Serological best practices

The Rh (RH) blood group system consists of 56 antigens located on large, multi-looped glycoprotein products of two closely linked genes, RHD and RHCE.¹

In general, antibodies from the Rh blood group system are potentially clinically significant, even in cases where reports of hemolytic transfusion reactions are rare. If possible, and with the exceptions noted below, antigen-negative donor red blood cells should be provided when the transfusion recipient is shown to have the corresponding antibody.

Anti-C^w

Key points 

• Anti-C^w may be clinically significant.^{1, 2} People with anti-C^w should receive red blood cell units that are crossmatch compatible by IAT at 37°C for transfusion. 
• People with sickle cell disease who have anti-C^w should be provided with C^w-negative red blood cell units for transfusion. 

Background 

C^w (RH8) is a relatively low-frequency antigen,²⁵ occurring at an estimated frequency of 2% in the general white population, and up to 7% to 9% in the Latvian, Laplander and Finnish populations. Its antithetical antigens include C^X (RH9, an abnormal C^w) and the high-incidence MAR (RH51) antigen. 

Anti-C^w was first described in 1946 in a patient with lupus (who also produced anti-Lu^a). It was named for its association with the C antigen and in recognition of the red blood cell donor, Willis. It is a relatively common antibody and usually occurs naturally, although it may be immune stimulated (i.e., by transfusion-related or pregnancy-related red blood cell exposure). Anti-C^w is almost always an IgG antibody, with rare case reports of IgM, and often occurs in association with other antibodies. Due to the location of C^w on the RHCE gene, gene linkage dictates that it almost always coincides with the C antigen (particularly of the DCe [R1] haplotype). When associated with C^w, the C antigen shows weakened expression, hence an allo–anti-C may rarely be made in C+C^w+ individuals. Most, but not all, C-negative red blood cells will also be C^w-negative.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include C^w typing; therefore, requests for C^w-negative red blood cell units result in additional manual phenotyping or genotyping. 

Management: Pre-transfusion and prenatal testing

Unlike many other Rh antigens, in the context of transfusion, anti-C^w is not typically clinically significant. There are no case reports associating anti-C^w with acute or delayed hemolytic transfusion reactions. However, it has been associated with mild to moderate hemolytic disease of the fetus and newborn (HDFN), with rare case reports of severe HDFN, including at least one case of hydrops fetalis.²⁶

When anti-C^w is detected in a patient’s pre-transfusion sample, there is usually no requirement for selection of C^w-negative donor red blood cells. In fact, 98% of donor units are C^w-negative. Routine donor antigen typing (phenotyping) at Canadian Blood Services does not include C^w typing; therefore, a request for C^w-negative red blood cell units results in additional manual phenotyping and/or genotyping. For current and historic anti-C^w, red blood cell units that are crossmatch compatible at the 37°C IAT phase should be selected for transfusion.¹ See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease 

If a person with sickle cell disease develops an anti-C^w antibody, then C^w-negative units should be provided, along with matching for the full phenotype. Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility as possible with the patient’s antigen and antibody profile.^{5, 14}
 

Return to top
 

Anti-V

Key points

• Anti-V is not usually clinically significant. People with anti-V should receive donor red blood cells that are V-negative if possible, and crossmatch compatible by IAT at 37°C.  
• People with sickle cell disease with anti-V should be provided with V-negative red blood cell units for transfusion.

Background

V (RH10) is rare in the white population, occurring at an estimated frequency of 1%, but is very common in people of African ancestry, occurring at an estimated frequency of 30%.²⁷ Though not an antithetical pair, the V antigen is associated with the VS antigen; both are thought to be “uncovered” by a mutation in the transmembrane domain of the RHCE gene product (partial e). Thus, most V+ people are also VS+.

Anti-V was first described in 1955 and is named after the first letter of the last name of the person in which the antibody was initially found.

Most reported anti-V are IgG antibodies that react best in the 37°C IAT phase, but some saline-reactive anti-V have been reported. These frequently occur in association with other antibodies, especially anti-D.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include V typing; therefore, requests for V-negative red blood cell units result in additional genotyping.

Management: Pre-transfusion and prenatal testing

In the context of transfusion, anti-V is not clinically significant. No clinical association has been made between anti-V and acute hemolytic transfusion reactions or HDFN. Nevertheless, when anti-V is detected in a pre-transfusion sample, donor red blood cells that are crossmatch compatible by IAT at 37°C should be selected for transfusion. In addition, because the V antigen is present in a substantial number of particular donor groups (30% of those of African descent), V-negative units should be provided if possible.²⁷ This cautious approach reflects the potential for Rh antibodies to cause hemolysis and the possibility of the transfusion of a V-positive unit, given the prevalence in some donor populations. A request to Canadian Blood Services for V-negative red blood cell units may require genotyping of donor samples to identify antigen-negative red blood cell units. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease

If a person with sickle cell disease develops an anti-V, then V-negative units should be provided, along with matching for the full phenotype, as discussed previously.^{5, 14} Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility as possible with the person’s antigen and antibody profile.

Return to top

There are currently 28 Lutheran antigens found on two red blood cell surface glycoproteins known as Lutheran (Lu) and basal-cell adhesion molecule (B-CAM). Lu/B-CAM is a part of the immunoglobulin superfamily of receptors that differ from one another only in their intracellular domain. They are high-affinity receptors for laminin, an extracellular matrix protein integral to the basement membrane of all cells. Lu/B-CAM is known to be overexpressed on sickle red cells and thought to contribute to circulatory stasis and vaso-occlusive episodes by facilitating erythrocyte adhesion to vascular endothelial cells.⁵

Anti-Lu^a

Key points

• Anti-Lu^a is rarely clinically significant. People with anti-Lu^a should receive red blood cell units that are crossmatch compatible by IAT at 37°C for transfusion.
• People with sickle cell disease who have anti-Lu^a should be provided with Lu^a-negative red blood cell units for transfusion.

Background

Lu^a is an antigen of the Lutheran blood group system. It was first described in 1945 in a patient with lupus erythematosus (who also produced anti-C^w, anti-c and anti-Kp^c). Lu^a is named after the red blood cell donor, although it was later discovered that the donor’s name was actually Lutteran but had been misspelled as Lutheran on the sample tube.^{25, 28}

Among the relatively uncommon antibodies against Lutheran antigens, Anti-Lu^a is the most common, which is why Lu^a positive cells are typically included in screening panels. Anti-Lu^a is usually an immune-stimulated antibody (i.e., stimulated by transfusion-related or pregnancy-related red blood cell exposure) but may also occur naturally, often in association with other antibodies. It is usually an IgM antibody but may have some associated IgG and IgA components. As such, most anti-Lu^a will directly agglutinate red blood cells at room temperature as well as in the 37°C IAT phase. As Lu/B-CAM is heterogeneously expressed on red blood cells, antibodies may display a mixed-field agglutination on testing.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include Lu^a typing; therefore, requests for Lu^a-negative red blood cell units result in additional manual phenotyping or genotyping. 

Management: Pre-transfusion and prenatal testing

In the context of transfusion, anti-Lu^a is rarely clinically significant for most people. It does not contribute to acute hemolytic transfusion reactions and is very rarely associated with mild, delayed hemolytic transfusion reactions and possible mild hemolytic disease of the fetus and newborn (HDFN). With respect to HDFN, fetal Lu^a is present on placental tissues; consequently, adsorption of maternal IgG antibodies onto placental cells prevents antibody transfer to the fetus. This has no apparent impact on pregnancy viability.

When anti-Lu^a is detected in a pre-transfusion sample, there is usually no requirement for selection of Lu^a-negative donor red blood cells. In fact, 92% of donor units are Lu^a-negative. Routine donor antigen typing (phenotyping) at Canadian Blood Services does not include Lu^a typing; therefore, a request for Lu^a-negative red blood cell units usually requires genotyping. For current and historic anti-Lu^a, red blood cell units that are crossmatch compatible in the 37°C IAT phase should be selected for transfusion. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease

If a person with sickle cell disease develops an anti-Lu^a, then Lu^a-negative units should be provided, along with matching for the full phenotype, as discussed previously. Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility with the patient’s antigen and antibody profile as possible.^{5, 14}

Return to top

The Kell antigens are located on the red blood cell transmembrane glycoprotein, CD238, and consist of a large group of 38 antigens. After the ABO and Rh blood group systems, the Kell system contains some of the most immunogenic antigens.

The Kell antigens are expressed early in erythropoiesis. These antibodies have the potential to cause severe hemolytic disease of the fetus and newborn (HDFN) through mechanisms including immune destruction and suppression of erythropoiesis. This leads to the potential for prolonged anemia in the neonatal period for those infants impacted by HDFN due to Kell system antibodies.⁴

Most Kell system antibodies are clinically significant with the potential to cause hemolytic transfusion reactions as well as HDFN and require antigen-negative blood for transfusion.

Anti-Kp^a

Key points

• Anti-Kp^a is rarely clinically significant.^{29, 30} People with anti-Kp^a should receive red blood cell units that are crossmatch compatible by IAT at 37°C for transfusion.
• People with sickle cell disease who have anti-Kp^a should be provided with Kp^a-negative units for transfusion.

Background

Anti-Kp^a is an antibody to an antigen of the Kell blood group system and is extremely rare. It was first identified in 1957 and is named “K” after Kell group (after “Kelleher”, the first producer of an anti-K antibody) and “p” after the name of the first identified anti-Kp^a producer, “Penny”. The International Society of Blood Transfusion nomenclature of Kp^a is KEL3 and is part of a triplet of antithetical antigens (Kp^a, Kp^b and Kp^c).

Kp^a is estimated to be present in only 2% of the white population and is not present in people of African or Japanese ancestry. Anti-Kp^a may be naturally occurring (i.e., arising without stimulus by transfusion-related or pregnancy-related red blood cell exposure), but is more likely to be an alloimmune-stimulated antibody. As such, it is usually an IgG antibody, predominantly IgG1.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include Kp^a typing; therefore, requests for Kp^a-negative red blood cell units result in additional manual phenotyping or genotyping.

Management: Pre-transfusion and prenatal testing

In the context of transfusion, an anti-Kp^a is rarely clinically significant for most people due to the relative rarity of the Kp^a antigen. Due to its low prevalence, there is usually no requirement to select Kp^a-negative donor red blood cells. Instead, red blood cell units that are serologically crossmatch compatible (at 37°C IAT phase) should be selected for transfusion. Routine donor antigen typing (phenotyping) at Canadian Blood Services does not include Kp^a typing; therefore, a request for Kp^a-negative red blood cell units results in additional manual phenotyping or genotyping.

Anti-Kp^a has been rarely reported to cause mild to moderate transfusion reactions, including delayed hemolytic reactions.^{29, 30} As it is expressed on umbilical cord red blood cells, it may also be associated with HDFN. There are only a small number of case reports in which anti-Kp^a has been implicated in a severe reaction. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

In perinatal cases, most guidelines recommend that any antibody to a Kell system antigen (including anti-Kp^a) should be worked up and monitored in the same manner as an anti-K. Traditionally, it has been thought that the severity of HDFN does not correlate well with the titre of Kell system antibodies and early consultation with a high-risk obstetrical provider is recommended when a Kell system antibody is identified during pregnancy. ^{31, 32} Given the paucity of cases implicating anti-Kp^a, it is likely that future recommendations for handling anti-Kp^a will continue to be extrapolated from anti-K.

People with sickle cell disease

If a person with sickle cell disease develops anti-Kp^a, then Kp^a-negative units should be provided, along with matching for the full extended phenotype, as discussed previously.^{5, 14} Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility as possible with the patient’s antigen and antibody profile.

Return to top

Anti-Js^a

Key points

• Anti-Js^a is clinically significant. People with anti-Js^a should receive Js^a-negative blood that is crossmatch compatible by IAT at 37°C for transfusion.
• People with sickle cell disease who have anti-Js^a should be provided with Js^a-negative red blood cell units for transfusion.

Background

Js^a (KEL6) is an antigen of the Kell blood group system and is almost exclusive to people of African ancestry, occurring at a frequency of 20% in this population. It is very rare in white populations (< 0.01%) and absent in people of Asian descent.

Js^a was first described in 1958 and is named after John Sutter, the first producer of the antibody. It was later assigned to the Kell system in 1965.

Anti-Js^a may occur naturally (i.e., arising without stimulus by transfusion-related or pregnancy-related red blood cell exposure) and although this is rare, it is more likely to be an immune-stimulated IgG antibody. Anti-Js^a is associated with rare reports of HDFN^{33, 34}

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include Js^a typing; therefore, requests for Js^a -negative red blood cell units result in additional genotyping. 

Management: Pre-transfusion and prenatal testing

In the context of transfusion, anti-Js^a may be clinically significant and has been implicated in acute and delayed hemolytic transfusion reactions as well as HDFN, which can be severe given the presence of Kell antigens on erythroid precursors. When anti-Js^a is detected in pre-transfusion sample testing (current or historical), it is usually detected in the context of screening for other antibodies since Js^a-positive cells are usually not present on routine screening cells. If anti-Js^a is detected, and if time permits, antigen-negative blood that is serologically crossmatch compatible in the 37°C IAT phase should be provided. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease

If a person with sickle cell disease develops an anti-Js^a, then Js^a-negative units should be provided, along with matching for the full extended phenotype, as discussed previously.^{5, 14} Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility as possible with the patient’s antigen and antibody profile.

Return to top

The Lewis antigens are glycoproteins that are found on the surface of many cells and secreted in various body fluids. As such, Lewis, along with ABO and H are sometimes referred to as “histo-blood groups,” given the fact they are present on many different tissue types.^{2, 4} There are presently six antigens described for this system. Red blood cells acquire these antigens on their membranes by adsorption from circulating antigens. Like the modification of the ABO system, the Lewis precursor oligosaccharide is readily modified by an autosomal dominant-coded fucosyltransferase enzyme (FUT3) as shown in Figure 4. Depending on the combination of Lewis (FUT3) and secretor gene (FUT2), three main phenotypes are possible; Le(a), Le(b) or Le(a-b-). The phenotype Le(a+b+) is also possible but is generally only found in people of East Asian descent who possess a weak secretor phenotype.

Image

Figure 4. Simplified diagram of how the Lewis system is related to the H and ABO. In non-secretors (no FUT2 enzyme) Le^a is formed by addition of a fucose residue to the H precursor. In “secretors” (active FUT2 enzyme) the H antigen in secretions is modified by the FUT3 enzyme to Le^b. Le(a-b-) lack an active FUT3 enzyme and can be secretors or non-secretors depending on FUT2 but is not phenotypically obvious.³⁵

This image is reproduced with permission from John Wiley and Sons with the license number 5973171262380 from the article:

Hu, D.-y., Shao, X.-x., Xu, C.-l., Xia, S.-l., Yu, L.-q., Jiang, L.-j., Jin, J., Lin, X.-q. and Jiang, Y. (2014), FUT2 and FUT3 in Crohn's disease. J Gastroenterol Hepatol, 29: 1778-1785. https://doi.org/10.1111/jgh.12599

Return to top
 

Anti-Lewis

Key points

• Anti-Le^a, anti-Le^b and anti-Le^ab are considered clinically insignificant unless detected at 37ºC.¹ People with anti-Le^a, anti-Le^b and anti-Le^ab should receive red blood cell units that are crossmatch compatible by IAT at 37ºC for transfusion.
• People with sickle cell disease who have anti-Le^a, anti-Le^b or anti-Le^ab should be provided with antigen-negative red blood cell units for transfusion.^{5, 14}

Background

Anti-Le, commonly anti-Le^a, Le^b or Le^ab, are antibodies directed to antigens of the Lewis blood group system. Anti-Le^a, Le^b or Le^ab are fairly common antibodies, and they are usually naturally occurring (i.e., arising without stimulus by transfusion-related or pregnancy-related red blood cell exposure) but can also be immune stimulated. In either case, they are predominantly IgM with some associated IgG components and are most commonly found in people with Lewis-negative (Le(a-b-)) phenotype. Very rarely, anti-Le^a can be found as purely IgG.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include Lewis antigens; therefore, requests for Lewis negative red blood cell units result in additional manual phenotyping.

Management: Pre-transfusion and prenatal testing

In the context of transfusion, anti-Le^a, Le^b or Le^ab are almost always clinically insignificant. Only in rare case reports, and mostly with anti-Le^a, have they been associated in hemolytic transfusion reactions. The reasoning is three-fold: first, due to IgM predominance, the antibodies are generally not active at body temperature; second, as it is also a highly secreted antigen, the donor plasma antigens neutralize recipient antibodies; lastly, transfused cells easily shed their antigen and eventually the donor membrane antigen matches the recipient phenotype.^{4, 36}

Similarly, anti-Le^a, Le^b or Le^ab are not generally associated with hemolytic disease of the fetus and newborn (HDFN). Lewis antibodies are commonly detected in prenatal sera because pregnant persons tend to lose their Lewis antigens during pregnancy, resulting in a temporary Le(a-b-) phenotype and the ability to transiently make anti-Le^a, Le^b or Le^ab until their true phenotype returns, at approximately 6 weeks postpartum. However, these antibodies are predominantly IgM and do not readily cross the placenta. Additionally, while Lewis antigens can be detected in the serum of neonates, they are not expressed on fetal or neonatal red blood cells.

Lewis antibodies are more likely to be IgM and as such will usually react best at room temperature (immediate spin phase). When anti-Le^a, Le^b or Le^ab are detected in a patient’s pre-transfusion sample testing at the 37ºC IAT phase, they have the potential to be clinically significant; however, there is usually no requirement for selection of antigen-negative donor red blood cells. Instead, red blood cell units that are crossmatch compatible at the 37°C IAT phase should be selected for transfusion. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease

If a person with sickle cell disease develops an anti-Le^a, Le^b or Le^ab, then antigen-negative units should be provided, along with matching for the full extended phenotype, as discussed previously.^{5, 14} Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context requires as complete compatibility with the patient’s antigen and antibody profile as possible.

Other disease associations

• Le^b is one of the gastric epithelial receptors for Helicobacter pylori.
• Lewis antigens are expressed in renal tissue, especially on distal tubule epithelium and vessel endothelium. Several studies have implicated Lewis antibodies in renal allograft loss in Le(a-b-) recipients; however, there is insufficient evidence to include Lewis as a routine histocompatibility marker.

Return to top

The Diego antigens are located on the red blood cell membrane glycoprotein AE1 (also known as band 3 or CD233), which plays an essential role in cellular gas exchange and anion equilibrium. There are currently 23 antigens that have been described for this system.^{2, 4}

The distribution of some Diego antigens varies substantially among individuals from different ethnic groups.

Anti-Wr^a

Key points

• Anti-Wr^a is clinically significant. People with anti-Wr^a should receive red blood cell units that are crossmatch compatible by IAT at 37°C for transfusion.
• People with sickle cell disease who have anti-Wr^a should be provided with Wr^a-negative red blood cell units for transfusion.

Background

Wr^a (DI3) is an antigen of the Diego blood group system. It is a low-frequency antigen in all ethnic groups, occurring at a frequency of less than 0.01%.²

Anti-Wr^a was first described in 1953 when it was implicated in a case of hemolytic disease of the fetus and newborn (HDFN). It was later assigned to the Diego blood group system in 1995 and is named after the family in which the antibody was first found.

Anti-Wr^a often occurs naturally (i.e., arising without stimulus by transfusion-related or pregnancy-related red blood cell exposure) or can be immune stimulated. It occurs in up to 2% of blood donors, is frequently found in patients with autoimmune hemolytic anemia, and is often found in association with other antibodies. Anti-Wr^a in healthy donors is predominantly an IgM antibody with or without an associated IgG component. It can also be predominantly IgG, with IgG1 and IgG3 subclasses most often identified in pregnant or previously transfused patients.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include Wr^a typing; therefore, requests for Wr^a -negative red blood cell units result in additional manual phenotyping.

Management: Pre-transfusion and prenatal testing

In the context of transfusion, anti-Wr^a is clinically significant. It is associated with acute and delayed hemolytic transfusion reactions, sometimes severe.^{1, 37} It is also associated with mild to severe HDFN. Nevertheless, anti-Wr^a is not routinely included on antibody screening cells. Despite the relatively common occurrence of the antibody, the combination of a predominantly IgM antibody—which, for the most part, does not react well at body temperatures—with the rarity of the Wr^a antigen on donor cells, makes the probability of incompatible units extremely low, especially with serological crossmatch. Thus, for both current and historically detected anti-Wr^a, red blood cell units that are crossmatch compatible in the 37°C IAT phase should be selected for transfusion. Routine donor antigen typing (phenotyping) at Canadian Blood Services does not include Wr^a typing; therefore, a request for Wr^a-negative red blood cell units results in additional manual phenotyping. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease

If a person with sickle cell disease develops an anti-Wr^a, then Wr^a-negative units should be provided, along with matching for the full extended phenotype.^{5, 14} Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility with the patient’s antigen and antibody profile as possible.

Return to top

Anti-Di^a

Key points

• Anti-Di^a may be clinically significant. People with anti-Di^a should receive red blood cell units that are crossmatch compatible by IAT at 37°C for transfusion.
• People with sickle cell disease who have anti-Di^a should be provided with Di^a-negative red blood cell units for transfusion.

Background

Di^a is an antigen of the Diego blood group system and is a low-frequency antigen, occurring at a frequency of 0.01% in most populations. However, Di^a occurs at a higher incidence among people of South American Indigenous ancestry (2%–54%), Japanese ancestry (12%), North American Chippewa Indigenous ancestry (11%), Chinese ancestry (5%), Hispanic ancestry (1%) and Polish ancestry (0.47%).^{2, 38}

Di^a was first identified in 1955 when it was implicated in a case of HDFN. It is named after Mrs. Diego, the first known producer of anti-Di^a.

Anti-Di^a is typically immune stimulated (i.e., by transfusion-related or pregnancy-related red blood cell exposure) IgG1 plus IgG3 and is best detected by IAT at 37°C. Only rare examples of naturally occurring anti-Di^a have been reported.

Routine donor antigen typing (phenotyping) performed on automation at Canadian Blood Services does not include Di^a typing; therefore, requests for Di^a -negative red blood cell units result in additional manual phenotyping or genotyping.

Management: Pre-transfusion and prenatal testing

In the context of transfusion, anti-Di^a may be clinically significant. It can cause mild to severe HDFN, but there are only infrequent reports of it being clearly implicated in a hemolytic transfusion reaction.³⁷ Given the general rarity of Di^a antigen, for both current and historically detected anti-Di^a, red blood cell units that are crossmatch compatible in the IAT phase at 37°C should be selected for transfusion. Routine donor antigen typing (phenotyping) at Canadian Blood Services does not include Di^a typing; therefore, a request for Di^a-negative red blood cell units results in additional manual phenotyping and/or genotyping. See the summary table for an overview of recommendations for red blood cell transfusion in patients with non-ABO antibodies.

People with sickle cell disease

If a person with sickle cell disease develops an anti-Di^a, then Di^a-negative units should be provided, along with matching for the full extended phenotype.^{5, 14} Given the increased risk of hyperhemolysis in people with sickle cell disease, this clinical context necessitates as complete compatibility with the patient’s antigen and antibody profile as possible.

Return to top

Authors: Youness Elkhalidy, MD, and Gwen Clarke, MD, FRCPC

Original publication date: October 2020

Date of review and update: November 7, 2025

Primary target audiences: Medical laboratory technologists in a hospital laboratory, transfusion medicine physicians

The purpose of this document is to clarify the relevance of ABO subgroups for hospital blood banks selecting a red blood cell unit for transfusion and the importance of working with Canadian Blood Services to investigate potential donors with an ABO subgroup that may cause typing discrepancies.

Key points

• The ABO blood group system includes subgroups with weak expression of A or B antigen on red blood cells.
• Quantitative and/or qualitative differences in ABO subgroups can result in irregularities or discrepancies observed during ABO typing.
• During ABO confirmation testing of a red blood cell unit by the hospital blood bank, a weak reaction may be an indication of an ABO subgroup.
• A weak reaction observed by the hospital blood bank should be confirmed with Canadian Blood Services. If a subgroup has previously been identified through Canadian Blood Services’ routine donor testing, the red blood cell unit is safe to use as labelled; if not previously identified, the unit should not be issued.

What are ABO subgroups?

The A and B blood group antigens are produced by enzymes that modify glycoproteins found on the surface of red blood cells. The precursor glycoprotein to both the A and B antigens is known as the H antigen. As with other blood group systems, the ABO system has variant phenotypes with a genetic basis. The term “subgroup” refers to phenotypes with variations in the structure or number of the A and B antigens, related to genetic variations in the enzymes that produce them. A and B subgroups can result in either of the following:

• reduced numbers of structurally typical A and B antigens
• production of slightly altered A and B antigens, which may or may not be reduced in number

Overall, A subgroups are much more common than B subgroups. Several A subgroups have been described with the A₂ subgroup beingthe most commonly recognized (up to 20% of group A donors are A₂). Some individuals with an A subgroup may form antibodies to the conventional  A antigen (known as A₁ antigen) because of either quantitative (i.e., they produce so little A antigen) or qualitative differences (i.e., A subgroup phenotypes are distinguished from the A₁ phenotype by glycolipid antigen expression).^39,40

How do ABO subgroups affect blood donor testing?

ABO subgroups can result in irregularities or discrepancies observed during ABO typing of a red blood cell unit. For instance, a subgroup that leads to reduced expression of A antigen, such as the A₂ subgroup, may result in a weaker forward reaction during typing. Other variants can result in mixed field reactions, which is the case with the A₃ and B₃ subgroups. A minority of subgroups may result in discrepancies between forward and reverse typing with an unexpected reverse typing, such as an anti-A₁ in a group A₂ individual.

However, it should be noted that many factors other than ABO subgroups can also affect serologic investigations, leading to difficulties in interpretation and ABO discrepancies. For instance, certain disease states can result in altered expression of ABO antigens or reduced production of naturally occurring ABO antibodies. Medication, passive antibodies, or cold-reacting antibodies may also interfere with testing. It is important to consider all factors that could possibly interfere with serological test results to ensure accurate ABO typing of red blood cell units and safe transfusions.

Some of the various ABO subgroups and their effects on forward and reverse grouping are shown in Table 2. Although not included in the table, subgroups can also occur in AB phenotypes. Blood group AB individuals with an associated A or B subgroup will frequently show ABO discrepancies in forward and reverse grouping. A relatively common example is an A₂B person with an anti-A₁, where the forward group shows a reaction with both anti-A and anti-B, but the reverse group has an unexpected reaction with A₁ reagent cells.

Table 2. Various ABO subgroups and the ways they affect red blood cell typing (wk = weak, mf = mixed field)⁴¹

^† Anti-H (a lectin that agglutinates the ABO precursor molecule) can be used to assess the relative “strength” of an A subgroup. The normal A₁ enzyme is very efficient at converting H to A₁ such that very small, undetectable amounts of H are left behind. For A subgroups, the more detectable unaltered H precursor that is left (i.e., stronger anti-H reaction in forward typing), the weaker the A subgroup; more H antigen indicated relative inefficiency in converting H to the A subgroup antigen, or in other words, a “weak” phenotype. This is not applicable for B subgroups as the normal B enzyme is not as efficient at converting the H precursors to the B antigen and usually leaves significant amounts of unaltered H antigens. As such, the presence of anti-H is not useful in distinguishing a typical B type from a B subgroup.

‡ For routine blood bank testing, A₁ cells are the stereotypical cells used in most ABO typing. The presence of agglutination to A₁ cells in the reverse typing indicated that an A subgroup individual is able to make antibodies to the normal A₁ antigen.

^§ Although the A₂ subgroup is considered a weak A subgroup, it usually does not affect the forward typing. Instead, it is most commonly identified via the presence of an unexpected anti-A or anti-A₁.

As part of routine donor typing, Canadian Blood Services investigates all weak or discrepant reactions to resolve the discrepancies and ensure accurate reporting of the ABO blood group. Often subgroup investigations follow identification of weak reactions seen during forward typing. A weak reaction is defined as less than 2+ agglutination either by solid phase or manual methods. As part of routine work-up, weak reactions are enhanced by altering cellular concentration, prolonging the incubation time, and/or decreasing the incubation temperature. These are the “wk” reactions seen in Table 2. Anti-A₁–specific lectins are also utilized to identify the presence or absence of the A₁ antigen. Cells lacking the A₁ antigen are presumed to contain an A subgroup. Mixed field reactions may also help to identify an A₃ or B₃ subgroup (see Table 2). In some cases, further testing at a reference lab is required, including adsorption and elution techniques, flow cytometric immunophenotyping to enumerate the antigens, or sequencing of the ABO genes.

Donors identified with blood groups corresponding to A₂, A₃, or B₃ are labelled as either A or B units depending on the subtype, serologic studies, and presence or absence of an anti-A₁ antibody. If a red blood cell unit with an ABO subgroup is assigned an A/B typing, it is safe to use as labelled (e.g., like any other unit of that A/B typing). If weak or discrepant reactions cannot be resolved, or if an ABO subgroup incompatible with transfusion is identified, the donor will be deferred.

How can hospital blood banks determine if a donor has been evaluated for an ABO subgroup?

As required by the Canadian Standards Association (CSA Z902-25),⁴² hospital blood banks must perform forward typing confirmation of red blood cell units received by Canadian Blood Services in order to allow for release of units via electronic crossmatch. If a weak reaction (< 2+) is identified during ABO confirmation testing, the unit may have an ABO subgroup.

In these cases, it is important to contact Canadian Blood Services distribution to ascertain whether a subgroup was identified and investigated by Canadian Blood Services donor testing during routine typing procedures (Figure 5). A Canadian Blood Services technical specialist can view donor history and confirm whether a subgroup was identified.

If Canadian Blood Services’ records confirm that a subgroup was identified, then the weak reaction seen during a hospital’s confirmation testing is an expected result of the ABO subgroup; the unit is then safe to use according to the ABO type on the end label (e.g., an A₃ unit can be transfused to a group A individual). However, if no history is found, a code will be added to the donor’s file which will cause a manual work-up to be done on the next donation to identify a possible ABO subgroup or other interference. In cases when a weak reaction is unexpectedly observed during confirmation testing (i.e., the subgroup has not been previously identified by Canadian Blood Services), the unit should not be issued. A hospital customer feedback form and return of the unit to Canadian Blood Services may be required and discussion with a hospital transfusion medicine physician is recommended.

Image

Figure 5. Algorithm for hospital blood banks for follow-up of potential ABO subgroup donor units

Author: Matthew Yan, MD, FRCPC 
Original publication date: January 2019
Date of review and update: November 7, 2025
Primary target audience: Medical laboratory technologists in a hospital laboratory, transfusion medicine physicians

Key points

1. Whole blood donations from donors with red cell antibodies can be made into red blood cell units with the antibody specificity listed on the end-label.  These comprise a very small percentage (< 0.5%) of red blood cell units issued by Canadian Blood Services.
2. The transfusion of red blood cell units from donors with red cell antibodies is safe in the adult population given the minimal amount of residual plasma in red blood cell units and dilution effect.

Background 

A very small percentage of red blood cell units (< 0.5%) are manufactured from blood donors with red blood cell antibodies. These donors may have antibodies directed against common non-ABO antigens (e.g., Rh or Kell) detected during routine screening or may represent a rare blood donor with the corresponding antibody. Since 2010, Canadian Blood Services has standardized the issuing of red blood cell units with antibodies to all hospitals it serves (Figure 6). However, given the small percentage of red blood cell units with antibodies, some centres may not have experience in the handling and transfusion of these units. In general, these red blood cell units are considered safe to transfuse for all adult populations, regardless of the patient’s red blood cell phenotype, and should be avoided in pediatric patients.

Image

Figure 6. Example of a product label for a red blood cell unit from a donor with known antibodies. The red box indicates the label location for red blood cell antibody information. For common antibodies, the specificity will be provided (e.g., “contains Anti-K”). For antibodies against rare antigens, “Other-AB” will be printed on the label. If the specificity has not been determined, “UNID-AB” will be printed on the label. 

Safety of units 

Red blood cell units with donor red blood cell antibodies are considered safe for transfusion in the adult population. These units are not typically transfused into the pediatric population as a precaution given the small blood volume of the patients.⁴³ The safety of units is largely due to the small volume of residual plasma present in the red blood cell unit, which is further diluted by additives (e.g., saline, adenine, glucose, mannitol [SAGM]) and by the large plasma volume of the transfusion recipient. Red blood cell units manufactured by Canadian Blood Services typically contain less than 29 mL of residual plasma volume.⁴⁴ This plasma volume is diluted by approximately 110 mL of SAGM additive.⁴⁴ 

To evaluate the effect of additive dilution on passive antibodies, Nobiletti et al.⁴⁵ examined the supernatant of 169 Adsol red blood cell units from donors with known antibodies. They found that 46% of the units had no detectable antibody. In 54% of the units with detectable antibodies, further five-fold dilution of the supernatant rendered the antibody undetectable. Similarly, Hill et al.⁴⁶ examined donor testing samples and red blood cell unit segment samples from 39 donors with known antibodies. There was a median two-tube decrease in the titre of the antibody in the segment sample compared to the donor sample. The antibodies were undetectable in the segment sample in 28% of the cases. 

Another feature that contributes to the safety of these red blood cell units is the low titre of the antibodies in the residual plasma of the units. Hill et al.⁴⁶ reported a median titre of one in the red blood cell unit segments from their study of 39 donors. This is in contrast to anti-A and anti-B titres from group O red blood cell segments, which were found to be much higher with a median titre of 32. Despite having a higher titre, group O red blood cells are commonly transfused to non-O recipients without much concern for hemolysis. This may be in part due to the presence of ABH substances present on tissue and plasma negating the effect of the antibodies, in addition to the mitigating factors already mentioned. 

Theoretical calculation 

Consider the worst-case scenario of a unit from a donor with a high-titre antibody transfused into a small female recipient (i.e., blood volume is smaller than average patient’s blood volume). Assuming the donor has an antibody with a titre of 1024 that is present in the residual plasma at an amount of 29 mL, dilution with 110 mL of SAGM results in at least a two-tube decrease to a titre of 256 (based on the study by Hill et al.⁴⁶). This volume of 139 mL (residual plasma and additive) is then transfused into a 150 cm, 45 kg female. The blood volume of the recipient is approximately 2900 mL. Assuming this is an inappropriate transfusion, and the recipient is not anemic and therefore has a hematocrit of 44%, the plasma volume in the recipient is approximately 1600 mL. This would theoretically result in a further 11-fold dilution of the 139 mL from the red blood cell unit that contained antibodies. In all likelihood, this dilution would render any antibody insignificant. 

Examples in practice 

Since Canadian Blood Services standardized the distribution of red blood cell units from donors with antibodies in 2010, there have not been any reported hemolytic reactions from these units. The few rare case reports of hemolysis from the published literature involved whole blood units, which contain a significant amount of plasma, and differ from current production methods used in Canada. In these reports, inter-donor incompatibility occurred when an antigen-negative recipient received an antigen-positive unit followed by a unit containing the corresponding antibody.^47-50

In the only published report evaluating the practice of transfusing antibody-positive red blood cell units, Combs et al.⁵¹ assessed the clinical impact of 259 red blood cell units from donors with a total of 312 antibodies. Approximately 90% of the units had clinically significant antibodies. Of the 99 recipients with post-transfusion samples, only 10 had detectable passive antibodies, but no evidence of a hemolytic reaction. The authors concluded the practice of transfusing antibody-positive red blood cell units to adult recipients was safe with minimal impacts on the hospital’s workload. This practice allows the continued inclusion of blood donors with antibodies in Canadian Blood Services’ donor pool and thus contributes to the continued sufficiency of the blood supply for Canadians.

Return to top 

^†Patients with sickle cell disease who develop any one of the antibodies listed here should be provided with antigen-negative red blood cell units for transfusion if available within the time required 

^‡The recommendation for antigen negative units as well as crossmatch compatible units is a reflection of the prevalence of the related antigens in some populations, the propensity of the antibodies in these blood group systems to cause hemolysis and other adverse effects, and the likelihood of IgG isotype antibodies that are reactive at 37°C. 

Download the table here: 

Return to top

Courses

For an introduction to immunohematology and the foundations of blood bank compatibility testing, visit LearnSerology.ca, an online educational resource developed by transfusion medicine specialists in Canada. The curriculum consists of six modules and includes an interactive module for completing an antibody investigation panel.

Additional reading

Books

Reid M, Lomas Francis C and Olsson M. The Blood Group Antigens Facts Book, 3rd ed. San Diego: Elsevier Science & Technology; 2012.

Daniels G. Human Blood Groups, 3^rd ed. Oxford: John Wiley & Sons; 2013 

Cohn, C; Delaney, M; Katz, L; and Schwartz, J, editors. Technical Manual, 21st Ed. Bethesda: AABB; 2023. https://www.aabb.org/aabb-store/product/technical-manual-21st-edition---digital-16919069

Guidance documents

National Health Services Blood and Transplant (NHSBT). Spn214/5 – the Clinical Significance of Blood Group Alloantibodies and the Supply of Blood for Transfusion. National Health Services Blood and Transplant (NHSBT) clinical guidelines, National Health Services, 2022: p. 39. https://hospital.blood.co.uk/clinical-guidelines/nhsbt-clinical-guidelines/.

The Canadian Haemoglobinopathy Association. Section 2. Transfusion. Part 1: Disease-Modifying Therapy. Sickle Cell Disease Consensus Statement. Ottawa; 2024. p. 12-20. Available from: https://canhaem.org/wp-content/uploads/2024/09/Transfusion.pdf

Davis BA, Allard S, Qureshi A, Porter JB, Pancham S, Win N, Cho G, Ryan K. Guidelines on Red Cell Transfusion in Sickle Cell Disease. Part I: Principles and Laboratory Aspects. Br J Haematol 2017; 176: 179-91.

Trompeter S, Massey E, Robinson S, Committee of the Task Force of the British Society of Haematology Guidelines. Position Paper on International Collaboration for Transfusion Medicine (ICTM) Guideline ‘Red Blood Cell Specifications for Patients with Hemoglobinopathies: A Systematic Review and Guideline’. Br J Haematol 2020; 189: 424-7. https://onlinelibrary.wiley.com/doi/abs/10.1111/bjh.16405.

We would like to acknowledge the contributions of Danielle Meunier, MD; Sophia Peng, MD; Jacqueline Côté, MLT; and Debra Lane, MD, FRCPC to the first edition of this best practices resource.

1. National Health Service Blood and Transplant. (2022). SPN214/5 – The Clinical Significance of Blood Group Alloantibodies and the Supply of Blood for Transfusion. 1-39. https://hospital.blood.co.uk/clinical-guidelines/nhsbt-clinical-guidelines/ 
2. Reid, M., Lomas Francis, C., & Olsson, M. (2012). The Blood Group Antigens Facts Book. Section II: The blood group systems and antigens; Rh Blood Group System, CW antigen. (3rd ed.). Elsevier Science & Technology. 
3. Daniels, G. (2013). Human Blood Groups. Chapter 3: MNS Blood Group System. 137. 
4. Melland, C., & Palk, M. (2023). Technical Manual. Chapter 12: Other blood group systems and antigens (Cohn, Ed. 21st ed.). Association for the Advancement of Blood & Biotherapies. 
5. Davis, B. A., Allard, S., Qureshi, A., Porter, J. B., Pancham, S., Win, N., Cho, G., & Ryan, K. (2017). Guidelines on red cell transfusion in sickle cell disease. Part I: principles and laboratory aspects. Br J Haematol, 176(2), 179-191. https://doi.org/10.1111/bjh.14346 
6. Morin, P., R, S., & G, C. (2020). Phenotype matching and storage age of blood for sickle cell patients: A review and recommendations for transfusion practice. https://profedu.blood.ca/en/transfusion/publications/phenotype-matching-and-storage-age-blood-sickle-cell-patients-review-and 
7. Yasuda, H., Ohto, H., Nollet, K. E., Kawabata, K., Saito, S., Yagi, Y., Negishi, Y., & Ishida, A. (2014). Hemolytic disease of the fetus and newborn with late-onset anemia due to anti-M: a case report and review of the Japanese literature. Transfus Med Rev, 28(1), 1-6. https://doi.org/10.1016/j.tmrv.2013.10.002 
8. Chapman, J. F., Elliott, C., Knowles, S. M., Milkins, C. E., & Poole, G. D. (2004). Guidelines for compatibility procedures in blood transfusion laboratories. Transfus Med, 14(1), 59-73. https://doi.org/10.1111/j.0958-7578.2004.00482.x 
9. He, Y., Gao, W., Li, Y., Xu, C., & Wang, Q. (2023). A single-center, retrospective analysis of 17 cases of hemolytic disease of the fetus and newborn caused by anti-M antibodies. Transfusion, 63(3), 494-506. https://doi.org/10.1111/trf.17249 
10. Delaney M., J. S., Katz LM., Shwatrz J. (2023). Technical Manual, 21st Edition - Digital. Association for the Advancement of Blood and Biotherapies. 
11. Jain, M. D., Cabrerizo-Sanchez, R., Karkouti, K., Yau, T., Pendergrast, J. M., & Cserti-Gazdewich, C. M. (2013). Seek and you shall find--but then what do you do? Cold agglutinins in cardiopulmonary bypass and a single-center experience with cold agglutinin screening before cardiac surgery. Transfus Med Rev, 27(2), 65-73. https://doi.org/10.1016/j.tmrv.2012.12.001 
12. Jackson, M. E., Grabowska, K., Lieberman, L., Clarke, G., & Yan, M. T. S. (2024). Management of Pregnancies Alloimmunized with Non-Rh and Non-K Alloantibodies. J Obstet Gynaecol Can, 46(1), 102189. https://doi.org/10.1016/j.jogc.2023.07.008 
13. Stetson, B., Scrape, S., & Markham, K. B. (2017). Anti-M Alloimmunization: Management and Outcome at a Single Institution. AJP Rep, 7(4), e205-e210. https://doi.org/10.1055/s-0037-1607028 
14. Canadian Haemoglobinopathy Association. Sickle Cell Disease Consensus Statement. Part I: Disease-modifying therapy. Transfusion. https://canhaem.org/wp-content/uploads/2024/09/Transfusion.pdf 
15. Tippett, P., Reid, M. E., Poole, J., Green, C. A., Daniels, G. L., & Anstee, D. J. (1992). The Miltenberger subsystem: is it obsolescent? Transfus Med Rev, 6(3), 170-182. https://doi.org/10.1016/s0887-7963(92)70167-9 
16. Reid, M. E., & Tippett, P. (1993). Review of a terminology proposed to supersede Miltenberger. Immunohematology, 9(4), 91-95. 
17. Heathcote, D. J., Carroll, T. E., & Flower, R. L. (2011). Sixty years of antibodies to MNS system hybrid glycophorins: what have we learned? Transfus Med Rev, 25(2), 111-124. https://doi.org/10.1016/j.tmrv.2010.11.003 
18. Broadberry, R. E., & Lin, M. (1996). The distribution of the MiIII (Gp.Mur) phenotype among the population of Taiwan. Transfus Med, 6(2), 145-148. https://doi.org/10.1046/j.1365-3148.1996.d01-64.x 
19. Palacajornsuk, P., Nathalang, O., Tantimavanich, S., Bejrachandra, S., & Reid, M. E. (2007). Detection of MNS hybrid molecules in the Thai population using PCR-SSP technique. Transfus Med, 17(3), 169-174. https://doi.org/10.1111/j.1365-3148.2007.00747.x 
20. Mallari, R. A., Chan, A., Powers, R. J., Pandipati, S., Bensing, K. M., Biese, D., & Denomme, G. A. (2020). Fetal inheritance of GP*Mur causing severe HDFN in an unrecognized case of maternal alloimmunization. Transfusion, 60(4), 870-874. https://doi.org/10.1111/trf.15709 
21. Molthan, L. (1981). Intravascular hemolytic transfusion reaction due to anti-Vw+Mia with fatal outcome. Vox Sang, 40(2), 105-108. https://doi.org/10.1111/j.1423-0410.1981.tb00678.x 
22. Agrawal, S., & Chowdhry, M. (2019). A case report on anti-Mia antibody in a multi-transfused patient from India. Transfus Apher Sci, 58(5), 625-627. https://doi.org/10.1016/j.transci.2019.08.027 
23. Pahuja, S., Sehgal, S., Sharma, G., Singh, M., & Yadav, R. (2019). The Anti-Mia Antibody – Report of Four Cases in a Tertiary Care Hospital with Review of Literature. Global Journal of Transfusion Medicine, 4(1), 79-83. https://doi.org/10.4103/gjtm.Gjtm_2_19 
24. Hamilton, J. R. (2020). Low prevalence red blood cell antigens: transfusions, babies, and changing demographics. Transfusion, 60(4), 659-662. https://doi.org/10.1111/trf.15764 
25. Garratty, G., Dzik, W., Issitt, P. D., Lublin, D. M., Reid, M. E., & Zelinski, T. (2000). Terminology for blood group antigens and genes-historical origins and guidelines in the new millennium. Transfusion, 40(4), 477-489. https://doi.org/10.1046/j.1537-2995.2000.40040477.x 
26. Kollamparambil, T. G., Jani, B. R., Aldouri, M., Soe, A., & Ducker, D. A. (2005). Anti-C(w) alloimmunization presenting as hydrops fetalis. Acta Paediatr, 94(4), 499-501. https://doi.org/10.1111/j.1651-2227.2005.tb01924.x 
27. Daniels, G. L., Faas, B. H., Green, C. A., Smart, E., Maaskant-van Wijk, P. A., Avent, N. D., Zondervan, H. A., von dem Borne, A. E., & van der Schoot, C. E. (1998). The VS and V blood group polymorphisms in Africans: a serologic and molecular analysis. Transfusion, 38(10), 951-958. https://doi.org/10.1046/j.1537-2995.1998.381098440860.x 
28. Parsons, S. F., Lee, G., Spring, F. A., Willig, T. N., Peters, L. L., Gimm, J. A., Tanner, M. J., Mohandas, N., Anstee, D. J., & Chasis, J. A. (2001). Lutheran blood group glycoprotein and its newly characterized mouse homologue specifically bind alpha5 chain-containing human laminin with high affinity. Blood, 97(1), 312-320. https://doi.org/10.1182/blood.v97.1.312 
29. Rossi KQ, S. S., Lang C, O’Shaughnessy R. (2013). Severe hemolytic disease of the fetus due to anti-Kpa antibody. Int J Blood Transfus and Immunohaem, 3, 19-22. 
30. Padmore, R., Berardi, P., Erickson, K., Desjardins, D., Giulivi, A., Tokessy, M., Neurath, D., & Saidenberg, E. (2014). Acute extravascular hemolytic transfusion reaction due to anti-Kpa antibody missed by electronic crossmatch. Transfus Apher Sci, 51(2), 168-171. https://doi.org/10.1016/j.transci.2014.08.011 
31. Vlachodimitropoulou, E., Shehata, N., Ryan, G., Clarke, G., & Lieberman, L. (2024). Management of pregnancies with anti-K alloantibodies and the predictive value of anti-K titration testing. Lancet Haematol, 11(11), e873-e877. https://doi.org/10.1016/s2352-3026(24)00239-4 
32. Slootweg, Y. M., Lindenburg, I. T., Koelewijn, J. M., Van Kamp, I. L., Oepkes, D., & De Haas, M. (2018). Predicting anti-Kell-mediated hemolytic disease of the fetus and newborn: diagnostic accuracy of laboratory management. Am J Obstet Gynecol, 219(4), 393.e391-393.e398. https://doi.org/10.1016/j.ajog.2018.07.020 
33. Anderson, R. R., Sosler, S. D., Kovach, J., & DeChristopher, P. J. (1997). Delayed hemolytic transfusion reaction due to anti-Js(a) in an alloimmunized patient with a sickle cell syndrome. Am J Clin Pathol, 108(6), 658-661. https://doi.org/10.1093/ajcp/108.6.658 
34. Levene, C., Rudolphson, Y., & Shechter, Y. (1980). A second case of hemolytic disease of the newborn due to anti-Jsa. Transfusion, 20(6), 714-715. https://doi.org/10.1046/j.1537-2995.1980.20681057162.x 
35. Hu, D. Y., Shao, X. X., Xu, C. L., Xia, S. L., Yu, L. Q., Jiang, L. J., Jin, J., Lin, X. Q., & Jiang, Y. (2014). Associations of FUT2 and FUT3 gene polymorphisms with Crohn's disease in Chinese patients. J Gastroenterol Hepatol, 29(10), 1778-1785. https://doi.org/10.1111/jgh.12599 
36. Milkins, C., Berryman, J., Cantwell, C., Elliott, C., Haggas, R., Jones, J., Rowley, M., Williams, M., & Win, N. (2013). Guidelines for pre-transfusion compatibility procedures in blood transfusion laboratories. British Committee for Standards in Haematology. Transfus Med, 23(1), 3-35. https://doi.org/10.1111/j.1365-3148.2012.01199.x 
37. Bahri, T., de Bruyn, K., Leys, R., & Weerkamp, F. (2018). Fatal Acute Hemolytic Transfusion Reaction due to Anti-Wr(a). Transfus Med Hemother, 45(6), 438-441. https://doi.org/10.1159/000488863 
38. Daniels, G. (2013). Human Blood Groups. Chapter 10: Diego Blood Group System (3rd ed.). Wiley-Blackwell. 
39. Svensson, L., Rydberg, L., De Mattos, L. C., & Henry, S. M. (2009). Blood group A1 and A2 revisited: an immunochemical analysis. Vox Sang, 96(1), 56-61. https://doi.org/10.1111/j.1423-0410.2008.01112.x 
40. Daniels, G. (2013). Human Blood Groups, 3rd edition. Wiley-Blackwell. 
41. Thakral, B., Saluja, K., Bajpai, M., Sharma, R. R., & Marwaha, N. (2005). Importance of Weak ABO Subgroups. Lab Med, 36(1), 32-34. https://doi.org/10.1309/x59taaypepcnbluj 
42. Canadian Standards Association Group. (2020). CAN/CSA-Z902:20 - Blood and blood components. CSA. 
43. Garratty, G. (1998). Problems associated with passively transfused blood group alloantibodies. Am J Clin Pathol, 109(6), 769-777. https://doi.org/10.1093/ajcp/109.6.769 
44. Canadian Blood Services. (2020). Circular of Information for the Use of Human Blood Components: Red Blood Cells, Leukocytes Reduced (LR). https://www.blood.ca/en/hospital-services/products/component-types/circular-information 
45. Nobiletti J, B. S., Cable R, et al. (1998). Unexpected red cell antibodies are not detected in 46% of additive red cells from antibody positive donors. Transfusion, 1998(38), Supplemntal 87S. 
46. Hill, E. A., & Bryant, B. J. (2014). Comparison of antibody titers in donor specimens and associated AS-1 leukoreduced donor units. Transfusion, 54(6), 1580-1584. https://doi.org/10.1111/trf.12486 
47. Abbott, D., & Hussain, S. (1970). Intravascular coagulation due to inter-donor incompatibility. Can Med Assoc J, 103(7), 752-753. 
48. Franciosi, R. A., Awer, E., & Santana, M. (1967). Interdonor incompatibility resulting in anuria. Transfusion, 7(4), 297-298. https://doi.org/10.1111/j.1537-2995.1967.tb05520.x 
49. West, N. C., Jenkins, J. A., Johnston, B. R., & Modi, N. (1986). Interdonor incompatibility due to anti-Kell antibody undetectable by automated antibody screening. Vox Sang, 50(3), 174-176. https://doi.org/10.1111/j.1423-0410.1986.tb04873.x 
50. Zettner, A., & Bove, J. R. (1963). Hemolytic transfusion reaction due to interdonor incompatibility. Transfusion, 3, 48-51. https://doi.org/10.1111/j.1537-2995.1963.tb04603.x 
51. Combs, M. R., Bennett, D. H., & Telen, M. J. (2000). Large-scale use of red blood cell units containing alloantibodies. Immunohematology, 16(3), 120-123.