Complications of Transfusions

Overview
Although RBC transfusions can help ameliorate many of the acute and chronic complications of SCD-and, at times, can be life-saving-their administration is associated with a wide variety of complications. Some transfusion-associated events are relatively mild, while others can be severe or even fatal. Health care providers should become familiar with the range of transfusion complications and learn their signs and symptoms as well as appropriate diagnostic testing, prevention strategies, and therapeutic interventions when warranted.
This section discusses alloimmunization, autoimmunization, iron overload, hemolysis, and hyperviscosity-the most commonly occurring side effects of transfusion. After a description of the side effects and a summary of the evidence, this section concludes by identifying some areas in which additional research is needed.

Alloimmunization and Autoimmunization

Background
Human erythrocytes express a large number of surface proteins, glycoproteins, polysaccharides, and glycolipids that are potentially immunogenic. Following an erythrocyte transfusion, if the donor erythrocytes have a different antigenic profile from those of the recipient’s own erythrocytes, an immunological response by the recipient against the “foreign” antigens can result in a process known as alloimmunization. Polysaccharide antigens generally elicit only immunoglobulin M (IgM) responses, but other erythrocyte antigens elicit an immune response that begins with production of polyclonal IgM alloantibodies within 3- 7 days of antigenic stimulation and then evolves to polyclonal IgG alloantibodies over several weeks.

Immunoglobulin G (IgG) alloantibodies persist for many years, although their titer may wane to low or undetectable levels. Almost all IgM alloantibodies, and some lgG alloantibodies, can bind to the transfused erythrocytes and fix complement, a set of serum proteins that bind to the erythrocyte and cause direct hemolysis. The result of alloimmunization is usually destruction of transfused erythrocytes that express the antigen, but the pathophysiology of red blood cell destruction and immune-mediated clearance is complex and depends upon several features including the antibody isotype, titer, and ability to fix complement. Occasionally, the recipient’s own erythrocytes become immunogenic and stimulate an immune response known as autoimmunization; most autoantibodies are lgG, and some fix complement. Autoantibody formation can occur at any time but occurs most frequently in patients who have already developed multiple alloantibodies.
Alloimmunization usually limits the ability to find compatible blood for future transfusions and increases the
risk for delayed hemolytic transfusion reactions, so efforts to avoid alloantibodies seem warranted.

Summary of the Evidence
The systematic review summarized more than 60 longitudinal and cross-sectional studies, involving more than 6,000 participants, in which alloimmunization or autoimmunization was described in adults and children with SCD undergoing transfusion. Rates of alloantibody formation ranged from 6 percent to 85 percent, while autoantibody formation ranged from 4 percent to 10 percent. These studies provide incidence and prevalence data only, and none compared the effectiveness of preventive strategies.

Most alloimmunization developed against erythrocyte antigens in the Rh blood system (D, Cc, Ee) and other minor blood groups (e.g., Kell, Kidd, Duffy). Phenotype matching of these antigens between transfusion donor and recipient may lower the alloimmunization rate, with a reported rate of 0- 7 percent described in studies where strict matching criteria were employed.

Iron Overload

Background
Transfused erythrocytes, whether administered through sporadic or repeated procedures, present an iron load to the recipient. The vast majority of the iron is carried by hemoglobin within the erythrocytes. As a rough calculation, 1 milliliter of erythrocytes contains approximately 1 milligram of iron, so for every 3-4 units of packed erythrocytes, 1 gram of iron enters the body. This process is clinically relevant, because adults normally have a total of only 4-5 grams of iron in their entire body, so this amount increases quickly after repeated transfusions. More importantly, there is no physiologic means to remove excess iron. Regulation of iron homeostasis normally occurs at the level of absorption through the hormone hepcidin, which inhibits the transport of gastrointestinal iron into the body. Because transfused blood represents iron that circumvents the normal pathways of iron regulation, this excess iron accumulates in tissues and can become pathological.

Hemosiderosis is a condition that reflects a large iron burden affecting normal organ function. The liver, pancreas, and heart are particularly vulnerable to iron overload. Chelation therapy can be used to remove excess iron. A number of different medications are used for chelation, but a thorough review of chelation dosing and management is beyond the scope of these guidelines. Deferoxamine is given by subcutaneous or intravenous route and leads to iron excretion through both urine and feces, whereas deferasirox is given orally once a day and removes iron primarily through the gastrointestinal tract. Deferiprone is taken orally three times a day and requires close monitoring due to the risk of agranulocytosis. Patients on monthly chronic transfusions typically receive chelation therapy to reduce iron burden, to attempt to normalize ironstores, and to avoid organ damage
from hemosiderosis.

Diagnostic Tools for Assessing Iron Overload
Changes in serum ferritin (SF) roughly correlate with iron loading, but the relationship is too inaccurate to use as a reliable method for evaluating iron status. Rather, SF is used as a biomarker to track qualitative trends of iron loading and chelation efficacy over time. Liver biopsy has been the gold standard in the diagnosis of iron overload but carries procedural risks and the possibility of sampling error. To avoid this invasive procedure, new diagnostic tools using MRI have been developed; these tools image the whole organ to quantify liver iron. Data are limited on the sensitivity and specificity of these new technologies to quantify liver iron in individuals with SCD. However, a significant body of literature supports the use of MRI as a substitute for liver biopsy for diagnosing iron overload in individuals with thalassemia. There is no reason to believe that the quantification of tissue iron would be different in individuals with SCD, and there is literature where MRI was used as (1) a screening tool for identifying patients eligible to participate in a trial of chelation therapy (80 patients; Cappellini et al.), (2) a tool to monitor outcomes in a study of chronically transfused SCD patients (15 patients; Hernandez et al. 1988232), or (3) a tool in studies examining different chelation regimens (15 patients; Voskaridou et al. 2005; Cancado et al. 2009; Levin et al. 1995; Ghoti et al. 2010). Therefore, the
expert panel considered the results in thalassemia patients when making recommendations for individuals with SCD.

Summary of the Evidence
A total of 50 studies (2 RCTs, 35 observational, and 13 cross-sectional) plus 9 case reports related to transfusion-acquired iron overload were identified. One RCT compared the use of deferasirox (oral) to deferoxamine (subcutaneous injections) in adults and children. The trial included 195 patients who were all iron overloaded (SF of at least 1,000 ng/mL, along with liver iron content of at least 2 mg iron/g dry weight of liver tissue in patients receiving simple transfusions, and 5 mg iron/g dry weight of liver tissue in patients receiving exchange transfusions) and demonstrated that the two approaches yielded similar results. The second RCT was the STOP trial, which did not evaluate treatments for iron overload; however, enrolled children in this trial received chronic transfusion, which was associated with a rise in SF in the first year of the trial and which necessitated treatment with deferoxamine in several children. Twenty other observational studies compared different chelation agents, and all have consistently demonstrated reduction of iron overload as measured by several methods. Data regarding the comparison among the different chelating agents or against alternative approaches such as hydroxyurea and exchange transfusion are unavailable or of very low quality.

Most studies used an SF level >1,000 ng/mL to diagnose patients with possible iron overload (often an inclusion criterion in the study). However, some studies used cutoffs of 1,500 ng/mL or higher. SF changes were
nonlinear. Levels less than 1,500 ng/mL indicated mostly acceptable iron overload; levels of 3,000 ng/mL or greater were specific for significant iron overload and were associated with liver injury. Using a cutoff of 2,500 ng/mL, Karam et al reported that SF had sensitivity of 62.5 percent and specificity of 77.8 percent for identifying liver iron concentrations of 7 mg iron/g dry liver tissue or greater. One observational study defined iron overload by liver iron concentration of at least 2.2 mg iron/g dry weight of liver tissue.422 Sufficient data were not found to allow the estimation of diagnostic accuracy of MRI, although many chelation studies used MRI findings as inclusion criteria.

Hemolysis

Background
Hemolysis (the breakdown and destruction of donor erythrocytes) can occur during or after a transfusion. It is important to note that the mechanism of transfusion-related hemolysis is immunologic, in contrast to the hemolysis of sickled erythrocytes, which is an intrinsic red cell defect. Most transfusion-associated hemolysis occurs 1 to 4 weeks after red cell transfusion and is called a delayed hemolytic transfusion reaction (DHTR). DHTR is related to immune-mediated mechanisms. The most common pathophysiology is preexisting or new IgG alloantibodies that bind to the erythrocytes and lead to accelerated clearance by macrophages in the extravascular compartment within the spleen, liver, marrow, and other parts of the reticuloendothelial system (RES). If the antibodies also fix complement, then erythrocyte destruction is further accelerated through lysis directly within the intravascular compartment. Both extravascular and intravascular hemolysis are manifest by shortened red blood cell survival, worsening anemia, and increased titers of antibodies found either on the erythrocytes themselves (positive direct antiglobulin, or “Coombs” test) or in the serum (positive indirect antiglobulin test) after the transfusion.
DHTRs can be associated with hyperhemolysis or bystander hemolysis. In this life-threatening complication of transfusion, patients will hemolyze not only the transfused blood but also their own RBCs, causing a profound
anemia. This complication is recognized when the hemoglobin falls below pretransfusion levels and is often
associated with reticulocytopenia and a positive direct Coombs test suggesting autoimmune destruction of RBCs.

Clinicians should have a high index of suspicion for hemolysis after transfusions, and they should coordinate diagnostic testing with the appropriate blood bank or transfusion service. Avoidance of future hemolytic events depends on proper diagnostic testing and avoidance of offending erythrocyte antigens.

Key Question

Summary of the Evidence
Three RCTs, 17 observational studies, and 47 case reports were identified related to hemolysis in association with transfusions. The RCTs included more than 300 patients. The studies described a prevalence of hemolytic reactions that ranged from 2 percent to 25 percent and an incidence of hyperhemolysis of 6 percent. There were no studies providing comparative effectiveness data on therapy. Descriptive studies reported successful management of the DHTR/hyperhemolysis (DHTR/H) syndrome with steroids, erythropoietin, and transfusion of phenotypically matched RBCs. The quality of evidence for management of these complications in SCD is very low, and data from transfusion in other populations may be indirectly applicable.

Hyperviscosity

Background
Transfusion of erythrocytes will increase the hematocrit of circulating blood, and increased viscosity could be problematic for patients with SCD. Avoidance of hyperviscosity is an important goal to prevent triggering a
voc.

Summary of the Evidence
No studies were found that described the effectiveness of a particular preventive or therapeutic approach for hyperviscosity in SCD.