Blood types (or groups) are determined by specific antigens found on the surface of erythrocytes. In humans, there is the ABO system of blood types, whereas animals have a variety of different blood types. Knowledge of blood types in the different species is important as transfusion of incompatible blood (the donor animal has a different blood type from the recipient animal) can result in severe hemolytic transfusion reactions and even death, in some instances.
There are two types of antibodies to blood group antigens; naturally occurring antibodies and antibodies acquired after exposure to the blood group antigen. Naturally occurring antibodies occur in most species and vary in their pathological significance, i.e. some will not produce a transfusion reaction. Acquired antibodies are produced after exposure to an incompatible blood type, which is from exposure to blood or products containing erythrocytes or their antigens. The most common route of exposure is from previous blood transfusions, however there are less obvious sources of exposure, such as vaccinations that contain foreign red blood cell antigens. Antibodies that are pathogenic (i.e. induce a hemolytic reaction) can cause agglutination and/or hemolysis of red cells.
Blood typing (for the most common blood groups) is offered by a few specialized veterinary diagnostic laboratories (e.g. the Comparative Coagulation Laboratory of the Animal Health Diagnostic Center at Cornell University, the Equine Blood Testing Laboratory in Kentucky, the Stormont Laboratory in California). Ideally, any animal that is routinely used as a blood donor should be blood typed for the most common antigens that produce a hemolytic reaction and (ideally) should be negative for these antigens. Blood type compatibility (or incompatibility) is determined in the laboratory using crossmatching procedures. Since administration of typed negative blood will not prevent a transfusion reaction to less well-characterized red cell antigens, crossmatching should always be performed in an individual that has been previously exposed to blood group antigens.
Canine blood types
There are 8 major blood groups in the dog, labeled as DEA (dog erythrocyte antigen) 1 to 8. These are illustrated in the table below. The major antigens are DEA 1.1 and DEA 1.2. Dogs can be positive for either (not both) DEA 1.1 or 1.2 or are negative for both. Naturally occurring antibodies occur in 20% of DEA 3-negative, 10% of DEA 5-negative, and 20-50% of DEA 7-negative dogs. New blood groups in dogs are being detected, including those in Dalmations, Doberman Pinschers and Shih Tzus, amongst other breeds) (Dal) (Blias et al 2007, Goulet et al 2017) and Kai-1 (IgM) and 2 (IgG) (Euler et al 2016).
Acute hemolytic transfusion reactions only occur in DEA 1.1 and 1.2 negative dogs. As these dogs do not have naturally occurring antibodies, a reaction will only be seen after sensitization of the dog through exposure to DEA 1.1 or 1.2 positive blood (antibody production takes 7-10 days after exposure). The normal lifespan of compatible transfused erythrocytes in dogs is approximately 21 days. In an acute hemolytic transfusion reaction, the lifespan of incompatible transfused erythrocytes ranges from minutes to 12 hours. Although DEA 3-, 5- and 7-negative dogs do have naturally occurring antibodies to DEA 3, 5 and 7 positive red cells, these blood groups do not incite severe hemolytic reactions. Rather, transfusion of incompatible blood is hemolysed more rapidly (within 4 to 5 days) than compatible blood would be (so-called delayed hemolytic reaction). Therefore, crossmatching in dogs does not need to be done on the first transfusion. Neonatal isoerythrolysis has been reported in DEA 1 negative female dogs (previously sensitized to DEA 1 positive cells) mated to DEA 1.1 positive male dogs.
|DEA group||“old” name||Population
|Natural antibody||Transfusion significance|
|1.1||A1||40-60%||No||Acute hemolytic reaction|
|1.2||A2||10-20%||No||Acute hemolytic reaction|
acute hemolytic transfusion reaction
(Melzer et al 2003)
|Kai-1||94%***||Not as yet||Unknown|
|Kai-2||1%***||Not as yet||Unknown|
*Incidence is breed-dependent, e.g., most Greyhounds are negative for DEA 1.1 (explaining their choice as blood donors) but are positive for DEA 3, whilst large numbers of Labrador retrievers are DEA 1.1 positive.
** Of 43-75 dogs tested, 100% of non-Dalmation dogs were positive for Dal, 58% were positive DEA-1 (extended gel), 13% were DEA-3 positive, 100% were DEA-4 positive, and 23% were positive for DEA-7 (Kessler et al 2010). In an extended study of 1130 dogs, there were 85.6-100% Dal+ Dalmations (n=2-90), 43-79% Dal+ Doberman Pinschers (n=14-158), 21-100% Dal+ of Shih Tzus (n=2-12) versus 99-100% of other breeds (Goulet et al 2017). Dal-reactive antibodies were only found in previously transfused dogs, but we have seen cases of transfusion reactions in Dal-negative dogs after the first transfusion. It is unclear how the latter dogs acquired the presumptive anti-Dal antibodies. Due to the low prevalence of Dal-negative dogs, it is difficult to find a compatible donor for a Dal-negative dog with anti-Dal antibodies.
*** Of 503 dogs surveyed (in the same study, 60% were DEA 1 positive) (Euler et al 2016).
Feline blood types
Only 1 blood group system, the AB system, has been identified in cats. In this system, there are 3 blood types; A, B and AB. Similar to humans, the blood group antigens are defined by specific carbohydrates on erythrocyte membranes. A N-glycolyl-neuraminic acid determines the A antigen and a N-acetyl-neuraminic acid determines the B antigen with equal amounts of both acids being found on AB erythrocytes. Cats with blood group B lack a hydroxylase enzyme that converts N-acetyl-neuraminic acid to N-glycolyl-neuraminic acid. The blood group antigens are inherited as a simple autosomal trait with A being dominant over B. The inheritance of the AB allele is, as yet, unknown (it is not due to codominance of A and B). The percentage of cats that are A or B positive is breed-dependent (see table below). The overall incidence of A positive DSH and DLH cats varies between countries, with a higher incidence in the USA (94 to 99%) than in the United Kingdom (87%) and Australia (73%). AB cats are quite uncommon (5% in the United Kingdom and <1% in the USA and Australia).
Cats have naturally occurring antibodies (alloantibodies) which are responsible for potentially life-threatening transfusion reactions. In B cats, the anti-A antibodies are strong agglutinins and hemolysins, especially of the IgM class. In contrast, anti-B antibodies in type A cats are weaker agglutinins and hemolysins (and are of the IgG and IgM class). Type AB cats lack naturally occurring antibodies and can safely receive blood from either type A or B cats (universal recipients).
The half-life of transfused erythrocytes in matched feline transfusions (i.e. type A blood to a type A cat or type B blood to a type B cat) is 29 to 39 days. Transfusion of A blood into a B cat results in rapid destruction of the donated type A blood (mean half-life of 1.3 hours) with severe clinical signs (hypotension, defecation, vomiting, hemoglobinemia, neurologic depression) and even death. In contrast, transfusion of type B blood into A cats produces milder clinical signs and the transfused erythrocytes have a mean half-life of 2.1 days. Due to the presence of these naturally occurring antibodies, cats must be crossmatched before their first transfusion (especially in those breeds with a high incidence of type B or AB blood). In addition, neonatal isoerythrolysis can occur in kittens bearing the A or AB blood group antigen from a mating of B queens to an A or AB tom.
In 2007, a new antigen was detected in cats, Mik, named after the cat in which it was first identified. This antigen has naturally occurring antibodies and can cause a positive crossmatch in a non-transfused cat. Transfusion does incite an antibody response, which can cause an acute hemolytic transfusion reaction (Weinstein et al 2007). Since this time, other antigens have been discovered. These anti-nonAB antibodies (called feline erythrocyte antigens or FEA) are responsible for positive crossmatch reactions in non-transfused cats and can cause delayed hemolytic reactions in previously transfused cats (Binvel et al 2021). The bottom line is that all cats should be crossmatched even if typed as A+.
|Type B frequency||Breeds|
|None||Siamese and related breeds, Burmese, Tonkinese, Russian Blue|
|1-10%||Maine Coone, Norwegian Forest, DSH, DLH|
|11-20%||Abyssinian, Birman, Himalayan, Persian, Somali, Sphinx, Scottish Fold|
|20-45%||Exotic and British Shorthair cats, Cornish and Devon Rex|
|Type AB||DSH, Scottish Fold, Birman, British Shorthair, Somali, Bengal, Abyssinian|
|A||Most DSH and DLH (98%) cats, all Siamese are type A.||In theory, do not need to crossmatch type A cats (or breeds with high likelihood of being cat A) if transfusing with blood from another type A cat.|
|B||Exotic breeds like Himalayan, Abyssinian, Somali, Birman, British shorthair, Devon Rex, and Persian.||Agglutinating antibodies are predominantly IgM. These should be crossmatched before a first transfusion because they are likely to die if transfuse with A blood.|
|AB||Scottish fold, Birman, British Shorthair and DSH.||AB kittens born to a B queen are in danger of neonatal isoerythrolysis, but can safely receive A or B blood.|
|Mik||Was discovered in a cat after a transfusion reaction of compatible A blood.||Pre-existing antibodies exist and will not be detected with typing (can be detected with crossmatching if sufficient antibody titer).|
Equine blood types
There are over 30 blood groups in horses, of which only 8 are major systems. Of these 8, 7 are internationally recognized (A, C, D, K, P, Q and U), whilst the T system is primarily of research interest. Of these, the Aa and Qa are most important for hemolytic reactions, especially neonatal isoerythrolysis (NI) (Becht et al 1983). Other blood groups can occasionally give NI reactions, including Qb, Qc, Db, Ua, Ab and Pa (Zaraby et al 1992, MacLeay 2001, Boyle et al 2005). In addition, all horses lack a unique red cell antigen to donkeys, so they will produce antibodies (and NI) when exposed to donkey blood (such as in mule pregnancies) (McClure et al 1994, Traub-Dargatz et al 1995). Natural antibodies do exist, particularly to Ca antigens, which can cause strong agglutination and hemolytic cross-match reactions (Fenn et al 2020) however the antibodies to Ca do not appear to produce a significant hemolytic reaction in vivo (Bailey et al 1988). The incidence of Aa and Qa is breed-dependent. The table below gives the percentage of animals in the listed breeds that are negative for the factor (Bowling and Clark 1985).
|a Proverbio et al 2020, b Kakoi et al 2021|
Using a biotinylation technique, Owens and colleagues showed that fresh red blood cells in autologous transfusions have a lifespan of approximately 99 days, with a sequential decreased lifespan with duration of cold storage. A much shorter mean lifespan of 39 days is seen with transfusions of fresh allogeneic blood that is crossmatch and blood type compatible (Mudge et al 2012). In contrast, transfusion of crossmatch incompatible allogeneic blood into adult horses (these were all anti-Ca reactions, based on a personal communication with Dr. Tomlinson) results in much shorter red blood cell lifespans, ranging from 5 to 11 days for crossmatches showing ≥2+ or ≥1+ agglutination reactions, respectively (1+ = 3-5 clumps, 2+ = small and large clumps by microscopic assessment) in adult horses (Tomlinson et al 2015). The half-life after transfusion of blood into horses with anti-Aa or anti-Qa antibodies is not known.
Ruminant blood types
- Cattle: There are 11 major blood group systems in cattle, A, B, C, F, J, L, M, R, S, T and Z. The B group has over 60 different antigens, making it difficult to closely match donor and recipient. The J antigen is a lipid that is found in body fluids and is adsorbed onto erythrocytes (therefore, it is not a “true” antigen). Newborn calves lack this antigen, acquiring it in the first 6 months of life. Some animals have only a small amount of J antigen on erythrocytes and none in serum; these so-called “J-negative” animals can develop antibodies against the J-antigen and develop transfusion reactions if transfused with J-positive blood. Neonatal isoerythrolysis is not a naturally occurring phenomenon in cattle. Bouts of NI have occurred secondary to blood-derived vaccines (e.g. against anaplasmosis, babesiosis). The most common antigens that cattle were sensitized to were the A and F systems.
- Sheep: Seven blood group systems have been identified in sheep (A, B, C, D, M, R and X). Similar to cattle, the B system is highly polymorphic. The R system is similar to the J system in cattle, in that the antigen is soluble. The M-L system is involved in active red cell potassium transport and polymorphisms in this system result in breeds of sheep with varying erythrocyte potassium content. Neonatal isoerythrolysis has been reported in lambs administered bovine colostrum. This is due to the presence of antibodies to sheep erythrocytes in bovine colostrum (called “heterophile” antibodies), which is a common occurrence. They are antibodies produced to common cross-reactive antigens present on the surface of bacteria and protozoa that are identical to epitopes on blood group antigens.
- Goats: Blood group antigens in goats are similar to those in sheep and the same reagents are used to type both species. Five major systems have been identified in goats; A, B, C, M and J (the latter is also a soluble antigen like in cattle).