysis. Factors that could cause non-immune hemolysis may also need to be considered. Improper storage of blood (very low temperatures at which blood freezes) or malfunction of a blood warmer may cause thermal destruction of red cells before transfusion. Infusing blood under pressure or through a small-gauge needle may also cause hemolysis. Finally, incompatible intravenous solutions will cause hemolysis if administered with blood.

Allergic and Anaphylactic Reactions Mild allergic reactions manifested primarily by urticaria are common following blood transfusion, occurring in approximately 1–3% of transfusions. Severe allergic reactions associated with hypotension, respiratory distress, and other cardiac and gastrointestinal symptoms of anaphylaxis are esti-mated to occur in 1:20,000 to 1:47,000 transfusions. In the French Haemovigilance system approximately 31% of reported events were allergic in nature. In most allergic reactions, patients appear to have preformed antibodies to a component of donor plasma, usually a plasma protein. In severe reactions, the patient may be IgA-deficient and have anti-IgA antibodies. On occasion the patient will have IgA but will lack one of the isotypic or allotypic determinants of the IgA class. Other patients have antibodies to other immunoglobulin classes or to proteins such as haptoglo-bin. Passive transfer of allergens or passive transfer of IgE antibodies in the donor product should also be considered. Although allergic reactions are generally thought to be IgE-mediated and histamine release from mast cells is the primary mediator, IgE antibody is not always demonstra-ble. Complement-derived anaphylatoxins such as C3a and C5a generated by immune complexes or other secondary mediators such as cytokines are responsible for anaphylac-toid symptoms.

Mild allergic reactions are usually treated with an oral or parenteral (intravenous or intramuscular) antihista-mine (diphenhydramine 25–50 mg). If a patient has a documented history of recurrent allergic reactions, prophy-lactic administration of an antihistamine is indicated. The mild allergic reaction may be the only reaction where there is general agreement that the transfusion could be inter-rupted, an antihistamine given, and the transfusion restarted when symptoms subside. Severe allergic or ana-phylactoid transfusion reactions should be treated as any other type of anaphylaxis is treated (i.e., epinephrine, volume expansion, oxygen supplementation and other respiratory support, etc.). Any patient having an anaphy-lactoid reaction should be tested for antibodies to IgA. Patients who have IgA antibodies must receive plasma and platelets from IgA-deficient donors. These products are usually available only from large regional blood centers. Red blood cells washed with 2 liters of normal saline will be satisfactory for transfusion. IgA-deficient patients who may need plasma derivatives pose a special problem because the standard labeling on these products does not indicate whether there may be trace amounts of IgA pres-ent. As little as 1 mg of IgA may trigger an allergic reaction. Clinicians should speak with the manufacturer’s represen-tative to determine whether a particular product and lot number is safe.

Circulatory Overload Circulatory overload is a frequent complication of transfusion therapy. It usually occurs in elderly patients with renal insufficiency, diminished car-diac reserve, or severe anemia. Circulatory overload is often included in acute pulmonary problems in many reports and accurate data on the incidence are not available. However, a report from the Mayo Clinic suggested that it occurred in 1/3,168 patients receiving red cell transfusions. The early symptoms of volume overload are nonspecific and include an increase in blood pressure, headache, devel-opment of a new cough, and a sensation of pressure in the chest. The development of any of these signs and symp-toms in a patient receiving transfusion therapy should be an indication to slow the infusion and monitor the patient more often. As cardiac decompensation progresses, dyspnea, orthopnea, tachycardia, cyanosis, and frank pulmonary edema may develop.

When the patient’s signs and symptoms suggest circula-tory overload, the transfusion should be stopped and the patient placed in a sitting position. Oxygen supplementa-tion should be provided and diuretics given to reduce the intravascular volume. If diuretics are ineffective, consider-ation should be given to phlebotomy.

Febrile Non-hemolytic Reactions The definition of a febrile non-hemolytic reaction is an increase in tempera-ture of ≥1°C following transfusion that cannot be explained by the patient’s clinical condition. This increase in temperature is usually accompanied by chills and rigors and sometimes headache, nausea, and vomiting. In most patients, fever develops during the transfusion. Usually the increase in temperature is ≤2°C. When the temperature increase is ≥2°C, bacterial contamination of the blood product and the development of an intercurrent infection must be considered. Some patients have chills, rigor, and feel cold but do not develop fever. The diagnosis of a febrile non-hemolytic reaction can be established only by exclud-ing other types of transfusion reactions accompanied by fever. In a community hospital population, approximately 0.5–1.0% of red cell transfusions are associated with febrile non-hemolytic reactions. In chronically transfused patients, the frequency is much higher.

Recipient antibodies directed against antigens on donor leukocytes are generally considered to be the cause of febrile non-hemolytic reactions. Initially it was believed that endogenous pyrogens (interleukin-1β, interleukin-6, and tumor necrosis factor) from donor leukocytes caused the febrile reaction. Recently, it has been suggested that complement activation following the interaction of recipi-ent antibodies with donor leukocytes causes activation of recipient monocytes. These activated monocytes are thought to release proinflammatory cytokines causing the reaction. Finally, since the widespread use of leukoreduc-tion filters to eliminate these reactions, it appears that cytokines produced by donor leukocytes prior to leuko-reduction may also cause febrile non-hemolytic reactions.

In the general population, only 15% of patients having a febrile reaction to a red cell product are likely to have a recurrent febrile reaction with the next transfusion. Therefore, many transfusion services do not recommend premedication or leukoreduction for the average patient until a second febrile reaction occurs. Febrile reactions are more common following platelet transfusions than red cell transfusions and are more common with older products than relatively fresh products. If reactions are mild, they can often be prevented by premedication with an anti-pyretic. If reactions are severe or if premedication does not prevent the reaction, leukoreduced products are indicated. In some multitransfused patients, it may be necessary to provide pre-storage leukoreduced products. Pre-storage leukoreduction will not remove all biologically active mediators because some of these are derived from platelets. Removal of most of the plasma from platelet products just prior to transfusion helps reduce reactions in patients not responding to pre-storage leukoreduction strategies.

Other Hypotensive Reactions Hypotension not uncom-monly accompanies severe immune reactions in which antibodies in the recipient react with donor cells or vice versa and transfusions where bacterial contamination of a blood product has been documented. In addition to these circumstances, hypotension has been described following the administration of leukoreduced products (both platelets and red cells), following the administra-tion of plasma-containing products to patients taking angiotensin-converting enzyme (ACE) inhibitors and fol-lowing the administration of plasma protein fraction to patients undergoing cardiopulmonary bypass. These reac-tions appear to be related to the generation of bradykinin under circumstances where enzymes important in bradykinin inactivation are either inhibited or a tissue con-taining these enzymes has been excluded from the circula-tion (cardiopulmonary bypass). Contact of plasma with negatively charged blood filters leads to contact activation of the intrinsic coagulation system and bradykinin genera-tion. There are at least five metallopeptidases responsible for the inactivation and metabolism of bradykinin. ACE is responsible for the hydrolysis of 60% of bradykinin in normal subjects and inactivation by aminopeptidase P (APP) is a second major metabolic pathway. Clinical reports have documented unexplained hypotensive reac-tions in patients on ACE inhibitors receiving platelet prod-ucts leukoreduced at the bedside and in patients under-going therapeutic plasma exchange. In vitro studies have documented an increase in bradykinin levels following fil-tration of platelet concentrates. The generated bradykinin was rapidly degraded and was undetectable after 1 hour of storage.

Treatment of a hypotensive reaction in a patient on ACE inhibitors should include immediate discontinuation of the product and appropriate resuscitation. If additional red cells are needed, washing is advised. If platelets are needed, the product should be pre-storage leukoreduced or filtered an hour prior to administration. Reduction of the residual plasma on the platelet product may also be helpful.

Transfusion-Transmitted Bacterial Infection Sepsis related to bacterial contamination of blood products is the second most common cause of fatal transfusion reactions reported to the FDA. In the United States between 1976 and 1999, 10% of transfusion-related deaths were caused by bacterial contamination. It is estimated that 0.2% of whole blood collections are contaminated. Bacterial contamination of platelets is more common than contam-ination of red blood cells but prevalence estimates vary widely. Estimates for red cells vary from 0.002% to 1.0% and for platelets from 0.04% to 10%. Some studies have suggested that contamination of pooled platelet concen-trates is more frequent than contamination of single donor platelets. The organisms contaminating blood products are usually from donor skin flora (i.e., Staphylococcus species, Propionibacterium acnes). However, asymptomatic donor bacteremia (Yersinia enterocolitica) and contamination of products from environmental sources (Pseudomonas species) may also lead to bacterial contamination of blood components.

The most common symptoms and signs associated with bacterial contamination of blood products are chills, fever, tachycardia, shock or hypotension, shortness of breath, back pain, and nausea and/or vomiting. Occasional patients may have an increase in blood pressure. Although these symptoms often develop immediately or within the first hour after transfusion is initiated, some patients may not develop symptoms or signs for several hours.

Two national studies of bacterial contamination of blood products performed in France and the United States provide the most current data on this complication of transfusion. In the United States from 1998 to 2000, suspected cases of bacterial contamination of blood prod-ucts were reported to the Centers for Disease Control (CDC) by blood collection facilities and transfusion serv-ices associated with the American Red Cross, the American Association of Blood Banks, and the Department of Defense. This study was given the acronym BaCon (Assessment of the Frequency of Bacterial Contamination Associated with Transfusion Reaction). In France, there is a national mandatory reporting system for adverse reactions to transfusion. From November 1996–1998, all adverse reactions suspected to be related to bacterial contamina-tion were investigated as part of the French BACTHEM Study.

The case definition for BaCon included the presence of one or more of the following signs or symptoms develop-ing within 4 hours of transfusion: fever ≥39°C or a change of ≥2°C from the pre-transfusion value; rigors; tachycardia ≥120 beats per minute, or a change of ≥40 beats per minute from the pre-transfusion value; a rise or drop of ≥30 mmHg in systolic blood pressure. Of 56 reported cases, 34 were confirmed with the same organism being cultured from the recipient and the component. There were 9 deaths. The rate of transfusion-transmitted bacteremia (in events per million units distributed) was 9.98 for single donor platelets, 10.64 for pooled platelets, and 0.21 for red cells. The rate of fatal reactions was 1.94 for pooled platelets, 2.22 for single donor platelets, and 0.13 for red blood cells. Fatal transfusion reactions were more likely to be associ-ated with contamination with gram-negative organisms.

Results from BACTHEM were similar. Of 158 suspected cases, 41 were confirmed. Twenty-five were associated with red blood cells and 16 with platelet products. This led to an estimated incidence rate per million components issued of 5.8 for red blood cells, 31.8 for apheresis platelets, and 71.8 for pooled platelet concentrate. Fatal reactions were uni-formly associated with gram-negative organisms.

In comparing results from current studies and earlier published reports, the organisms identified as contaminants have changed over time. In earlier studies, Yersinia enterocol-itica and Pseudomonas species were the predominant organ-isms isolated from red cells. In the BaCon study Serratia species accounted for most of the cases of red blood cell contamination and Acinetobacter was the second most frequent species cultured from blood components in the BACTHEM study. Currently, manufacturers are focusing on the development of systems to detect bacterial contamination in blood products and the development of nucleic acid bind-ing compounds that will prevent the proliferation of bacteria accidentally introduced into units at the time of phlebotomy.

Transfusion-Related Acute Lung Injury Transfusion-related acute lung injury (TRALI) is one of the most serious, underdiagnosed complications of transfusion. Typically, acute respiratory distress develops within 1–2 hours of starting a transfusion of a blood component that contains plasma, but some patients have developed symptoms as late as 6 hours after transfusion. Patients have severe hypoxemia and pulmonary edema. Hypotension and fever may also occur. The SHOT study from the UK reported that 11 of 169 (7%) reports involved acute lung injury. Data from the Mayo Clinic suggest that the incidence of TRALI may be 1:5,000 with plasma-containing transfusions. In Canada, data from Quebec reported 3 cases of TRALI in a population receiving 190,000 red cells and 3,000 units of plasma. At a hospital in Ontario, one case of TRALI was observed annu-ally and 12,000 red cell units were transfused.

In most investigations, donor plasma has contained anti-bodies either to granulocytes or to HLA antigens (both class I and class II). In a small percentage of cases, the recipient serum has contained such antibodies (prior to transfusion). However, antibodies have not been identified in all cases, and it has been hypothesized that lipid products from neutrophils may cause TRALI. A case of TRALI following the administra-tion of intravenous immune globulin was recently reported. Treatment for TRALI depends on the severity of the reaction. Supplemental oxygen and respiratory support including mechanical ventilation may be required. Hypotension is cor-rected and corticosteroids are usually given.

When a blood product is implicated in TRALI, a donor serum sample is obtained and tested for antibodies to leukocytes. Most such donors are multiparous women. When antibodies are identified, donors are usually no longer allowed to donate. If they have an unusual blood type, the red cells are collected and either frozen or washed, processes that eliminate residual plasma.

Delayed Reactions

Delayed adverse reactions to blood transfusion may be immunologic, related to transmission of a viral or parasitic disease, or caused by iron overload secondary to trans-fusion. Immunologic adverse effects include delayed hemolytic transfusion reaction (DHTR), immunization to red cell, platelet, leukocyte or plasma antigens, auto-immune phenomena triggered by alloimmunization, and graft-versus-host disease.

Delayed Hemolytic Transfusion Reactions Delayed hemolytic transfusion reactions usually result from failure to detect existing antibody because the antibody concentration has dropped below the detection level for the method used in antibody screening and crossmatching tests. The trans-fusion stimulates antibody production causing an acceler-ated destruction of transfused cells. Common presenting signs and symptoms include unexplained fever, a decrease in hemoglobin unexplained by clinical events, and an increase in bilirubin several days to weeks after transfusion. When a patient blood sample is examined, a positive direct antiglobulin test is present and often antibody can be eluted from the transfused cells. Antibody may also be detectable in patient serum at this time. Delayed reactions are usually mild and may go unrecognized. Clinically detectable hemolysis is reported to vary from 1:5,000 to 1:10,000 transfusions. Some delayed reactions are serious, causing disseminated intravascular coagulation, renal failure, and death. In the first 2 years of the SHOT study, 51 of 366 (14%) reports concerned delayed hemolytic transfusion reactions; and in two cases death was attrib-uted to the transfusion reaction.

Alloimmunization Alloimmunization to antigens on red cells is estimated to occur in 1–1.6% per donor unit provided that D-negative units are given to D-negative recipients. However, immunization may become a serious problem when a patient requires chronic transfusion and the red cell phenotypes of the blood donors vary from that of the recipient population. For example, most patients with sickle cell disease are African-American and in most communities in the United States the majority of blood donors are Caucasian. Some specialists in the treatment of sickle cell disease have recommended prospective match-ing of clinically important blood groups for patients on chronic transfusion protocols to prevent immunization and lower the risk of delayed transfusion reactions. Immunization to HLA antigens through routine blood transfusion may also pose problems for patients. Red blood cells and platelet products contain a large number of white blood cells that normally express HLA antigens. Patients having many transfusions with unmodified red cell or platelet products may become immunized to antigens of the HLA system. Immunization to HLA antigens will make a patient with thrombocytopenia refractory to random donor platelet transfusions and may make finding a com-patible solid organ donor impossible for a patient needing a heart, kidney, or small bowel transplant. The prophylactic use of leukoreduced red cell and platelet products may be indicated for these patient populations.

Immunization to platelet-specific antigens may also occur following transfusion. In addition to refractoriness to platelet transfusions this type of immunization may lead in rare instances to the disorder post-transfusion purpura. In post-transfusion purpura patients develop severe throm-bocytopenia 5–10 days following transfusion. Patients are usually women who have been previously pregnant or transfused. Although several platelet-specific antigen systems have been reported to cause this disorder, in most cases patients lack a platelet-specific antigen HPA-1a (Pl^A1), and have made an anti-HPA-1a. HPA-1a is a high-frequency platelet-specific antigen with only 2% of the population being HPA-1a-negative. Through mechanisms that are not well understood the anti-HPA-1a destroys the transfused HPA-1a-positive transfused platelets and the patient’s own HPA-1a-negative platelets. Post-transfusion purpura is usu-ally self-limited with spontaneous recovery within 3 weeks. Occasional patients with severe symptomatic thrombo-cytopenia require treatment. High-dose intravenous immune globulin (IVIG) is the treatment of choice. Random platelets are generally HPA-1a-positive and will not increase the platelet count. Antigen-negative platelets may be helpful when given with IVIG.

Graft-Versus-Host Disease Transfusion-associated graft versus-host disease (TA-GVHD) is a rare but usually fatal complication of transfusion in which viable lymphocytes in donor blood products transfused to an immunologically compromised patient engraft. These foreign lymphocytes recognize the HLA antigens of the transfusion recipient as foreign and generate a characteristic immune response characterized by rash, fever, liver function abnormalities, diarrhea, and marrow dysfunction. Symptoms usually develop 8–10 days after transfusion. TA-GVHD was origi-nally described in immunocompromised individuals. It was described in newborns having exchange transfusion, patients with congenital forms of immune deficiency, and patients immunosuppressed by intense chemotherapy. Subsequently, it was reported in immunologically normal individuals transfused with blood from an individual who was either HLA-identical or homozygous for a shared HLA haplotype. The risk of such matched transfusions in unre-lated populations varies.

Treatment of TA-GVHD is rarely effective, with 90% of patients dying from the disorder. Transfusion guidelines are focused on prevention through blood product irradia-tion. Irradiation of blood products requires a specific physician order. However, most blood bank/transfusion services have developed guidelines for blood product irradiation. Transfusions to premature infants, neonates, individuals with congenital immune deficiency, recipients of hematopoietic stem cell transplants, patients receiving intense immunosuppressive chemotherapy (for leukemia, Hodgkin’s disease, and other lymphomas), and directed blood product donations from family members are gener-ally irradiated routinely. Physicians should be informed about irradiation guidelines at the institutions where they practice. Products for some patients may automatically be irradiated based on admitting diagnosis but other cases may require a specific order. It is always best to specifically order irradiated blood, rather than to depend on a stan-dard operating procedure. This practice minimizes the opportunity for error.

Iron Overload Transfusion hemosiderosis is common when patients with hematologic disorders require chronic transfusion. Typically these are patients with hemo-globinopathies such as thalassemia or sickle cell disease. However, patients with myelodysplastic syndromes are also at risk for this disorder. Chronically transfused patients should have ferritin levels monitored and should be treated with iron-chelating agents such as desferoxamine when needed.

Transfusion-Transmitted Viral and Parasitic Infection Currently blood donations for allogeneic use have been tested by FDA-licensed tests and found negative for antibodies to human immunodeficiency virus (anti-HIV), hepatitis C virus (anti-HCV), human T-cell lym-photrophic virus (anti-HTLVI/II) and hepatitis B core antigen (anti-HBc), as well as HIV-antigen (HIV-1-Ag, also called p24 antigen) and hepatitis B surface antigen (HbsAg). Beginning in March 1999, nucleic acid amplifica-tion testing (NAT) for HIV and HCV has been performed on most of the blood collected in the United States under an investigational new drug application (IND) approved by the FDA. Traditional tests are performed on individual samples, but NAT testing is performed on pooled samples. Two manufacturers have developed test systems using NAT. Gen-Probe (San Diego, CA) developed a multiplex system that tests for HCV and HIV at the same time. Roche Molecular Systems (RMS, Pleasanton, CA) developed inde-pendent NAT tests for HIV and HCV. Since the imple-mentation of NAT under FDA IND one case of HIV transmitted by blood transfusion has been reported in the United States. Although the new tests are very sensitive, the window period for HIV remains at 10–11 days. NAT testing is not yet legally mandated, but will be as soon as suf-ficient licensed kits are available to screen the entire US blood supply. In the meantime, the FDA is requiring that all allogeneic units be NAT-tested unless there is an extremely urgent situation such as the 9/11 terrorist attacks. At the present time, NAT testing for HBV on pooled samples does not appear to be superior to serologic tests for HbsAg. It is unclear when NAT testing for HBV will be introduced.

The Gen-Probe test for HIV and HCV was licensed in February 2002. Licensure of the RMS test is imminent. Table 49.2, derived primarily from the American Red Cross experience, provides data on the window period reduction with NAT using pooled testing for transfusion-transmitted viruses. For HIV the residual risk with NAT decreased from 1:1,300,000 to 1:1,900,000 per million red cells transfused. For HCV, the window period could become 10–12 days. Although the estimated reduction in the window period for HCV (time from infection to seropositivity) is calcu-lated to be 32.5–40 days, there is a very high level of viremia in HCV infection, with virus becoming detectable 10–12 days after exposure. Thus NAT may reduce the HCV window period to 10–12 days. If these estimates are correct, the residual risk could be as low as 1:1,600,000.

Blood is not routinely screened for cytomegalovirus (CMV) or parvovirus B19 even though these infections can be transmitted by transfusion. However, the standard of care is to provide CMV-seronegative blood for individuals who are CMV-seronegative and are either immunodefi-cient (infants, congenital or acquired immunodeficiency) or are candidates for hematopoietic stem cell transplanta-tion or solid organ transplantation where the tissue donor is also CMV-seronegative. Blood centers screen part of their inventory for CMV to provide CMV-seronegative blood for these special patients. Most adult blood donors are CMV-seropositive. The prevalence of seropositivity for CMV varies in different geographic areas of the United States. Most of these donors have latent infections but there is not currently a test available to distinguish between donors who are infectious and those who are not. In the absence of CMV-seronegative blood, leukoreduced blood components are considered an appropriate alternative.

Other infections such as malaria, babesiosis, and try-panosomiasis may be transmitted by transfusion. There is no specific testing for malaria in the United States, and exclusion of infected donors is through the medical his-tory. For other potential infections, local blood centers often add specific questions related to local infectious risks.

NAT for HIV and HCV was licensed in the United States in 2003. Investigative NAT testing for West Nile virus, which can be transmitted by transfusion and transplantation, is currently being conducted under FDA-IND in the United States. Systems for detecting bacterial contamination of platelets are now licensed.