Febrile, non-hemolytic reactions

Febrile, non-hemolytic reactions are the most commonly described transfusion reactions in cats and dogs. These reactions are characterized by a temperature increase of >1°C during or immediately after transfusion. Other clinical signs can include vomiting and tremors. Most febrile, non-hemolytic reactions are mild and self-limiting. The pathogenesis of these reactions is likely multifactorial. Transfused blood products contain residual white blood cells (WBCs) from the donor. Circulating antibodies in the recipient react with the transfused WBCs, inducing release of cytokines and other mediators from the donor WBC. Blood products also contain biological response modifiers such as lipids, complement protein, and cytokines that induce a febrile response in the recipient (Blumberg et al. 2006; Fung and Heddle 2013). Fever can also be the first sign of an acute hemolytic transfusion reaction or sepsis, therefore any patient with a fever at the onset of a transfusion should be carefully monitored for the development of more serious clinical signs.

White blood cells, and in some instances platelets, are removed from blood products by passing whole blood through a leukoreduction filter (McMichael et al. 2010). Leukoreduction filtration is usually performed shortly after collection (“pre-storage”), but can also be performed immediately prior to transfusion (“post-storage”). Because >99% of WBCs are removed from the blood product, pre-storage leukoreduction abrogates in vitro production of cytokines (Corsi et al. 2014). Leukoreduction causes a significant decline in the number of febrile non-hemolytic transfusion reactions in humans (King et al. 2004; Paglino et al. 2004; Yazer et al. 2004). The effect of leukoreduction on inflammation caused by blood transfusion has been studied in healthy dogs. In one canine study, leukoreduction mitigated the inflammatory response caused by autologous transfusion when compared to transfusion of non-leukoreduced blood (McMichael et al. 2010). However, the magnitude of inflammation was similar after transfusion of leukoreduced versus non-leukoreduced blood in another study (Callan et al. 2013). Further study of the clinical benefits of leukoreduction in veterinary patients is needed.

Acute hemolytic reactions

Acute hemolytic transfusion reactions result from type II hypersensitivity (antigen–antibody) reactions, due to the presence of antibodies in recipient plasma against donor red blood cell (RBC) antigens. Antibodies (primarily immunoglobulin (Ig)M, also can include IgG) adsorb to the donor RBC surface and can activate the complement cascade. This results in intravascular hemolysis of transfused RBCs during or within 24 hours of transfusion. Extravascular hemolysis can also occur if complement activation is minimal or absent (Massey et al. 2013). Clinical signs range in severity and include fever, vomiting, defecation, tachypnea, hemoglobinuria, hemoglobinemia, tachycardia, recumbency, dyspnea/apnea, hypotension, and other signs of shock. Acute kidney injury, disseminated intravascular coagulation, and death are potential sequelae (Auer et al. 1982; Giger and Bücheler 1991).

Cats have naturally occurring alloantibodies to their non-self blood type; type A cats will have alloantibodies to the B blood group and vice versa (Bücheler and Giger 1993). In particular, type B cats can have strongly agglutinating and hemolyzing anti-A antibodies. Therefore, type B cats can have severe hemolytic reactions within minutes of transfusion of a small amount of type A blood (Auer et al. 1982). Transfused type A RBCs administered to type B cats have an average half-life of 1.3 hours, with all type A RBCs destroyed within 24 hours. Reports of type A cats receiving type B blood document relatively less severe hemolytic reactions, but this depends on the anti-B titer of the recipient. The average half-life of type B RBCs transfused to type A recipients is 2 days. In contrast, the half-life of type-matched transfused RBCs ranges from 29 to 39 days in cats (Giger and Bücheler 1991). Due to the presence of naturally occurring alloantibodies in cats and the potential for severe, fatal transfusion reactions, it is imperative that all cats are blood typed prior to transfusion and that type-matched blood is always administered.

Dog erythrocyte antigen (DEA) 1 is considered the most antigenic and clinically relevant canine blood type, as it usually causes the most serious alloantibody-mediated transfusion reactions (Hale 1995). In recent years, this blood group system has undergone nomenclature changes. Previously, the DEA 1 blood group system described at least two subgroups: DEA 1.1 and DEA 1.2. A third type, DEA 1.3 (A₃), was also described, but could not be verified due to lack of reagent availability (Symons and Bell 1991; Andrews et al. 1992; Hale 1995). Quantitative flow cytometry research has since discovered a large variation in DEA 1 expression among DEA 1 positive dogs. Dogs previously classified as DEA 1.2 positive are instead thought to have weak DEA 1 expression. Current clinical recommendations are to classify dogs as DEA 1 positive or negative; DEA 1 positive dogs should also be classified based on the strength of their DEA reaction (weak to strong) according to the blood-typing method used (Acierno et al. 2014).

Dogs do not have naturally occurring alloantibodies to DEA 1, but can develop them after a DEA 1 mismatched transfusion. The expected half-life of transfused RBCs ranges from 43 to 104 days in dogs (Novinger et al. 1996; McDevitt et al. 2011). Previous sensitization of a DEA 1 negative dog to DEA 1 positive blood can lead to an acute hemolytic reaction upon subsequent transfusion of DEA 1 positive blood (Giger et al. 1995). It is unknown if dogs with weak DEA 1 expression will have a transfusion reaction to blood with strong DEA 1 expression. Current clinical recommendations are to transfuse weak DEA 1 positive recipients with DEA 1 negative blood (Acierno et al. 2014). Natural antibodies to DEA 3, 5, and 7 exist in some dogs, but transfusion reactions caused by these incompatibilities are typically mild and delayed (Hale 1995). Sensitization to other DEAs and subsequent hemolytic transfusion reactions have been reported (Callan and Giger 1995; Hale 1995; Melzer et al. 2003). Current evidence shows that pregnancy does not induce DEA alloantibody formation in bitches (Blais et al. 2009).

Some patients lack high-frequency RBC antigens (Blais et al. 2007; Weinstein et al. 2007). Certain Dalmatians, Doberman Pinschers, and Shih Tzus lack an antigen (Dal) commonly found on the surface of most dog RBCs. Transfusion with Dal-positive blood sensitizes Dal-negative transfusion recipients, leading to acute or delayed hemolytic reactions with subsequent transfusions (Blais et al. 2007; Goulet et al. 2014). Similarly, a common RBC antigen (Mik) is absent in some cats. Some Mik-negative cats have naturally occurring alloantibodies against this antigen, which can lead to an acute hemolytic transfusion reaction despite administration of compatible AB blood type. Other Mik-negative cats acquire alloantibodies after transfusion with Mik-positive blood and are at risk of a hemolytic reaction during subsequent transfusion. Pre-existing anti-Mik alloantibodies might be detected on a crossmatch, but this depends on the titer present in the recipient (Weinstein et al. 2007).

Allergic/anaphylactic reactions

Allergic transfusion reactions range in severity from mild to severe anaphylaxis and occur within minutes to hours of transfusion initiation. Type I hypersensitivity responses are the typical cause of allergic transfusion reactions. These reactions can be triggered by patient exposure to a protein or other substance in the donor plasma, especially if the recipient has previously been sensitized to this allergen. The allergen binds to antibody (IgE or IgG) bound by Fc receptors on the surfaces of mast cells and basophils. As mast cells and basophils become activated, granules and chemical mediators (e.g., histamine, platelet activating factor) are released to cause a type I hypersensitivity reaction. Biological response modifiers might also cause non-allergen dependent hypersensitivity reactions. Chemokines and cytokines such as vascular endothelial growth factor and transforming growth factor β1 can directly activate mast cells and basophils, leading to a similar allergic reaction (Fung and Heddle 2013). Plasma and platelet products increase the risk of allergic transfusion reaction in humans. Removing plasma from blood products significantly decreases incidence of allergic transfusion reactions, suggesting that plasma proteins are important mediators of these complications (Tobian et al. 2011).

Delayed hemolytic transfusion reactions

Delayed hemolytic transfusion reactions occur >24 hours (usually 3–5 days) after the transfusion and are usually due to a secondary or anamnestic immune response. After a first mismatched transfusion, patients can be sensitized to a specific RBC antigen and mount an antibody titer. This reaction might not result in clinical abnormalities and, over time, the antibody titer declines. Because of this low antibody titer, crossmatching prior to transfusion might not detect the incompatibility. After repeated exposure to the same RBC antigen during a subsequent transfusion, the secondary immune response rapidly induces a higher level of antibody formation, leading to a delayed hemolytic reaction (Massey et al. 2013). Delayed hemolytic transfusion reactions can also occur after first exposure to a novel RBC antigen, triggering a primary immune response. Hemolysis can occur within weeks after transfusion in these recipients, upon sufficient alloantibody production (Massey et al. 2013). Delayed hemolytic transfusion reactions are typically mild in comparison to acute hemolytic reactions and are less likely to be detected clinically. Signs of a delayed hemolytic transfusion reaction include an unexpected decrease in the hematocrit or packed cell volume (PCV) after transfusion, icterus, and rarely hemoglobinemia/hemoglobinuria.

Some dogs that are negative for DEA 3, 5, or 7 have low titers of naturally occurring alloantibodies against the corresponding antigen (Swisher and Young 1961; Hale 1995), therefore transfusion of the corresponding antigen can lead to a delayed transfusion reaction. Although an acute immunologic transfusion reaction is unlikely upon transfusion of DEA 1 positive RBCs to an unsensitized DEA 1 negative recipient, a delayed hemolytic reaction can occur. Anti-DEA 1 antibodies can form within 2 weeks, causing a rapid decline in transfused RBCs as the anti-DEA 1 titer increases in the patient (Young 1949). This delayed reaction is unlikely to cause clinical signs, but it decreases transfusion efficacy and duration of benefit.

Immune complex (type III hypersensitivity) formation

Type III hypersensitivity reactions are caused by antigen–antibody complex formation. Immune complexes can be deposited in tissues where the complement cascade is activated and chemotactic factors are released. Neutrophils migrate to this site and release enzymes and oxidants that cause acute inflammation and tissue damage. Immune complexes can also be formed in circulation when an antigen directly enters the blood. Circulating immune complexes are commonly deposited in glomeruli, resulting in glomerulonephritis. Circulating complexes can also deposit in endothelial cells, causing vasculitis, and the synovium causing arthritis (Tizard 2013). Clinical signs occur within hours to weeks of antigen exposure.

Human serum albumin (HSA) has been used to treat hypoalbuminemia in critically ill dogs and cats. Despite the potential benefits of HSA, severe adverse reactions following transfusion of this product have been reported. In a report of six healthy dogs given 25% HSA, one dog developed an immediate hypersensitivity reaction and all dogs developed delayed reactions consistent with type III hypersensitivity within 2 weeks of transfusion. Clinical signs included peripheral edema, primary vasculitis, and lameness. Two of six dogs developed more severe signs, including disseminated intravascular coagulation (DIC) in one dog and oliguric kidney failure in both dogs. Despite supportive treatment, both of the severely affected dogs died (Francis et al. 2007). Other studies have reported immediate and delayed hypersensitivity reactions in healthy dogs receiving initial and subsequent 25% HSA transfusion (Cohn et al. 2007). Retrospective studies of critically ill and hypoalbuminemic dogs and cats transfused with 5% and 25% HSA do not reveal any severe short-term hypersensitivity reactions, but long-term data is lacking and delayed reactions might not have been reported (Mathews and Barry 2005; Viganó et al. 2010). It is possible that critically ill dogs mount a diminished immune response to HSA compared to healthy dogs, explaining the more severe reactions after HSA transfusion in healthy dogs. However, one report describes type III hypersensitivity reactions leading to vasculitis in two critically ill dogs after HSA transfusion (Powell et al. 2013). Based on these reports, the safety of HSA transfusions is questionable in dogs and cats.

Post-transfusion purpura

A marked thrombocytopenia occurring 5–10 days after transfusion has been reported in human patients. Post-transfusion purpura is thought to result from the secondary immune response, leading to production of alloantibodies against specific platelet antigens after repeated transfusions. This reaction also destroys the patient’s own platelets and can cause clinical signs of bleeding (Murphy 2013). Post-transfusion purpura has been reported in a dog after receiving multiple transfusions (Wardrop et al. 1997a). The overall incidence of this transfusion reaction in veterinary patients is unknown but appears to be rare.

Non-immunologic hemolysis

Non-immunologic hemolysis of RBCs can occur ex vivo (prior to or during the transfusion) secondary to RBC aging, improper storage, bacterial contamination of the product, or improper administration (Prittie 2003; Patterson et al. 2011). Hemolysis of stored RBCs secondary to aging is a marked manifestation of storage lesions, which compromise cell viability.

Improper RBC storage can lead to hemolysis; freezing RBCs will cause cell membrane rupture. To minimize hemolysis of RBCs prior to transfusion, RBC products should be stored in a dedicated refrigerator with a constant temperature set between 1 and 6°C. Temperature changes outside of recommended storage conditions can lead to decreased cell deformability and increased osmotic fragility, resulting in decreased RBC survival (Perrotta and Snyder 2001). Storage at warmer temperatures increases the metabolic rate of RBCs, causing the cells to use nutrients in storage media at a faster rate. Refrigeration slows down storage lesions, compared to RBCs stored at room temperature. Refrigeration also inhibits bacterial growth within RBC units. Bacterial contamination can decrease oxygen within a stored red cell product, decreasing stored RBC lifespan. Release of endotoxins and bacterial enzymes can cause excessive hemolysis and discoloration of the blood product (Hess 2010).

Bloodborne infectious disease transmission

Infectious disease screening is strongly recommended prior to using dogs and cats as blood donors and routine parasite prophylaxis is suggested in donors to decrease the risk of bloodborne infections (Wardrop et al. 2005). Ideally, annual infectious disease screening should be performed and cats that go outdoors should be omitted from the donor pool. Specific infectious diseases that are screened for vary according to the geographic region and donor lifestyle, but include both vector-borne and non-vector-borne agents (Wardrop et al. 2005; Tasker and Warman 2013). Transmission of Babesia gibsonii and Leishmania spp. via blood transfusion resulting in clinical disease has been reported in dogs (Owens et al. 2001; Stegeman et al. 2003).

Transfusion-related immunomodulation

Transfusion of blood products can induce pro-inflammatory reactions and quell the recipient’s immune response (Izbicki et al. 2004; McMichael et al. 2010; Jackman 2013). As a result, blood transfusions can negatively impact patient outcome. Transfusion-related immunomodulation (TRIM) is a broadly used term encompassing both the transient immunosuppression, as well as the pro-inflamamtory response observed in human patients after blood transfusion (Vamvakas and Blajchman 2007). The exact mechanism causing TRIM is unknown, but is likely to be multifactorial and involve mediators such as WBCs and soluble plasma proteins within RBC units. During storage of RBC units, a more acidic environment develops, leading to WBC degradation over approximately 2 weeks. The WBCs become activated, leading to release of cytokines and eventual WBC death (Vamvakas and Blajchman 2007; Vamvakas 2013). As a result, the supernatant of RBC units becomes rich in pro-inflammatory and procoagulant mediators over time (Sparrow and Patton 2004). Transfusion of donor WBCs might also cause direct suppression of the recipient’s immune system (Bordin et al. 1994). Ultimately, the suspected mediators of TRIM appear to cause a shift in T helper (Th) cell type profile from Th1 to Th2, resulting in suppression of cell-mediated immunity. The functions of natural killer cells and macrophages, as well as other immune system roles, are also suppressed after blood transfusion (Vamvakas 2013).

TRIM has been associated with both beneficial and detrimental patient outcomes. Human patients receiving blood transfusion prior to renal transplants had improved graft survival, a potential benefit of TRIM (Opelz et al. 1973, 1997; Marti et al. 2006; Vamvakas 2013). Suspected detrimental effects of TRIM in human patients include recurrence or metastasis of cancer, increased rate of perioperative and other infections, and increased short-term mortality. However, there is conflicting evidence describing the role of autologous blood transfusion in such adverse sequelae (Vamvakas 1996, 2007; McAlister et al. 1998; Motoyama et al. 2004; Chalfin et al. 2014; Kluth et al. 2014; Rohde et al. 2014; Yeoh et al. 2014;).

Pre-storage leukoreduction appears to mitigate activity of the WBC-derived pro-inflammatory and immunomodulating mediators implicated in TRIM, but does not reduce other soluble plasma factors such as HLA molecules (Sparrow and Patton 2004; Vamvakas 2007, 2013). While leukoreduction was associated with decreased patient mortality in people after cardiac surgery, this strategy has not consistently improved the suspected consequences of TRIM such as increased mortality or post-operative infections in other studies (Hébert et al. 2003; Vamvakas 2007; Englehart et al. 2009).