Serological weak D phenotypes: a review and guidance for interpreting the RhD blood type using the RHD genotype

Transcription

1 review Serological weak D phenotypes: a review and guidance for interpreting the RhD blood type using the RHD genotype S. Gerald Sandler, 1 Leonard N. Chen 1 and Willy A. Flegel 1,2 1 Department of Pathology and Laboratory Medicine, MedStar Georgetown University Hospital, Washington, DC, and 2 Department of Transfusion Medicine, NIH Clinical Center, National Institutes of Health, Bethesda, MD, USA Summary Approximately 02 1% of routine RhD blood typings result in a serological weak D phenotype. For more than 50 years, serological weak D phenotypes have been managed by policies to protect RhD-negative women of child-bearing potential from exposure to weak D antigens. Typically, blood donors with a serological weak D phenotype have been managed as RhD-positive, in contrast to transfusion recipients and pregnant women, who have been managed as RhD-negative. Most serological weak D phenotypes in Caucasians express molecularly defined weak D types 1, 2 or 3 and can be managed safely as RhD-positive, eliminating unnecessary injections of Rh immune globulin and conserving limited supplies of RhD-negative RBCs. If laboratories in the UK, Ireland and other European countries validated the use of potent anti-d reagents to result in weak D types 1, 2 and 3 typing initially as RhD-positive, such laboratory results would not require further testing. When serological weak D phenotypes are detected, laboratories should complete RhD testing by determining RHD genotypes (internally or by referral). Individuals with a serological weak D phenotype should be managed as RhD-positive or RhD-negative, according to their RHD genotype. Keywords: serological weak D phenotype, partial D phenotype, RhD blood group, RHD gene, blood transfusion. Since the identification of the Rh factor more than 7 decades ago (Levine & Stetson, 1939; Landsteiner & Weiner, 1940), recipients of blood transfusions and blood donors have been categorized as either RhD-positive [D+ red blood cells (RBCs)] or RhD-negative (D ) RBCs. In 1946, the first D variant antigen was reported, that is, RBCs that did not agglutinate when RhD typed by certain anti-d sera, but did agglutinate when typed with other anti-d sera (Stratton, Correspondence: Dr. S. Gerald Sandler, Transfusion Service, MedStar Georgetown University Hospital, 3800 Reservoir Road, N.W., Washington, DC 20007, USA. sandlerg@gunet.georgetown.edu 1946). Stratton named these D variants D U. Subsequently, case reports revealed that some women with a D U phenotype who had been exposed to D+ RBCs by transfusion or pregnancy formed anti-d (Argall et al, 1953; Simmons & Krieger, 1960; Ostgard et al, 1986; Mayne et al, 1991; Domen & Hoetge, 1997). Additional pregnancies were reported that were complicated by RhD haemolytic disease of the fetus and newborn (Hill et al, 1974; Lacey et al, 1983; White et al, 1983; Cannon et al, 2003). To protect RhD-negative women from exposure to the D antigen and forming anti-d (RhD alloimmunization) by transfusion of RBCs from a donor with a D U variant antigen, policies were developed in the United States requiring RBCs from blood donors who tested initially negative by anti-d to be retested with antiglobulin (a weak D test ) (Scientific Committee of the Joint Blood Council & Standards Committee of the American Association of Blood Banks, 1958). If RBCs agglutinated after addition of antiglobulin to anti-d typing, the RBCs were interpreted to be D+. If RBCs did not agglutinate after addition of antiglobulin, they were interpreted to be D. Of the five Blood Establishments in the UK and Ireland, only the Northern Ireland Blood Transfusion Service uses a weak D test for blood donors. The Irish Blood Transfusion Service does perform a weak D test on RhD-negative blood donors who are C+/E+. In the United States, a weak D test was not required for pregnant women or transfusion recipients. If RBCs from a pregnant woman or transfusion recipient typed negative by initial anti-d testing and a weak D test was not performed, the individual s RhD type was interpreted to be RhD-negative to ensure that the individual was not inadvertently exposed to D+ or D variant RBCs. In recent years, molecular laboratory methods have been developed that separate D variant antigens into three groups, namely, molecularly defined weak D phenotypes, partial D phenotypes and DEL phenotypes. The following review is intended as a guide for managing blood transfusions and Rh immunoprophylaxis for pregnant women whose RhD type has been reported by the laboratory as a serological weak D phenotype. Our intent is to review the molecular science determining D variant RBC antigens and to provide guidance for managing patients with a serological weak D phenotype based on the individual s RHD genotype. First published online 16 May 2017 doi: /bjh ª 2017 John Wiley & Sons Ltd

2 Clinical importance of the RhD blood group antigen From a clinical perspective, the Rh blood group system is the most important of the 36 blood group systems (Storry et al, 2016) after ABO. Among the 54 blood group antigens in the Rh system, the RhD antigen is the most immunogenic and important in clinical practice (Gunson et al, 1976; Urbaniak & Robertson, 1981; Storry et al, 2014). The consequence of an RhD-negative woman forming anti-d is that any subsequent pregnancy involving an RhD-positive fetus is at risk for morbidity and mortality associated with Rh haemolytic disease of the fetus and newborn. The consequence of any RhD-negative individual forming anti-d is that the option of transfusing D+ RBCs in an emergency is eliminated and there is now an absolute lifetime requirement for transfusing only D RBCs. Report of the first D variant antigen (D U ) and subsequent changes in terminology Stratton (1946) reported that RBCs from a blood donor at the Manchester Royal Infirmary failed to agglutinate with 20 anti- D sera, but agglutinated variably with 12 other anti-d sera. Stratton determined that the donor had inherited a D variant antigen from his father, as did his two brothers. He reported the D variant as a new Rh allelomorph and named it D U. As more sensitive RhD typing methods were developed, it became apparent that an individual s RBCs that typed as D U by a relatively insensitive manual tube method may be typed as a straightforward D+ (RhD-positive) if retested by a potent monoclonal anti-d reagent, or by a more sensitive manual or an automated RhD typing method (Agre et al, 1992). In 1992, Ortho Diagnostics (Raritan, NJ) convened a meeting of immunohaematologists to consider the impact of these more sensitive RhD typing methods on the terminology for D U RBCs. The participants cited D negative, D U positive, Rh negative, D U positive, D negative, D U positive, D U phenotype, Rh positive D U and others as examples of different names used to describe a serological weak D phenotype. The outcome of the meeting was a Letter to the Editor of Transfusion, recommending elimination of the term D U and wider use of the designation weak D for all weak expressions of the D antigen [to] eliminate much of the confusion caused by inconsistent and sometimes erroneous terminology (Agre et al, 1992). Within a decade, new laboratory methods were developed capable of determining RHD genotypes that were expressed as a serological weak D phenotype (Flegel et al, 1998; Wagner et al, 1999). Molecular scientists began to report RHD genotypes as weak D type 1, weak D type 2, and so forth, introducing terminology for RHD genotyping that could be confused easily with the weak D designation intended by Agre et al (1992) for serologically-determined weak expression of the D antigen. In 2015, the AABB (formerly, American Association of Blood Banks) and the College of American Pathologists (CAP) convened a Work Group on RHD Genotyping and charged it with developing recommendations to clarify clinical issues related to RhD typing (Sandler et al, 2015). The Work Group published its recommendations using the term serologic weak D phenotype to distinguish the result of serological weak D testing using anti-human globulin in clinical laboratories versus the results of RHD genotyping for weak D types based on molecular methods. A laboratory report of a serological weak D phenotype reflects sensitivity of the laboratory s method, as well as the molecular basis for the weak D antigen In the United States, a serological weak D phenotype is usually defined as reactivity of RBCs with an anti-d reagent giving no or weak ( 2+) reactivity in initial testing, but agglutinating moderately or strongly with anti-human globulin (a weak D test) (Jenkins et al, 2005; Daniels, 2013; Sandler et al, 2015). In Europe, such a weak D test is often understood to be what a reference laboratory does to resolve equivocal serological reactivity, whether by serological or molecular methods. Grading a serological reaction as 2+ is often subjective and there is a lack of consensus for the definition. In the UK and Ireland, most clinical laboratories use potent anti-d reagents and few use indirect antiglobulinreactive anti-d reagents. Any serological reaction of 2+ is likely to be referred to a Red Cell Immunohaematology (RCI) Laboratory. Typically, RCI Laboratories use the Quotient Advanced Partial RhD Typing Kit (Quotient/Alba Bioscience Limited, Edinburgh, Scotland), comprising of 12 9 IgG anti-d reagents, to identify serological weak D phenotypes. RCI Laboratories in the UK do not currently perform RHD genotyping for these patients samples. In Ireland, RHD genotyping for serological weak D phenotypes commenced in The observed prevalence of serological weak D phenotypes increases when the laboratory method of detection is relatively insensitive, for example, manual tube testing. In this situation, the first phase of anti-d detection may be too insensitive to agglutinate RBCs with a D variant antigen, but the RBCs will be agglutinated by the second weak D test phase and, therefore, the sample is interpreted to be a serological weak D phenotype. In contrast, the observed prevalence of serological weak D phenotypes decreases when the method of detection is more sensitive, for example, an automated gel column or solid-phase analyser using a blend of potent recombinant monoclonal anti-d reagents. In this situation, the highly sensitive first phase of D antigen testing by anti-d will agglutinate RBCs expressing a weak D antigen and the sample is interpreted to be a conventional RhD-positive. Thus, a blood sample from a patient or blood donor with a weak D variant antigen may be interpreted to be RhD-positive when tested by a laboratory using ª 2017 John Wiley & Sons Ltd 11

3 Table I. Prevalence of serological weak D phenotypes in different populations and by different laboratory methods. Population n Prevalence Method Reference Donors (France) Donors (United States) Donors (D C+ and/or E+) (England) Donors (D C E ) (England) Donors (Netherlands) Patients (D ) (Canada) Donors (United States) Prenatal (United States) Pregnant women (Croatia) % Groupamatic Garretta et al (1974) % Manual tube Stroup Walters (1988) % Groupamatic Contreras and Knight (1989) % Groupamatic Contreras and Knight (1989) % Manual tube Van Rhenen et al (1989) % Manual tube Denomme et al (2005) % Olympus PK7200 Jenkins et al (2005) % Manual tube Wang et al (2010) % Manual tube Monoclonal anti-d Lukacevic Krstic et al (2016) a sensitive RhD typing method. However, the same sample may be interpreted to express a serological weak D phenotype if tested by the same or a different laboratory using a less potent anti-d reagent. The prevalence of serological weak D phenotypes also varies by race and ethnicity (Table I). An estimated 02 10% of Caucasians inherit an RHD genotype that codes for a serological weak D phenotype (Garratty, 2005). In North London, the prevalence of weak D phenotypes was estimated to be 03% for white and 17% for black blood donors (Contreras & Knight, 1991). A serological weak D phenotype is the expression of an amino acid substitution in the RhD protein or an RHD-RHCE-D gene conversion causing a D variant antigen The molecular basis for a serological weak D phenotype can be determined by retesting the blood sample by one of several molecular methods to identify the underlying mutation or recombination (Monteiro et al, 2011; Tilley & Grimsley, 2014). The most common D variant antigen identified when a serological weak D phenotype is detected in a Caucasian is a molecularly defined weak D type. Less commonly, a serological weak D phenotype is associated with a partial D phenotype, which is a D variant most frequently expressed as D+, but may occasionally present as a serological weak D phenotype. A third category of D variants, DEL phenotypes, is included in this review for the purpose of a comprehensive overview of D variants, but the expression of the D antigen in DEL phenotypes is too weak to be detected as a serological weak D phenotype. Molecular studies of D antigens in different populations reveal a significant number of D variant alleles among individuals who type as RhD-negative by routine serological methods. A study of hospital patients in Toronto, Canada revealed that at least 096% of RhD-negative patients expressed an RHD variant allele (Denomme et al, 2005). A study of RhD-negative pregnant Dutch women detected 096% with a D variant allele (Stegmann et al, 2016). The prevalences of RHD alleles in studies conducted among RhD-negative donors in Europe, East Asia, South America and North Africa are summarized in a review (Denomme, 2013). Weak D types A molecularly defined weak D type is a variant of the RhD protein with an amino acid substitution in the trans-membranous or intracellular segment and expresses a decreased quantity of D antigen (Wagner et al, 1999; Flegel et al, 2007) (Fig 1). Most serological weak D phenotypes (>95% in Northern Europeans) are the expression of weak D types 1, 2, 3 or 40/41 (Flegel, 2011). To date, 147 weak D types have been listed on the Rhesus database ( info/) (Table II). Partial D phenotypes Partial D phenotypes were initially described as blood factors (Unger et al, 1959), then as mosaics (Tippett & Sanger, 1962; Weiner & Unger, 1962). Salmon et al (1984) introduced the term partial D (Issitt & Telen,1996). Molecular studies have determined that RBCs expressing a partial D phenotype have an amino acid substitution in at least one of the extracellular or RBC membrane surface loops (Wagner et al, 1999; Flegel et al, 2007) (Fig 1). Most RBCs express a partial D phenotype as D+ by routine serological methods 12 ª 2017 John Wiley & Sons Ltd

4 Fig 1. The D antigen in the red cell membrane. The RhD protein consists of 417 amino acids (circles). All amino acid positions involved in the known, molecularly-defined weak D types are marked: weak D type 1, 2 and 3 (red), weak D type 4 cluster (yellow) and the additional 11 weak D types (orange) of the original description by Wagner et al (1999). Many more amino acid substitutions have since been identified in one (grey) or several weak D types (blue) causing a serological weak D phenotype. Also, five rare subtypes of weak D type 1, 2 or 3 have been characterized that carry one additional amino acid substitution each (red ring). There are nine exon boundaries in the RHD cdna (black bars), as reflected in the amino acid sequence (Flegel, 2011). Amino acid no. 1 is lacking from the mature protein in the red cell membrane, and the arch depicts the Rh protein vestibule (modified from Flegel, 2006). and are not detected as serological weak D phenotypes. They are not detected by routine RhD typing unless the individual has been exposed to D+ RBCs and formed anti-d (Westhoff, 2005). Occasionally, partial D RBCs have decreased expression of the D antigen and are detected as a serological weak D phenotype by routine RhD typing (Westhoff, 2005; Stegmann et al, 2016). Approximately 5 10% of weak D phenotypes in the United States are estimated to be partial D phenotypes (Garratty, 2005). There are 105 partial D types listed by the Rhesus database ( Partial D types are also separated into D categories, of which DVI is the most common and most likely to be associated with formation of anti-d in Caucasian populations. The prevalence of DVI in South-western Germany among more than blood donors was 002% (Wagner et al, 1995). In South-western England, 50% of blood donors classified as D U were found to have the category DVI phenotype (Leader et al, 1990). In the United States, monoclonal anti-d blood typing reagents are selected to avoid detection of DVI RBCs (Wagner et al, 1995; Judd et al, 2005). The result is that transfusion recipients with a partial DVI phenotype, who would otherwise be routinely typed as RhD-positive, will be typed as RhD-negative. While this strategy protects DVI transfusion recipients from receiving transfusions of potentially immunogenic D+ RBCs, the practice does not protect transfusions recipients with other partial D types from alloimmunization by transfusion of random (wild-type) D+ RBCs (Von Zabern et al, 2013). For example, transfusion recipients who have inherited a DIV partial D phenotype are likely to be typed as RhD-positive and are at risk of alloimmunization by random D+ RBCs (Von Zabern et al, 2013). DEL phenotypes DEL variant antigens (formerly, D el ) express a D antigen that is too weak to be detected by routine serological methods as D+ or as a serological weak D phenotype. DEL variant antigens were first detected by adsorption of anti-d and elution (Okubo et al, 1984, 1991). DEL variant RBCs from blood donors routinely type as D (anti-d and weak D test negative), but transfusion of DEL variant RBCs to RhD-negative recipients has been reported to stimulate formation of anti-d (Yasuda et al, 2005; Kim et al, 2009; Shao, 2010; Yang et al, 2015). Transfusion recipients with a complete DEL phenotype and an RHD (1227G>A) allele (Asian-type DEL) are not at risk of forming anti-d following transfusion of D+ RBCs (Wang et al, 2014). Pregnant women with a complete DEL phenotype who deliver an RhD-positive newborn are not at risk for forming anti-d, but pregnant women with certain partial or hybrid DEL alleles are at risk for forming anti-d. There are significant differences in the prevalence of DEL phenotypes and the ª 2017 John Wiley & Sons Ltd 13

5 Table II. Interpreting results of RHD genotyping for clinical practice. Prevalence among serological weak D phenotypes in various populations Clinical practice per AABB-CAP Work Group* Weak D type Caucasians Africans Asians Recommendation Anti-D reports Evidence Reference Type 1, 2, 3 >90% Rare Rare RhD positive No reports Strong Wagner et al (2000); Flegel (2006) Type 40, 41, 43 <2% Common Not reported RhD negative Under investigation Precautionary Wagner et al (2000); Polin et al (2007); Sandler et al (2015) Type 42 (DAR) Rare Common Rare RhD negative Many reports Strong Hemker et al (1999); Wagner et al (2000) Type 11 <1% Not reported Not reported RhD negative Single report Precautionary Flegel (2006) Type 15 <1% Not reported Common RhD negative Single report Precautionary Wagner et al (2000); Luettringhaus et al (2006) Type 21 3 reports Not reported Not reported RhD negative Single report Precautionary M uller et al (2001); Polin et al (2007); McGann and Wenk (2010) Type 57 1 report Not reported Not reported RhD negative Single report Precautionary Le Marechal et al (2007) All other weak D types combined <5% Rare Rare RhD negative Not reported Precautionary Flegel (2006) *Following a precautionary strategy, the AABB (formerly, American Association of Blood Banks) and the College of American Pathologists (CAP) Work Group limited its recommendation to weak D types 1, 2, and 3 for managing serological weak D phenotypes as RhD-positive until more data are available (Sandler et al, 2015). RHD (1227G>A) allele (Wagner et al, 2001) among RhDnegative individuals in different racial and ethnic populations. Less than 1% of Chinese Han are RhD-negative (Yang et al, 2007; Gu et al, 2014) and of these, as many as 30% express the DEL phenotype (Shao et al, 2002; Wagner et al, 2005). In Japanese, 05% are RhD-negative (Okubo et al, 1991) and 28% of these express a DEL phenotype (Fukumori et al, 1997). In Koreans, 015% are RhD-negative and of these, 17% express a DEL phenotype (Kim et al, 2005; Luettringhaus et al, 2006). The prevalence of DEL phenotypes is significantly less in Caucasians, of whom approximately 15% are RhD-negative and only 01% of these express a DEL phenotype (Flegel et al, 2009). Among the 3 5% of African Americans who are RhD-negative, there are no reports of DEL phenotypes (Daniels, 2002). When are serological weak D phenotypes detected? Most serological weak D phenotypes are detected when a pregnant woman, potential transfusion recipient or blood donor has a blood sample routinely typed for RhD and the grade of RBC agglutination is weaker ( 2+) than expected for RhD typing using potent anti-d reagents (3+ to 4+). Also, serological weak D phenotypes are detected when a clinical laboratory types a blood sample as D+, but the laboratory s record of a prior RhD type is D. The discrepancy may reflect an error in patient or sample identification. Alternatively, the discrepancy may reflect the increased potency of new monoclonal RhD typing reagents in the laboratory compared to previously used plasma-derived anti- D reagents that were less effective for detecting weakly expressed D variant antigens. Applying RHD genotyping results in clinical practice In the United States, the longstanding laboratory practice of not performing a weak D test for transfusion recipients and pregnant women and/or managing serological weak D phenotypes as RhD-negative has proven to be a safe strategy in that it protects susceptible individuals from RhD alloimmunization and forming anti-d. However, that practice results in unnecessary transfusion of difficult-to-obtain D- RBCs for many transfusion recipients and unnecessary injections of Rh immune globulin for many pregnant women (Sandler et al, 2015). The following section describes how RHD genotyping transfusion recipients with a serological weak D phenotype can conserve inventories of RhD-negative RBCs without compromising transfusion safety. Also, RHD genotyping pregnant women when a serological weak D is detected can avoid unnecessary injections of Rh immune globulin without compromising the safety of their pregnancy or the fetus. Blood donors and transfusion recipients Following recognition that RhD haemolytic disease of the fetus and newborn was the result of RhD-negative mothers forming anti-d, clinical practice guidelines required that 14 ª 2017 John Wiley & Sons Ltd

6 RhD-negative transfusion recipients, particularly women of childbearing potential, receive only D RBCs when transfused. In 1958, the first guideline for managing transfusion recipients or blood donors with a serological weak D phenotype was published in the first edition of AABB Standards (Scientific Committee of the Joint Blood Council & Standards Committee of the American Association of Blood Banks, 1958). That initial guidance for transfusion practice required a weak D test if a blood donor s RBCs typed as D by direct agglutination using an anti-d reagent, but regarded a direct agglutination method to be sufficient for RhD typing of transfusion recipients RBCs. That guidance has remained unchanged for more than 50 years. The current 30th edition of Standards requires a weak D test for blood donors, thereby protecting RhD-negative transfusion recipients from inadvertent exposure by transfusion to potentially immunogenic RBCs with a serological weak D phenotype (Ooley, 2015). In contrast, the current 30th edition of Standards considers a weak D test for transfusion recipients to be optional, resulting in most recipients with a serological weak D phenotype being categorized as RhD-negative, protecting them from inadvertent exposure to RBCs that are either D+ or express a serological weak D phenotype (Ooley, 2015). In 2014, CAP conducted a survey of policies and practices for testing serological weak D phenotypes and administration of Rh immune globulin involving more than 3100 laboratories in the United States (Sandler et al, 2014a). This survey revealed that there was a lack of standard practice for interpreting the RhD type when a serological weak D phenotype was detected. Observational studies in central Europe indicate that transfusion recipients with a weak D type 1, 2 or 3 in the homozygous or hemizygous state are not at risk for forming alloanti-d when exposed to D+ RBCs (Wagner et al, 2000; Flegel, 2006). Approximately 90% of Caucasians in central Europe with a serological weak D phenotype have a weak D type 1, 2 or 3 and can be managed safely as RhDpositive (Flegel, 2007). The AABB-CAP Work Group recommended that RHD genotyping be performed for transfusion recipients when a serological weak D phenotype is detected by routine RhD typing. Those patients whose serological weak D phenotype is associated with a molecularly defined weak D type 1, 2 or 3 may be transfused safely with D+ RBCs (Fig 2). Although an automated DNA extraction system can extract DNA from whole blood in less than one hour, turnaround times increase if the laboratory uses manual methods for the extraction. Currently, in the United States, most RHD genotyping is performed in reference laboratories and, therefore, the turnaround time is more than 1 day, excluding the procedure for patients requiring an urgent transfusion. For patients requiring chronic transfusions, for example, sickle cell disease, thalassaemia and myelodysplastic syndrome, the results of once-in-a-lifetime RHD genotyping may not be available in time for the current transfusion, but would be available for future transfusions (Chou et al, 2013; Fasano & Chou, 2016). New methods for blood group genotyping, for example, direct polymerase chain reaction (PCR) amplification without DNA extraction, offer the promise of reducing the time for RHD genotyping Result of RhD Typing by Manual Tube or Automated Methods Negative Candidate for RhIG RhD-negative for transfusion! Discrepant/inconclusive or strength of reaction weaker than expected (serological weak D phenotype) Send for RHD genotyping for weak D types Positive (and concordant with patient history, if available) Not a candidate for RhIG RhD-positive for transfusion Weak D type 1, 2 or 3 Not detected Weak D type 1, 2 or 3 Detected May be at risk for forming anti-d Candidate for RhIG RhD-negative for transfusion Not at risk for forming anti-d Not a candidate for RhIG RhD-positive for transfusion Fig 2. Flow diagram for managing a laboratory result of a serological weak D phenotype. If the report of routine RhD typing is RhD-negative, the individual should be managed as RhD-negative; and if RhD-positive, the individual should be managed as RhD-positive. If the result of RhD typing is a serological weak D phenotype, the laboratory should retest the blood sample or refer it to a reference laboratory for RHD genotyping. If a weak D type 1, 2 or 3 is detected, the individual should be managed as RhD-negative. If a weak D type 1, 2 or 3 is detected, the individual can be managed safely as RhD-positive. (Reproduced from Sandler, S.G., Flegel, W.A., Westhoff, C.M., Denomme, G.A., Delaney, M., Keller, M.A., Johnson, S.T., Katz, L., Queenan, J.T., Vassallo, R.R. & Simon, C.D. (2015) It s time to phase in RHD genotyping for patients with a serologic weak D phenotype. Transfusion, 55, , with permission of Wiley Periodicals, Inc.) ª 2017 John Wiley & Sons Ltd 15

7 to minutes, making RHD genotyping feasible for real-time application in the hospital (Wagner et al, 2017). Patients with sickle cell disease benefit from RHD genotyping, not only because most have a requirement for chronic transfusion, but also because they are at increased risk of alloimmunization to certain Rh and other blood group antigens because of the differences that they inherit from their African ancestry compared to those inherited by the predominately Caucasian donors in Western countries (Vichinsky et al, 1990). Pregnant women In the United States, guidelines for managing pregnant women with a serological weak D phenotype were first introduced in 1981 (Oberman, 1981). The American Association of Blood Banks issued a standard that a woman s candidacy for Rh immune globulin should be determined by the same laboratory method as that for RhD typing of blood donors (Oberman, 1981). Thus, women with a serological weak D phenotype were categorized as RhD-positive and not considered candidates for Rh immunoprophylaxis with Rh immune globulin. Within a few years, there were reports of women with a serological weak D who delivered an RhD-positive newborn, did not receive Rh immune globulin, and who formed anti-d (White et al, 1983; Ostgard et al, 1986; Mayne et al, 1991; Domen & Hoetge, 1997). Currently, in the UK and Ireland, women with a serological weak D phenotype are often managed as RhD-positive. Although most of these women would be determined to be RhD-positive if RHD genotyped, a minority will have an RHD type other than 1, 2 or 3 that would qualify them as candidates for Rh immune globulin. The AABB standard was revised in the current 30th edition of AABB s Standards which determines a pregnant woman s candidacy for Rh immune globulin using the same RhD typing method for as that for a transfusion recipient, that is, the woman s anti-d typing of RBCs is negative and the test for weak D is optional (Ooley, 2015). A survey of hospital practice in the United States by CAP in 2014 revealed that only 198% of responding laboratories performed a weak D test when a patient s RBCs typed negative by the initial anti-d test (Sandler et al, 2014a). Thus, most pregnant women in the United States with a serological weak D phenotype are managed without a weak D test as RhDnegative for purposes of Rh immunoprophylaxis with Rh immune globulin. While this strategy is safe and prevents Rh alloimmunization of RhD-negative pregnant women, the practice results in many pregnant women receiving unnecessary injections of Rh immune globulin. The AABB-CAP Work Group reviewed data pertaining to the safety of managing pregnant women and women of childbearing potential with a serological weak D phenotype. The Work Group determined that pregnant women with a weak D type 1, 2 or 3 in the homozygous or hemizygous state are not at risk of forming alloanti-d when exposed to conventional D+ RBCs (Sandler et al, 2015). The Work Group recommended that RHD genotyping be performed when routine RhD typing resulted in a serological weak D phenotype for a pregnant women or a woman of childbearing potential. If the result was a weak D type 1, 2 or 3, the woman can be managed safely as RhD-positive, because she is not at risk of forming anti-d (Fig 2). The Work Group s recommendation for pregnant women and other females of childbearing potential was accepted. In 2015, a Joint Statement on Phasing-in RHD Genotyping for Pregnant Women and Other Females of Childbearing Potential With a Serologic Weak D Phenotype was issued by the AABB, America s Blood Centers, the American Red Cross, the American College of Obstetricians and Gynecologists, CAP and the Armed Services Blood Program ( tement aspx?pf=1). Cost effectiveness of RHD genotyping pregnant women with a serological weak D phenotype A study using a Markov-based model evaluated the costs of options for managing the administration of Rh immune globulin in pregnant women in the United States whose RhD type was reported to be a serological weak D phenotype (Kacker et al, 2015). The study determined that there would be cost saving when the cost of RHD genotyping was less than 256 USD. Genotyping would decrease net cost among non-hispanic Caucasian females, but would increase cost among non- Hispanic African Americans, non-hispanic American Indian/ Alaskans and Hispanic women (Kacker et al, 2015). The differences in cost for RHD genotyping different populations are a reflection of the higher prevalence of serological weak D phenotypes associated with weak D types 1, 2 or 3 in Caucasians compared to other racial and ethnic populations. It s time for a paradigm shift: Clinical laboratories should implement policies to increase detection of serological weak D phenotypes and resolve their interpretation by RHD genotyping, not avoid their detection or make detection optional Prior to the availability of molecular methods capable of distinguishing different D variant antigens that are expressed as serological weak D phenotypes, laboratory practice in the United States was avoidance of the issue. Serological weak D phenotypes were interpreted as RhD-negative for pregnant women and transfusion recipients, and as RhD-positive for blood donors (Sandler et al, 2014a). For the past 50 years, laboratories have used RhD typing policies and procedures selected for their avoidance of detecting D variant antigens. It s time to change the paradigm and select RhD typing reagents that will not only detect normal ( wild type ) RhD antigens, but also detect D variant antigens. Such a scenario 16 ª 2017 John Wiley & Sons Ltd

8 has been proposed and is feasible by RhD typing using two monoclonal anti-d reagents, one recognizing DVI and other clinically important partial D variants, and another not recognizing clinically important partial D variants. (Denomme et al, 2005; Garratty, 2005; Denomme & Flegel, 2008; Von Zabern et al, 2013). Discrepant results using the two-reagent protocol would prompt identification of the variant D allele by molecular methods. Laboratories and transfusion services should discontinue reporting serological weak D phenotype as a test result in response to a request to perform an RhD blood type Traditionally, clinical practice for Rh immunoprophylaxis with Rh immune globulin, as well as transfusion of RBCs, has been based on interpreting the results of RhD typing as RhD-positive or RhD-negative. Advances in molecular science have identified 147 weak D alleles of the RHD gene, but the longstanding practice of managing pregnant women and patients as either RhD-positive or RhD-negative continues to be safe and adequate. As the science of RhD typing increasingly relies on RHD genotyping to guide the interpretation of serological weak D phenotypes, molecular laboratories have the responsibility of providing reports that are readily interpreted for managing patients as either RhD-positive or RhD-negative. It s time for laboratories to discontinue the practice of reporting serological weak D phenotype in response to a request to perform an RhD type. Today, given the laboratory resources for resolving a serological weak D phenotype, a laboratory offering RhD typing for clinical services should have an internal procedure for resolving the occasional serological weak D result by reflexively (automatically) performing RHD genotyping or referring the blood sample to a molecular reference laboratory for resolution. Should in-hospital clinical laboratories or regional reference laboratories perform the molecular testing required to resolve serological weak D phenotypes? Many models exist that demonstrate the cost savings and opportunities for increased quality when low volume tests are centralized in a reference laboratory. As clinicians increasingly recognize the benefit of applying RHD genotyping to resolve serological weak D phenotype results, laboratories will have options for providing the new service. Adding an in-hospital molecular service for blood groups will require the purchase of an automated extractor for DNA, PCR work stations, centrifuges, a thermal cycler, hybridization oven, imaging system and a computer (Sapatnekar & Figueroa, 2014). The initial capital expense and the ongoing cost of maintaining technical expertise for a relatively low-volume service are not realistic for most hospitals. For hospitals, an alternative model, i.e., centralizing the required molecular services in a community-based regional blood group reference laboratory, is more realistic. In our opinion, the second option hospitals refer blood samples for RHD genotyping to reference laboratories has the advantages of cost-effective high-volume operations which can fund highly skilled molecular scientists and acquire upto-date technology (analysers) as this new laboratory science evolves (Hillyer et al, 2008; Sandler et al, 2014b). Acknowledgements The authors thank Mouna Ouchari for updating Figure 1. This work was supported by the Intramural Research Program (project ID Z99 CL999999) of the NIH Clinical Center. Author contributions S. Gerald Sandler: Wrote the first draft and reviewed subsequent drafts. Leonard Chen: Fact-checker and formatted references. Willy A Flegel: Wrote all sections pertaining to molecular science, edited all drafts, prepared Figure 1 and Table II. Disclaimers The recommendations and opinions expressed are those of the authors, not their institutions or organizations. The views expressed do not necessarily represent the view of the National Institutes of Health, the Department of Health and Human Services, or the U.S. Federal Government. Conflict of interest WAF receives royalties for RHD genotyping. SGS and LD declare having no conflicts of interest relevant to this article. References Agre, P.C., Davies, D.M., Issitt, P.D., Lamy, B.M., Schmidt, P.J., Treacy, M. & Vengelen- Tyler, V. (1992) A proposal to standardize terminology for weak D antigen. Transfusion, 32, Argall, C.I., Ball, J.M. & Trentelman, E. (1953) Presence of anti-d antibody in the serum of a DU patient. Journal of Laboratory and Clinical Medicine, 41, Cannon, M., Pierce, R., Taber, E.B. & Schucker, J. (2003) Fatal hydrops fetalis caused by anti-d in a mother with partial D. Obstetrics and Gynecology, 102, Chou, S.T., Jackson, T., Vege, S., Smith-Whitley, K., Friedman, D.F. & Westhoff, C.M. (2013) High prevalence of red blood cell alloimmunization in sickle cell disease despite transfusion from Rh-matched minority donors. Blood, 122, Contreras, M. & Knight, R.C. (1989) The Rh-negative donor. Clinical Laboratory Haematology, 11, Contreras, M. & Knight, R.C. (1991) Controversies in transfusion medicine: testing for D u. Transfusion, 31, ª 2017 John Wiley & Sons Ltd 17

9 Daniels, G. (2002) Human Blood Groups, 2nd edn. Blackwell Science, Oxford. Daniels, G. (2013) Variants of RhD current testing and clinical consequences. British Journal of Haematology, 161, Denomme, G.A. (2013) Prospects for the provision of genotyped blood for transfusion. British Journal of Haematology, 163, 3 9. Denomme, G.A. & Flegel, W.A. (2008) Applying molecular immunohematology discoveries to standards of practice in blood banks: now is the time. Transfusion, 48, Denomme, G.A., Wagner, F.F., Fernandez, B.J., Wei, L. & Flegel, W.A. (2005) Partial D, weak D types, and novel RHD alleles among 33,864 multiethnic patients: implications for anti-d alloimmunization and prevention. Transfusion, 45, Domen, R. & Hoetge, G. (1997) Alloanti-D following transfusion in patients with the weak D (Du) phenotype [abstract]. Blood, 90 (suppl 1, pt 2):130b. Fasano, R.M. & Chou, S.T. (2016) Red blood cell genotyping for sickle cell disease, thalassemia and other transfusion complications. Transfusion Medicine Reviews, 30, Flegel, W.A. (2006) How I manage donors and patients with a weak D phenotype. Current Opinion in Hematology, 13, Flegel, W.A. (2007) The genetics of the Rhesus blood group system. Blood Transfusion, 5, Flegel, W.A. (2011) Molecular genetics and clinical applications for RH. Transfusion and Apheresis Science, 44, Flegel, W.A., Wagner, F.F., M uller, T.H. & Gassner, C. (1998) Rh phenotype prediction by DNA typing and its application to practice. Transfusion Medicine, 8, Flegel, W.A., Denomme, G.A. & Yazer, M.H. (2007) On the complexity of D antigen typing: a handy decision tree in the age of molecular blood group diagnostics. Journal of Obstetrics and Gynaecology Canada, 29, Flegel, W.A., Zabern, I.V. & Wagner, F.F. (2009) Six years experience performing RHD genotyping to confirm D- red blood cell units in Germany for preventing anti-d immunizations. Transfusion, 49, Fukumori, Y., Hori, Y., Ohnoki, S., Nagao, N., Shibata, H., Okubo, Y. & Yamaguchi, H. (1997) Further analysis of D el (D-elute) using polymerase chain reaction (PCR) with RHD gene-specific primers. Transfusion Medicine, 7, Garratty, G. (2005) Do we need to be more concerned about weak D antigens? Transfusion, 45, Garretta, M., M uller, A. & Gener, J. (1974) Determination du facteur Du sur Groupamatic. Revue Francaise de Transfusion, 17, Gu, J., Wang, X., Shao, C.P., Wang, J., Sun, A.Y., Huang, L.H. & Pan, Z.L. (2014) Molecular basis of DEL phenotype in the Chinese population. BMC Medical Genetics, 15, 54. Gunson, H.H., Stratton, F. & Phillips, P.K. (1976) The primary Rho (D) immune response in male volunteers. British Journal of Haematology, 32, Hemker, M.B., Ligthart, P.C., Berger, L., van Rhenen, D.J., van der Schoot, C.E. & Wijk, P.A. (1999) DAR, a new RhD variant involving exons 4, 5, and 7, often in linkage with cear, a new Rhce variant frequently found in African blacks. Blood, 94, Hill, Z., Vacl, J., Kalasova, E., Calabkova, M. & Pintera, J. (1974) Haemolytic disease of newborn due to anti-d antibodies in a Du-positive mother. Vox Sanguinis, 27, Hillyer, C.D., Shaz, B.H., Winkler, A.M. & Reid, M. (2008) Integrating molecular technologies for red blood cell typing and compatibility testing into blood centers and transfusion services. Transfusion Medicine Reviews, 22, Issitt, P.D. & Telen, M.J. (1996) Introduction of the term partial D (letter). Transfusion, 36, Jenkins, C.M., Johnson, S.T., Bellissimo, D.B. & Gottschall, J.L. (2005) Incidence of weak D in blood donors typed as D positive by the Olympus PK Immunohematology, 21, Judd, W.J., Moulds, M. & Schlanser, G. (2005) Reactivity of FDA-approved anti-d reagents with partial D red cells. Immunohematology, 21, Kacker, S., Vassallo, R., Keller, M.A., Westhoff, C.M., Frick, K.D., Sandler, S.G. & Tobian, A.A. (2015) Financial implications of RHD genotyping of pregnant women with a serologic weak D phenotype. Transfusion, 55, Kim, J.Y., Kim, S.Y., Kim, C.A., Yon, G.S. & Park, S.S. (2005) Molecular characterization of D-Korean persons; development of a diagnostic strategy. Transfusion, 45, Kim, K.H., Kim, K.E., Woo, K.S., Han, J.Y., Kim, J.M. & Park, K.U. (2009) Primary anti-d immunization by DEL red blood cells. Korean Journal of Laboratory Medicine, 29, Lacey, P.A., Caskey, C.R., Werner, D.J. & Moulds, J.J. (1983) Fatal hemolytic disease of the newborn due to anti-d in an Rh-positive DU variant mother. Transfusion, 23, Landsteiner, K. & Weiner, A. (1940) An agglutinable factor in human blood recognized by immune sera for Rhesus blood. Experimental Biology and Medicine, 43, 223. Le Marechal, C., Guerry, C., Benech, C., Burlot, L., Cavelier, B., Porra, V., Delamaire, M., Ferec, C. & Chen, J.M. (2007) Identification of 12 novel RHD alleles in western France by denaturing high-performance liquid chromatography analysis. Transfusion, 47, Leader, K.A., Kumpel, B.M., Poole, G.D., Kirkwood, J.T., Merry, A.H. & Bradley, B.A. (1990) Human monoclonal anti-d with reactivity against category DVI cells used in blood grouping and determination of the incidence of the category DVI phenotype in the DU population. Vox Sanguinis, 58, Levine, P. & Stetson, R.E. (1939) An unusual case of intragroup agglutination. Journal of the American Medical Association, 113, Luettringhaus, T.A., Cho, D., Ryang, D.W. & Flegel, W.A. (2006) An easy RHD genotyping strategy for D- East Asian persons applied to Korean blood donors. Transfusion, 46, Lukacevic Krstic, J., Dajak, S., Bingulac-Popovic, J., Dogic, V. & Mratinovic-Mikulandra, J. (2016) Anti-D antibodies in pregnant D variant antigen carriers initially typed as RhD+. Transfusion Medicine and Hemotherapy, 43, Mayne, K.M., Allen, D.L. & Bowel, P.J. (1991) Partial D women with anti-d alloimmunization in pregnancy. Clinical Laboratory Haematology, 13, McGann, H. & Wenk, R.E. (2010) Alloimmunization to the D antigen by a patient with weak D type 21. Immunohematology, 26, Monteiro, F., Tavares, G., Ferreira, M., Amotim, A., Rocha, C., Araujo, F. & Cunha-Ribeiro, L.M. (2011) Technologies involved in molecular blood genotyping. ISBT Science Series, 6, 1 6. M uller, T.H., Wagner, F.F., Trockenbacher, A., Eicher, N.I., Flegel, W.A., Sch onitzer, D., Schunter, F. & Gassner, C. (2001) PCR screening for common weak D types shows different distributions in three Central European populations. Transfusion, 41, Oberman, H.A. (ed.) (1981) Standards for Blood Banks and Transfusion Services, 10th edn. American Association of Blood Banks, Washington, DC. Okubo, Y., Yamaguchi, H., Tomita, T. & Nagao, N.A. (1984) D variant, D el? Transfusion, 24, 542. Okubo, Y., Seno, T., Yamano, H., Yamaguchi, H., Lomas, C. & Tippett, P. (1991) Partial D antigens disclosed by a monoclonal anti-d in Japanese blood donors. Transfusion, 31, 782. Ooley, P. (2015) Standards for Blood Banks and Transfusion Services, 30th edn. American Association of Blood Banks, Bethesda, MD. Ostgard, P., Fevang, F. & Komstad, L. (1986) Anti-D in a D-positive mother giving rise to a severe hemolytic disease of the newborn. A dilemma in antenatal immunohaematological testing. Acta Paediatrica Scandinavia, 75, Polin, H., Danzer, M., Hofer, K., Gassner, W. & Gabriel, C. (2007) Effective molecular RHD typing strategy for blood donations. Transfusion, 47, Salmon, C., Cartron, J.P. & Rouger, P. (1984) The Human Blood Groups, pg Masson Publishing, New York. Sandler, S.G., Roseff, S.D., Domen, R.E., Shaz, B. & Gottschall, J.L.; College of American Pathologists Transfusion Medicine Resource Committee. (2014a) Policies and procedures related to testing for weak D phenotypes and administration of Rh immune globulin. Archives of Pathology and Laboratory Medicine, 138, Sandler, S.G., Horn, T., Keller, J., Langeberg, A. & Keller, M.A. (2014b) A model for integrating molecular-based testing in a transfusion service. Blood Transfusion, 14, Sandler, S.G., Flegel, W.A., Westhoff, C.M., Denomme, G.A., Delaney, M., Keller, M.A., 18 ª 2017 John Wiley & Sons Ltd

10 Johnson, S.T., Katz, L., Queenan, J.T., Vassallo, R.R. & Simon, C.D. (2015) It s time to phase in RHD genotyping for patients with a serologic weak D phenotype. Transfusion, 55, Sapatnekar, S. & Figueroa, P.I. (2014) How do we use molecular red blood cell antigen typing to supplement pretransfusion testing? Transfusion, 54, Scientific Committee of the Joint Blood Council & Standards Committee of the American Association of Blood Banks. (1958) Standards for a Blood Transfusion Service, p. 10. Joint Blood Council Inc., Washington, DC. Shao, C.P. (2010) Transfusion of RHD-positive blood in Asia type DEL recipients. New England Journal of Medicine, 362, Shao, C.P., Maas, J.H., Su, Y.Q., K ohler, M. & Legler, T.J. (2002) Molecular background of Rh D-positive, D-negative, D(el) and weak D phenotypes in Chinese. Vox Sanguinis, 83, Simmons, R.T. & Krieger, V.I. (1960) Anti-Rh (D) antibodies produced by isoimmunization in an Rh positive woman of the unusual genotype (CDue/CdE). Medical Journal of Australia, 2, Stegmann, T.C., Veldhuisen, B., Bijman, R., Thurik, F.F., Bossers, B., Cheroutre, G., Jonkers, R., de Ligthart, P., Haas, M., Haer-Wigman, L. & van der Schoot, C.E. (2016) Frequency and characterization of known and novel RHD variant alleles in Dutch D-negative pregnant women. British Journal of Haematology, 173, Storry, J.R., Castilho, L., Daniels, G., Flegel, W.A., Garratty, G., de Haas, M., Hyland, C., Lomas- Francis, C., Moulds, J.M., Nogues, N., Olsson, M.J., Poole, J., Reid, M.E., Rouger, P., van der Schoot, E., Scott, M., Tani, Y., Yu, L.-C., Wendel, S., Westhoff, C., Yahalom, V. & Zelinski, T. (2014) International Society of Blood Transfusion Working Party on red cell Immunogenetics and blood group terminology: Cancun report (2012). Vox Sanguinis, 107, Storry, J.R., Castilho, L., Chen, Q., Daniels, G., Denomme, G., Flegel, W.A., Gassner, C., de Haas, M., Hyland, C., Keller, M., Lomas-Francis, C., Moulds, J.M., Nogues, N., Olsson, M.L., Peyrard, T., van der Schoot, C.E., Tani, Y., Thornton, N., Wagner, F., Wendel, S., Westhoff, C. & Yahalom, V. (2016) International society of blood transfusion Working party on red cell Immunogenetics and terminology: report of the Seoul and London meetings. ISBT Science Press, 11, Stratton, F. (1946) A new Rh allelomorph. Nature, 158, Stroup Walters, M. (1988) More on Du. Immunohematology, 4, Tilley, L. & Grimsley, S. (2014) Is the next generation sequencing the future of blood group testing? Transfusion and Apheresis Science, 50, Tippett, P. & Sanger, R. (1962) Observations on subdivisions of the Rh antigen D. Vox Sanguinis, 7, Unger, L.J., Wiener, A.S. & Katz, L. (1959) Studies on blood factors RhA, RhB, and RhC. The Journal of Experimental Medicine, 110, Urbaniak, S.J. & Robertson, A.E. (1981) A successful program of immunizing Rh-negative male volunteers for anti-d production using frozen/ thawed blood. Transfusion, 21, Van Rhenen, D.J., Thijssen, P.M. & Overbeeke, M.A. (1989) Testing efficacy of anti-d sera by a panel of donor red cells with weak reacting D antigen and with partial D antigens. Vox Sanguinis, 56, Vichinsky, E.P., Earles, A., Johnson, R.A., Hoag, M.S., Williams, A. & Lubin, B. (1990) Alloimmunization in sickle cell anemia and transfusion of racially unmatched blood. New England Journal of Medicine, 322, Von Zabern, I., Wagner, F.F., Moulds, J.M., Moulds, J.J. & Flegel, W.A. (2013) D category IV: a group of clinically relevant and phylogenetically diverse partial D. Transfusion, 53, Wagner, F.F., Kasulke, D., Kerowgan, M. & Flegel, W.A. (1995) Frequencies of the blood groups ABO, Rhesus, D category VI, Kell and of clinically relevant high-frequency antigens in southwestern Germany. Infusionsther Transfusionmed, 22, Wagner, F.F., Gassner, C., M uller, T.H., Sch onitzer, D., Schunter, F. & Flegel, W.A. (1999) Molecular basis of weak D phenotypes. Blood, 93, Wagner, F.F., Frohmajer, A., Ladewig, B., Eicher, N.I., Lonicer, C.B., M uller, T.H., Siegel, M.H. & Flegel, W.A. (2000) Weak D alleles express distinct phenotypes. Blood, 95, Wagner, F.F., Frohmajer, A. & Flegel, W.A. (2001) RHD positive haplotypes in D negative Europeans. BMC Genetics, 95, Wagner, T., K orm oczi, G.F., Buchta, C., Vandon, M., Lanzer, G., Mayr, W.R. & Legler, T.J. (2005) Anti-D immunization by DEL red blood cells. Transfusion, 45, Wagner, F.F., Flegel, W.A., Bittner, R. & D oscher, A. (2017) Molecular typing for blood group antigens within 40 min by direct polymerase chain reaction from plasma or serum. British Journal of Haematology, 176, Wang, D., Lane, C. & Quillen, K. (2010) Prevalence if RhD variants, confirmed by molecular genotyping in multiethnic prenatal populations. American Journal of Clinical Pathology, 134, Wang, Q.P., Dong, G.T., Wang, X.D., Gu, J., Li, Z., Sun, A.Y., Shao, C.P., Pan, Z.L., Huang, L.H., Xie, W.X., Sun, G.M., Chen, J.J., Pei, H., Yang, X.J. & Shan, P.N. (2014) An investigation of secondary anti-d immunization among phenotypically RhD-negative individuals in the Chinese population. Blood Transfusion, 12, Weiner, A.S. & Unger, L.J. (1962) Further observation on the blood factors Rh-A, Rh-B, Rh-C and Rh-D. Transfusion, 2, Westhoff, C.M. (2005) Review: the Rh blood group D antigen...dominant, diverse, and difficult. Immunohematology, 21, White, C.A., Stedman, C.M. & Frank, S. (1983) Anti-D antibodies in D- and Du-positive women: a cause of hemolytic disease of the newborn. American Journal of Obstetrics and Gynecology, 145, Yang, Y.F., Wang, Y.H., Chen, J.C., Eng, H.L. & Lin, T.M. (2007) Prevalence of RHD1227A and hybrid Rhesus in the general Chinese population. Translation Research, 149, Yang, H.S., Lee, M.Y., Park, T.S., Cho, S.Y., Lee, H.J., Lim, G., Lee, D.D., Oh, S.H., Cho, D. & Park, K.U. (2015) Primary anti-d alloimmunization induced by Asian type RHD (c.1227g>a) DEL red cell transfusion. Annals of Laboratory Medicine, 35, Yasuda, H., Ohto, H., Sakuma, S. & Ishikawa, Y. (2005) Secondary anti-d immunization by Del red blood cells. Transfusion, 45, ª 2017 John Wiley & Sons Ltd 19