review

Variants of RhD – current testing and clinical consequences

Geoff Daniels

International Blood Group Reference Laboratory, NHS Blood and Transplant, Bristol, UK

Anti-D (-RH1) of the Rh blood group system is clinically In this review I will attempt to clarify the situation, recom-
important as it causes haemolytic transfusion reactions and mend a testing and transfusion strategy that minimizes the
haemolytic disease of the fetus and newborn. Although most risk of alloimmunization without undue waste of D blood
people are either D+ or D , there is a plethora of D variants, stocks, and describe the clinical issues relating to these aber-
often categorized as either weak D or partial D. These two rant forms of D.
types are inadequately defined and the dichotomy is potentially
misleading. DVI is the D variant most commonly associated Molecular genetics of Rh and the
with anti-D production and UK guidelines recommend that D polymorphism
patients are tested with anti-D reagents that do not react with
DVI. Weak D types 1, 2, and 3 are seldom, if ever, associated RHD and RHCE, a pair of homologous genes sharing 93
8%
with alloanti-D production, so a policy recommendation homology over all introns and coding exons (Okuda et al,
would be to treat patients with those D variants as D+, to 2000), encode the antigens of the Rh system. They are
preserve D stocks, whereas patients with all other D variants located on chromosome 1p36
11 and are closely linked, but
would be treated as D . All donors with D variant red cells, are in opposite orientation: 5′-RHD-3′–3′-RHCE-5′ (Wagner
including DVI, should be treated as D+. & Flegel, 2000) (Fig. 1). Each gene consists of 10 exons that
encode 417-amino acid polypeptides, although the N-termi-
Keywords: Rh blood group system, D antigen, D variant, nal methionine is cleaved from the mature proteins (Avent
weak D, partial D. et al, 1990; Ch
erif-Zahar et al, 1990). The RhD and RhCE
proteins differ by between 31 and 35 amino acids, depending
RhD (D or RH1), originally identified in 1939, was the first on the RHCE allele. They are hydrophobic molecules that
clinically important blood group to be found following the span the red cell membrane 12 times, with internal N- and
discovery of ABO 39 years earlier. A phenotypic relationship C-termini and six potential external loops (Fig. 1). Atypically
between D and an antigen on human red cells detected by for red cell membrane proteins, they are not glycosylated.
antibodies made in rabbits immunized with rhesus monkey Homology modelling based on crystal structures of bacterial
red cells, led to D being inappropriately named the Rhesus homologues of the Rh proteins and on the human non-ery-
antigen. A vestige of that term remains in Rh, the name of throid Rh family glycoprotein RhCG predict the presence of
the blood group system that contains D. an extracellular vestibule in the regions of the third and
The frequency of the D+ phenotype is about 85% in fourth external loops and sixth, seventh, and eighth mem-
Caucasians, around 95% in sub-Saharan Africa, and greater brane spanning domains that penetrates the cell membrane
than 99
5% in eastern Asia (Daniels, 2013). Although most permitting access to IgG antibodies (Conroy et al, 2005;
people are either D+ or D , a grey area exists in the form Avent, 2007; Gruswitz et al, 2010; Flegel, 2011).
of a plethora of D variants: the so-called weak D, partial D, The D phenotype usually results from a complete
and DEL phenotypes. These ill-defined terms are often a absence of the RhD protein, explaining the high immunoge-
cause of consternation beyond their clinical relevance. Over nicity of D. The RhCE protein is almost always present. In
its history of well over a century, blood group serology has Caucasians, the usual cause of the D phenotype is homozy-
been fraught with problems arising from nomenclature, and gosity for a complete deletion of RHD (Colin et al, 1991).
the terminology applied to D variants has led to confusion. Deletion of RHD is also the most common cause of D in
black Africans, but around 67% of D black Africans have
at least one copy of RHD*Ψ, a D gene inactivated by a 37 bp
duplication in exon 4 and a nonsense (Tyr>stop) mutation
This article is published with the permission of the Controller of in exon 6 (Singleton et al, 2000). Some RHD–CE–D hybrid
HMSO and the Queen’s Printer for Scotland. genes produce no D antigen and are present in about 15% of
Correspondence: Geoff Daniels, IBGRL, NHSBT, North Bristol Park, black Africans. They consist of exons 1, 2, 9, and 10 derived
500 Northway, Filton, Bristol BS34 7QH, UK. from RHD, but exons 4–8 derived from RHCE. Exon 3 is
E-mail: [Link]@[Link] usually part RHD and part RHCE (type 1), but may be

ª 2013 Crown copyright, British Journal of Haematology doi:10.1111/bjh.12275

Review

5′ 3′ 3′ 5′ Table I. Some D variants associated with production of alloanti-D

1 10 10 1
DII DFL DWI
DIII DFR Weak D type 1*
Rh
box RHD Rh
box TMEM50A RHCE DIVa DFV Weak D type 2*
C/c DIVb DHAR Weak D type 4
E/e
DV DHMI Weak D type 11
DVI DMH Weak D type 15
DVII DMI Weak D type 21
N N
C C DAR DNAK Weak D type 57
RhD RhCE DAU DNB DEL-5
DBT DOL DEL-ex 8 del
Fig. 1. Organization of RHD & RHCE and diagram to demonstrate
the conformation of the RhD and RhCE protein in the membrane. *Controversial, see text.

entirely RHCE (type 2) (Faas et al, 1997; Daniels et al, 1998; appear to have elevated D expression (e.g. DIVa) (Daniels,
Pham et al, 2009). Although there are many other mutated 2013). Associated low frequency antigens have subdivided
RHD genes and hybrid genes responsible for D phenotypes, categories IV, V, and VI, and assist in the definition of DIII,
these are relatively rare and the most common D genotypes DVII, DFR, DBT, DAU-5, and DHAR (Table II).
are homozygosity or compound heterozygosity for an RHD The term Du was initially used for the D antigen of those
deletion, RHD*Ψ, or RHD–CE–D. red cells that are not agglutinated by IgM anti-D, but which
react with IgG anti-D in an antiglobulin test. With modern,
monoclonal anti-D reagents, most red cells that would previ-
D variants ously have been classified as Du would be considered as
Before discussing testing strategies and clinical consequences, having a normal D by routine testing. In contrast to the
we must first understand what is meant by the term D vari- qualitative partial D antigens, Du was considered a quantita-
ant. It is commonly understood that there are two classes of tive variant of D, differing from normal D purely on the
D variant: weak D and partial D. number of antigen sites per red cell. Consequently there
The concept of the D mosaic, in which the normal D anti- could be no Du antigen and no anti-Du, so Du was replaced
gen is a mosaic with all pieces present, but that variants may with the term weak D in the 1990s (Agre et al, 1992).
have some pieces missing, was developed in the 1950s and
1960s (Tippett & Sanger, 1962; Wiener & Unger,1962). A Molecular genetics of D variants
person whose D antigen lacks parts of the mosaic and who is
immunized with normal D may produce an antibody to the Broadly, two types of molecular mechanisms are responsible
parts they lack; an antibody that will behave in most routine for D variants: (1) one or occasionally a few nucleotide
tests as anti-D. With the production of human monoclonal changes in RHD, resulting in amino acid substitutions in the
antibodies to D in the 1980s, the concept of the mosaic RhD protein; (2) the product of genetic recombination,
became replaced by the concept of D epitopes. By testing probably as a result of gene conversion, giving rise to an
numerous monoclonal antibodies against a wide range of D RHD–CE–D gene in which a section of RHD is replaced by
variant red cells in an international workshop, 30 ‘epitopes’ the corresponding section from RHCE. The converted section
were defined (Scott, 2002). may consist of part of an exon, a whole exon, or several
Qualitative variants, in which some or many epitopes are exons of RHCE. In DHAR, however, there is a deletion of
lacking, making the production of a D-like antibody possible,
are commonly termed partial D. Testing of partial D red cells Table II. D variants associated with Rh low frequency antigens
with the D-like antibodies produced by individuals with par- (LFAs)
tial D antigens led to classification of partial D antigens into
categories. These categories were numbered (in Roman D variant Rh LFAs
numerals) from DII (DI was made obsolete) to DVII, and DIII DAK (RH54)
some were subdivided on the basis of serological subtleties. DIVa Goa (RH30)
Further analyses, initially with monoclonal antibodies and DIVb Evans (RH37)
then by sequencing of the genes encoding the antigens, pro- DV Dw (RH23)
duced a plethora of partial D antigens, designated by names DVI BARC (RH52)
such as DBT, DFR, and DHAR (Table I). The strength of DVII Tar (RH40)
DBT RH32
reactivity of partial D antigens with those anti-D that react
DFR FPTT (RH50)
with them may be weaker than that of normal D+ cells (e.g.
DHAR RH33, FPTT
DVI), similar to normal D+ cells (e.g. DIII), or may even

2 ª 2013 Crown copyright, British Journal of Haematology

 Review

1 2 3 4 5 6 7 8 9 10 Whereas both types of molecular mechanism described

D above may account for partial D antigens, Wagner et al
DIIIa (1999) proposed that only amino acid substitutions located
DIIIb within the membrane spanning or cytosolic domains of the
DIIIc D protein are responsible for weak D. Through RHD
DIVa sequencing they identified 16 different types of weak D,
DIVb numbered weak D type 1 to weak D type 16. This has now
DIV-3 been expanded to weak D type 81, with several subtypes
DIV-4 (e.g. weak D type 4
1) (Rhesus base website, [Link]
DIV-5 [Link]/~fwagner/RH/RB). These weak D types cannot be
DVa distinguished serologically, though they are associated with
DV-1 different degrees of weakness. Among white German and
DVa-4 Austrian blood donors with weak D red cells, 70%, 18%, and
DVa-6 5% were weak D types 1, 2, and 3 respectively (Wagner et al,
DVI-1 1999), whereas among French patients the equivalent fre-
DVI-2 quencies were 38%, 27%, and 8% (Noizat-Pirenne et al,
DVI-3 2007).
DVI-4 The number of D antigen sites per red cell in the ‘weak D
DAR
types’ varies from about 60 to 3800. This compares with
DBT-1
13,000–24,000 for ‘normal’ D+ red cells and 3000 on DVI
DBT-2
type 2 and 33,000 on DIII type 4 red cells (Wagner et al,
DCS-1
2000).
DFR-1
DFR-2
DHAR Definitions of weak D and partial D

Fig. 2. RHD exons (black boxes) and RHCE exons (white boxes) of The terms weak D and partial D have often been used to
hybrid genes in some D variants. determine how a patient is transfused or whether anti-D
immunoglobulin prophylaxis should be given. These terms,
RHD (like D ), but exon 5 of RHCE has been replaced by however, are not adequately defined and, therefore, not suit-
exon 5 of RHD, responsible for expression of some D able for this purpose. Three types of differentiating defini-
epitopes (Fig. 2). tions have been used, but none is satisfactory.
Rh epitopes are conformational: antibody binding is 1 Weak D antigens have all D epitopes; partial D antigens
dependent on the shape of the molecule within the mem- lack one or more D epitopes. This is difficult to define
brane and epitopes may even involve more than one of the serologically, because a negative reaction with a particular
extracellular loops. For antibody binding, epitopes must monoclonal antibody or by a specific method could result
involve the extracellular loops or the proposed extracellular from weak expression of the epitope, rather than its
vestibule, but changes in parts of the molecule that are not absence. DIIIa appears to have all epitopes, but individu-
exposed to the outside of the membrane, such as those als with DIIIa red cells often make anti-D and so their
within the membrane-spanning a-helices, may weaken or red cells must be lacking at least one D epitope.
destroy some epitopes and even create new ones. 2 Individuals with partial D antigens can make anti-D;
Some serologically-defined D variants have been subdi- those with weak D antigens cannot. This is the usual
vided by their genotype. For example, there are four types of interpretation of the dichotomy, but it is dependent on
RHD*DVI. All have an RHD–CE–D hybrid gene, but the an immune response. If anti-D has not been found in any
RHCE-derived exons are 4 and 5 in type 1, 4–6 in type 2, person with a particular D variant, this does not mean
3–6 in type 3, and 3–5 in type 4 (Fig. 2). All encode the that another patient with the same variant will not make
same changes in the external loops of the RhD protein and anti-D following immunization with D+ red cells. For
all encode a D antigen with epitope characteristics typical of example, weak D types 4
2 and 15 have been classed as
DVI. The encoded phenotypes are not identical; they differ weak D, yet all have subsequently been found in numer-
in D antigen strength and expression of the DVI-associated ous patients who have made alloanti-D (Wagner et al,
antigen BARC (Mouro et al, 1994; Avent et al, 1997; Wagner 2000).
et al, 1998; Esteban et al, 2006). Likewise, there are 3 In partial D, the RhD proteins have amino acid changes
seven types of RHD*DV, all encoding Glu233Gln, but all outside of the membrane, whereas in weak D the RhD
except RHD*DV type 4 have additional mutations in exon 5 proteins have one or more amino acid substitutions
(Rouillac et al, 1995; Omi et al, 1999; Hyodo et al, 2000; within either the membrane-spanning domains or the
M€ uller et al, 2001). cytoplasmic loops of the protein, but not externally

ª 2013 Crown copyright, British Journal of Haematology 3

Review

exposed (Wagner et al, 1999; Flegel et al, 2007). One D Chinese women with a history of pregnancy, 44 had the
major flaw with this definition is that the precise locations DEL variant and no anti-D, yet 38 of 155 who were truly
of the amino acid residues of the RhD protein within the D had anti-D (Shao, 2010). The most common DEL allele
membrane are not known; different models predict differ- in Caucasians encodes Met295Ile (weak D type 11). The
ent locations for some amino acids. Also, this definition second most common has the splice site mutation RHD
is not functional from the point of view of transfusion (IVS3+1g>a) (Wagner et al, 2001; Gassner et al, 2005). Four
practice, as RHD genotyping is often not performed rou- individuals with RHD(IVS3+1g>a) had anti-D, with one of
tinely in many hospitals or even reference laboratories. the antibodies causing mild haemolytic disease of the fetus
Furthermore, the words ‘weak’ and ‘partial’ have well- and newborn (HDFN) (K€ orm€oczi et al, 2005; Gardener et al,
defined meanings in the English language and to use them 2012).
as code for structural difference in a membrane protein is
misleading.
Clinical importance of D variants in patients
In some cases, symbols suggestive of variant D antigen
After the ABO antibodies, anti-D is clinically the most
activity may have been awarded when an RHD mutation was
important blood group antibody in transfusion medicine,
detected, despite lack of evidence of abnormal D expression.
because anti-D can cause severe acute and delayed
For example, DMA (Leu207Phe) is encoded by an RHD gene
haemolytic transfusion reactions and severe HDFN (Klein &
that has only been found in trans with normal RHD, so there
Anstee, 2005). D is very immunogenic, with about 20–30%
is no information on the effect on D expression (Wagner
of D– patients who receive large volumes of D+ blood mak-
et al, 2003), and weak D types 19 (Ile204Thr) and 20
ing anti-D (Frohn et al, 2003; Klein & Anstee, 2005; Yazer &
(Phe417Ser) have normal D antigens as determined by sero-
Triulzi, 2007; Gonzalez-Porras et al, 2008). It is imperative,
logical tests, although there are reduced numbers of D anti-
therefore, that D+ red cells are not transfused to patients
gen sites (Yu et al, 2006). Without evidence of abnormal D
with anti-D or with a history of having anti-D, to D girls
expression, the terms weak D, partial D, and D variant are
and women of childbearing age, or to patients with a poten-
inappropriate.
tial for transfusion dependency.
The D variants DAR and weak D type 4
2 have been
Patients with many D variant phenotypes are capable of
found in individuals with alloanti-D and have the same three
producing alloanti-D if immunized with D+ red cells. Table I
amino acid substitutions, differing only by a single synony-
lists many of the reported D variants present in individuals
mous mutation (Hemker et al, 1999; Wagner et al, 2000).
with alloanti-D. This is not a complete list and absence of a
Consequently, the terms DAR and weak D type 4
2 represent
variant from this list does not imply that a patient with that
D variants with identical phenotypes and almost identical
variant cannot make alloanti-D as a result of transfusion
genotypes, yet the former is considered partial D and the lat-
with D+ red cells or a D+ pregnancy.
ter has a weak D notation.
Weak D types 1–3 are the most frequent of those variants
The weak D/partial D dichotomy is artificial and the
with a ‘weak D’ designation, representing about 93% of those
terminology used to describe these variants is misleading.
D variants in a Caucasian population (Wagner et al, 1999).
The terms weak D and partial D should now be replaced
Frequencies vary in different populations: the very rare weak
with a single collective name, such as D variant (Daniels
D type 38 (Gly278Asp) is relatively common in Portuguese
et al, 2007).
(Rodrigues et al, 2006); weak D type 42 is the most common
in Quebec, Canada (St-Louis et al, 2011), and weak D type 3
the most common in the Zagreb region of Croatia (Dogic
DEL
et al, 2011).
DEL is a very weak form of D that cannot be detected by There are a few reports of anti-D in patients with weak D
conventional serological tests and can only be recognized types 1 and 2 (Beckers et al, 2004; Roxby et al, 2004; Daniels
serologically by adsorption and elution tests. The maximum et al, 2007; Pelc-Klopotowska et al, 2010; Vege et al, 2011),
antigen site density per red cell was estimated to be 36, but D antibodies in patients with these phenotypes are gener-
though in most cases there are less than 22 (K€ orm€
oczi et al, ally autoantibodies (Wagner et al, 2000; Pham et al, 2011).
2005). In eastern Asia, between 10% and 33% of red cell In 220 patients with weak D types 1 and 2, Pham et al
samples found to be D– by conventional serological tech- (2011) identified 12 with anti-D and then showed that all 12
niques are DEL (Daniels, 2013) and the most common RHD were autoantibodies. Failure in a survey to reveal any alloan-
gene responsible has a synonymous mutation in the 3′ nucle- tibodies cannot be interpreted as meaning that alloanti-D
otide of exon 9 (1227G>A, Lys409Lys), that results in loss of cannot be produced by patients with weak D types 1 and 2,
exon 9 from the mRNA (Shao et al, 2006; Liu et al, 2010). but it does indicate that it could only be a rare event. If the
The presence of DEL in Asians is not usually associated red cells of a patient with anti-D give a negative direct anti-
with anti-D production. Of 104 Chinese pregnant women globulin test (DAT), no anti-D can be eluted from them,
with anti-D, none had the DEL phenotype. Of 199 apparent and they do not adsorb the anti-D from the patient’s own

4 ª 2013 Crown copyright, British Journal of Haematology

 Review

plasma, they should be considered to be alloanti-D. They mistaken for D+. Some D variants, such as DIIIa, will type
should only be dismissed as autoanti-D if there were a posi- as normal D+ with monoclonal reagents and therefore be
tive DAT, anti-D could be eluted from the patient’s red cells, reported as D+ and be at risk of making anti-D.
and/or the patient’s red cells adsorb the anti-D from the According to UK guidelines, when significantly weak
autologous plasma. It is important to classify antibodies as results are obtained on routine D typing of patient red cells,
autoanti-D or alloanti-D on the basis of commonly applied then patients should be considered D+ unless they are either
serological criteria, rather than on preconceived dogma based a female patient of childbearing potential or a patient who is
on the patient’s D phenotype or genotype. likely to require long-term transfusion support, in which case
Many D variants are rare and, although there are some they should be treated as D until the cause of the aberrant
frequency data for some variants, there is no published infor- result is resolved (Fig. 3) (Milkins et al, 2013). Further test-
mation on the prevalence of different variants in individuals ing with a panel of D monoclonal antibodies, or with a
with anti-D. Purely from observation, the most commonly commercial kit, such as that produced by Alba Bioscience
encountered D variant in patients with anti-D, at least in a (Penicuik, Midlothian, UK), will assist in the identification
predominantly Caucasian population, is DVI. In an African of some of the more common variants, especially weak D
population, or in people of African origin, DIIIa and DAR types 1 and 2 and DAR (but not DIIIa). More definitive
are probably the most commonly encountered D variants in identification can be achieved by molecular testing, either
patients with anti-D. with kits or arrays designed to identify many or most of the
D variant red cells express at least some epitopes of D, so known variants, or by sequencing RHD, which will also iden-
D antibodies made by D variant individuals might not tify new variants. Once the variant has been identified, then
have as broad a specificity and, therefore, have the same a policy must be in place for transfusion and giving anti-D
haemolytic potential as those made by D individuals. prophylaxis.
Anti-D in women with variant D red cells have been respon- It has been common practice that weak D patients should
sible for severe HDFN. Some cases, severe enough to warrant receive D+ red cells, to conserve stocks of D blood, and
exchange transfusion and one leading to fetal death, have that partial D patients should be given D red cells. In addi-
been published and involve mothers with DIIIa (Beckers tion, partial D pregnant women would be given anti-D
et al, 1996), DIVb (Okubo et al, 1991), DVI (Lacey et al, immunoglobulin prophylaxis, whilst weak D women would
1983), and DBT (Wallace et al, 1997), and D variants not. This policy is based on the premise that weak D patients
described as DU (Hill et al, 1974; White et al, 1983), ‘D+’
(Østgard et al, 1986), and RhB (Lowes, 1969). Although I am
not aware of any report of anti-D in a D variant patient Is the reaction grade with one or
causing a haemolytic transfusion reaction, D variant patients more anti-D positive but weak*?
with anti-D should be transfused with D red cells.
No
Yes
D testing in patients and in pregnancy Report and treat as
D+
The recommended practice in the UK is that patients,
Is the patient female
including pregnant women, should be tested in duplicate and < 50 years of
(unless samples are tested on secure automation) by direct age? Yes
agglutination with potent IgM monoclonal anti-D (Milkins Treat as D– (or hold if
et al, 2013). These reagents are selected to detect all but the No possible) and refer for
confirmation of D type
weakest examples of D and to give a negative reaction with Yes
DVI red cells. Consequently, patients who have the DVI phe- Is the patient likely to
notype and are, therefore, prone to make anti-D, are typed require chronic
transfusion support?
as D and treated as D for blood transfusion and adminis-
Weak D types 1, 2, 3
tration of anti-D immunoglobulin during pregnancy and fol- Treat as D+
No
lowing delivery of a D+ baby. Patients with very weak All other D variants
expression of D, including those with DEL phenotype, will Treat as D–
Report and treat as
also be typed as D . An antiglobulin test for detection of D+
some of the weaker forms of D variant phenotypes is not
recommended for patient testing in the UK (Milkins et al, Fig. 3. Reporting of D typing anomalies and selection of red cells.
2013), or in other countries such as France and Germany Adapted from: (Milkins et al, 2013) Guidelines for pre-transfusion
compatibility procedures in blood transfusion laboratories. Transfu-
(Mota et al, 2005), the Netherlands ([Link]
sion Medicine, 23, 3-35. With permission from John Wiley & Sons.
document/7584) or the USA (AABB, 2011), but is mandatory *Weak reaction is defined by local policy and in line with manufac-
in some countries. If an antiglobulin test is applied in D turers’ instructions – likely to be <3+ or <2+ depending on system
typing, a positive DAT could result in D red cells being used. Recommendation in dashed box is not in the UK guidelines.

ª 2013 Crown copyright, British Journal of Haematology 5

Review

cannot make anti-D, whereas partial D patients can. This is a haemolytic transfusion reactions, making the selection of
flawed logic for reasons already given and such a policy is suitable red cells for subsequent transfusions increasingly
not recommended. difficult (Yazdanbakhsh et al, 2012). SCD patients who are
It is well established that individuals with weak D types 1, of African origin, but living in countries where the donor
2, and 3 make alloanti-D only extremely rarely, if at all. So it pool is predominantly Caucasian, have an enhanced likeli-
is a reasonable policy to treat patients with those variants as hood of blood group antibody production. The D variants
D+, in order to conserve stocks of D red cells, and to treat DIIIa and DAR are almost unknown in Caucasians, but are
patients with any other D variant as D , which will have relatively common in people of African origin, and both
only a minimal effect on D stocks as all of the other vari- are often associated with anti-D production (Hemker et al,
ants are relatively rare. This is compatible with published 1999; Castilho et al, 2005; Westhoff et al, 2010). In a study
recommendations (Wagner et al, 2000; Pham et al, 2011) of 39 SCD patients with DIIIa, 18 had alloanti-D (Westhoff
and the policy that has now been adopted in England et al, 2010). From over 1000 Afro-Caribbeans living in
(Fig. 3). If, however, only serological testing is being used, Martinique, the allele frequencies of RHD*DIIIa and
this policy is difficult to apply because identification of vari- RHD*DAR were 0
88% and 0
39% respectively (Silvy et al,
ants is often not possible. Noizat-Pirenne et al (2007) 2011). Although DAR gives a characteristic pattern of reac-
observed a range of reactivity for weak D of the same type, tions with monoclonal anti-D, DIIIa red cells react with all
tested with the same reagent by a gel matrix technique. monoclonal anti-D and so cannot easily be detected by
Anti-D immunoglobulin should be offered to women serological methods. Transfusion of patients with these D
known to have D variant red cells, other than weak D types variants with D red cells would prevent anti-D produc-
1–3, during and after pregnancy, because the anti-D constitu- tion, but could be self-defeating as the D red cells would
ent that does not bind to the mother’s own D variant cells generally come from Caucasian donors, increasing the risk
should be available to suppress immunization (Lubenko of exposure to other red cell antigens (Noizat-Pirenne &
et al, 1989). This is particularly important in DVI mothers, Tournamille, 2011).
whose red cells lack most D epitopes, although in most cases
DVI mothers will be routinely typed as D and receive
D variants in blood donors
prophylactic anti-D automatically.
There is little information about immunization of It is essential that donor red cells capable of immunizing a
D women by D variant fetuses. DVa stimulated production D patient to make anti-D are labelled D+. Consequently,
of anti-D in a D– woman during her first pregnancy and reagents are selected that give a positive reaction with DVI
resulted in mild HDFN of her second baby, who also red cells and those of most other D variants. DVI individu-
appeared to have the variant D antigen (Mayne et al, 1990). als, therefore, are D+ donors, but D patients.
The mother had received anti-D immunoglobulin after the In the UK, the policy for D typing donors (UK Blood
first pregnancy. A weak D type 3 fetus appears to have stim- Transfusion & Tissue Transplantation Services, 2005) is as
ulated anti-D production in a D C+ mother, causing the follows:
neonatal red cells to give a positive DAT (Needs et al, 2007).
● The D blood group must be determined on each donation
Although there is no evidence that DVI and other weak
of blood.
forms of D can immunize a D mother during pregnancy or
● In the testing of donors being grouped for the first time,
at delivery, logic might deem that neonatal (umbilical cord)
two anti-D blood grouping reagents capable of detecting
D typing be performed with reagents that detect DVI, so that
between them DIV, DV and DVI should be used. If two
the mother would receive anti-D immunoglobulin if the baby
monoclonal anti-D are used, they should be from different
has DVI red cells. That is, neonates would be treated like
clones.
blood donors for D typing. This is the policy in some coun-
● Donors whose red cells give unequivocal positive or
tries, such as the Netherlands ([Link]
negative reactions with both anti-D reagents should be
ment/7584) and the USA (AABB, 2011), but in the UK the
regarded as D+ and D respectively.
BCSH guidelines (Milkins et al, 2013) recommend that the
● If the results with the anti-D reagents are discordant or
same reagents are used for testing mother and cord, in order
equivocal, the tests should be repeated. Where the D group
to avoid any confusion leading to the mother’s red cells
is in doubt it is safer to classify such donors as D+.
being tested with the wrong reagents.
● For known (repeat) donors, one anti-D reagent, or
blended reagent, that detects weak D, DIV, DV and DVI
D variants and sickle cell disease can be used.

Blood transfusion is an important tool in the management Some D variants, such as DEL, are not detected by agglu-
of sickle cell disease, especially for the prevention of stroke. tination tests, even by indirect antiglobulin tests. There is
Multiply transfused patients are prone to making blood evidence that DEL can cause primary immunization of
group antibodies, which can cause life-threatening delayed D patients through transfusion (Wagner et al, 2005; Chen

6 ª 2013 Crown copyright, British Journal of Haematology

 Review

et al, 2006), as can other very weak forms of D, including commercialized by Progenika (Derio, Spain), distinguishes
weak D type 2 (Flegel et al, 2000), weak D type 26 (Gassner about 80 RHD variant genes (Avent et al, 2007). Another
et al, 2005), weak D type 1 with dCe in trans (which weakens array-based system, employing coloured microbeads, distin-
expression) (Mota et al, 2005), and a D /D+ chimera with a guishes about 70 RHD variant genes (BioArray Solutions,
proportion of D+ cells too low to be detected by routine Warren, NJ, USA) and a non-commercialized platform,
serological testing (Wagner et al, 2001). This has led a few employing multiplex ligation-dependent probe amplification,
transfusion services in Europe, mostly in Switzerland and determines 79 RHD variant genes (Haer-Wigman et al,
parts of Germany, to introduce molecular screening for RHD 2013). Some other commercial blood group genotyping plat-
in D donors, so that transfusion of D patients with DEL forms apply polymerase chain reaction with sequence-specific
red cells can be avoided, especially to girls and women of primers and gel electrophoresis to distinguish a smaller num-
childbearing age (Flegel et al, 2009). In an international ber of variants: for example, Innotrain (Kronberg, Germany)
forum in 2006 with representation from 16 blood services and BAGene (Lich, Germany). Although it is usually not nec-
and a regulatory authority in 13 countries, only seven proven essary to identify most D variants by molecular testing in
cases of anti-D immunization in D recipients induced by routine transfusion practice, and some of the commercial kits
red cells from a donor typed D in routine testing were are marked, ‘for research use only’, the identification of weak
recorded and none of the contributors were of the opinion D types 1–3 in patients is valuable, as these can be treated as
that routine molecular testing for RHD should be recom- D+, the remainder of variants as D . Setting up ‘in house’
mended (Engelfriet & Reesink, 2006). Given that RHD*DEL platforms employing allele-specific primers for detecting
is usually linked to RHCE*Ce or RHCE*cE, several contribu- weak D types 1–3 is straightforward.
tors to the forum were in favour of transfusing D C E Probably the most important application of predicting D
red cells to D patients, especially females of childbearing phenotype from genomic DNA, at least in Europe, is for
potential. Testing of 5058 D Danish blood donors for RHD determining fetal D type in D pregnant women with anti-
revealed 13 with RHD, all with DEL phenotype (Krog et al, D, to determine whether the fetus is at risk of HDFN (Dan-
2011). In a subsequent look-back of the 136 transfusions iels et al, 2009). This testing is performed on cell-free fetal
with these red cells to D patients, only one recipient had DNA in the maternal plasma. In Denmark and the Nether-
made anti-D and that patient had also received D+ platelets. lands, non-invasive fetal D typing is offered to all D preg-
Eleven of the recipients had made antibodies to other blood nant women, so that routine antenatal anti-D prophylaxis
groups. The conclusion from this study is that it would be (RAADP) can be withheld from the roughly 40% who have a
very difficult to justify the expense of the introduction of D fetus, thus avoiding unnecessary exposure to blood
screening all D donors for RHD. products (Clausen et al, 2012; de Haas et al, 2012). Large tri-
als of targeted RAADP have also been carried out in Sweden
and England (Finning et al, 2008; Daniels et al, 2012; Wik-
Molecular D typing and D variants
man et al, 2012). In fetal RHD testing, false negative results
Since the cloning of RHD and the elucidation of the molecu- must be avoided as they could result in the pregnant woman
lar bases for the D polymorphism, it has become common failing to receive the appropriate treatment. With the current
practice to predict D phenotype from genomic DNA. The technology, a small level of false positive results is inevitable.
most common molecular background to the D phenotype Apparent false-positive results are around 1–2% (Finning
is homozygosity for a deletion of RHD, so the basis of et al, 2008; Clausen et al, 2012; de Haas et al, 2012). These
molecular D typing is amplifying part of RHD to assess can arise from RHD genes that contain inactivating muta-
whether the gene is present. Relatively common potential tions and so do not produce any D antigens. Some RHD
sources of error come from RHD*Ψ, which has all of the genes responsible for expression of very low levels of D anti-
RHD exons but produces no D antigen, and a variety of gen (DEL phenotype) may appear to give rise to false posi-
RHD–CE–D hybrid genes, which lack part or all of some RHD tive results, because the cord sample gives a D result by
exons, but still produce some D antigen. Most errors arising routine serological testing, but are not in fact false results
from these variants are avoided by amplifying more than one since a D antigen is expressed, though it is very unlikely that
exon of RHD and by including one pair of primers that are such a weak expression of D could immunize a mother
selected not to amplify RHD*Ψ. Alternatively, because through a feto-maternal bleed.
RHD*Ψ contains a 37-bp duplication, any amplification
across the region containing the duplication will give differ-
Conclusion
ent size products from RHD and RHD*Ψ. Any inactivating
mutation in RHD, however, will result in no production of Serological D variants have been known since 1946. Now
D antigen and is likely to give a positive result in molecular over 200 D variants are recognized, though in most cases
testing. Fortunately, such mutations are rare. they can only be differentiated by molecular genetical analy-
BloodChip, a microarray chip-based blood group genotyp- sis. Next generation sequencing, still looming on the horizon
ing platform, developed by a European consortium and as a method of blood grouping, promises to reveal many

ª 2013 Crown copyright, British Journal of Haematology 7

Review

more variants of D and of RHD in the next few years. This which case they are treated as D and referred for molec-
may be valuable from a scientific point of view, but from the ular testing if possible.
clinical aspect it is only necessary to be able to identify weak ● If molecular testing reveals weak D types 1–3, they would
D types 1, 2, and 3 so that they can be treated as D+ in then be treated as D+; otherwise they would be treated as
order to conserve D stocks and avoid waste of anti-D D (Fig. 3).
immunoglobulin. This can be achieved by straightforward ● The equivalent policy is applied for anti-D immunoglobu-
molecular tests. lin prophylaxis in pregnant and post-partum women: if
In summary, I recommend the following procedure. weak D types 1–3, withhold treatment; otherwise offer
treatment.
● Molecular testing for weak D types 1–3 should be made
as widely available as possible.
● Patients whose red cells give abnormal results in routine
Acknowledgements
testing with monoclonal IgM reagents (that give negative
results with DVI cells) be treated as D+ and transfused I am extremely grateful to Clare Milkins, Nicole Thornton,
with D+ red cells unless they are female of 50 years old or Shane Grimsley, and Louise Tilley for their valuable contri-
less or are likely to require chronic transfusion support, in butions in the preparation of this review.

References Castilho, L., Rios, M., Rodrigues, A., Pellegrino, J., Daniels, G., Poole, G. & Poole, J. (2007) Partial D
Saad, S.T.O. & Costa, F.F. (2005) High fre- and weak D: can they be distinguished? Transfu-
AABB. (2011) Standards for Blood Banks and quency of partial DIIIa and DAR alleles found sion Medicine, 17, 145–146.
Transfusion Services, 27th [Link], Bethesda, in sickle cell disease patients suggests increased Daniels, G., Finning, K., Martin, P. & Massey, E.
MD. risk of alloimmunization to RhD. Transfusion (2009) Non-invasive prenatal diagnosis of fetal
Agre, P.C., Davies, D.M., Issitt, P.D., Lamy, B.M., Medicine, 15, 49–55. blood group phenotypes: current practice and
Schmidt, P.J., Treacy, M. & Vengelen-Tyler, V. Chen, W., Li, L., Tsai, S.L. & Lin, K. (2006) Anti- future prospects. Prenatal Diagnosis, 29, 101–107.
(1992) A proposal to standardize terminology D immunization by Del red blood cells in Daniels, G., Finning, K., Wade, A., Massey, E.,
for weak D antigen. Transfusion, 32, 86–87. Taiwan: two case reports. Transfusion, 46(Suppl.), Soothill, P., Phillips, C.J., Martin, W.L. & Chit-
Avent, N.D. (2007) New insight into the Rh system: 129A. ty, L.S. (2012) Implementation of fetal RHD
structure and function. ISBT Science Series, 2, 35–43. Ch
erif-Zahar, B., Bloy, C., Le Van Kim, C., Blan- typing in all RHD-negative women: timing,
Avent, N.D., Ridgwell, K., Tanner, M.J.A. & chard, D., Bailly, P., Hermand, P., Salmon, C., costs, and efficiency. Vox Sanguinis, 103(Sup-
Anstee, D.J. (1990) cDNA cloning of a 30 kDa Cartron, J.P. & Colin, Y. (1990) Molecular clon- pl.1), 34 (abstract).
erythrocyte membrane protein associated with ing and protein structure of a human blood Dogic, V., Bingulac-Popovic, J., Babic, I., Hundric-
Rh (Rhesus)-blood-group-antigen expression. group Rh polypeptide. Proceedings of the Haspl, Z., Jurakovic-Loncar, N., Mratinovic-
Biochemical Journal, 271, 821–825. National Academy of Science USA, 87, 6243– Mikulandra, J., Vuk, T., Balija, M. & Jukic, I.
Avent, N.D., Liu, W., Jones, J.W., Scott, M.L., 6247. (2011) Distribution of weak D types in Croatian
Voak, D., Pisacka, M., Watt, J. & Fletcher, A. Clausen, F.B., Christiansen, M., Steffensen, R., population. Transfusion Medicine, 21, 278–279.
(1997) Molecular analysis of Rh transcripts and Jørgensen, S., Nielsen, C., Jakobsen, M.A., Mad- Engelfriet, C.P., Reesink, H.W. eds (2006) Interna-
polypeptides from individuals expressing the sen, R.D., Jensen, K., Krog, G.R., Rieneck, K., tional forum. Testing for weak D.. Vox Sangui-
DVI variant phenotype: an RHD gene deletion Sprogøe, U., Homburg, K.M., Grunnet, N. & nis, 90, 140–153.
event does not generate all DVIccEe phenotypes. Dziegiel, M.H. (2012) Report of the first nation- Esteban, R., Montero, R., Flegel, W.A., Wagner, F.
Blood, 89, 1779–1786. ally implemented clinical routine screening for F., Subirana, L., Parra, R., Ribera, A. & Nogu
es,
Avent, N.D., Martinez, A., Flegel, W.A., Olsson, fetal RHD in D pregnant women to ascertain N. (2006) The D category VI type 4 allele is pre-
M.L., Scott, M.L., Nogu
es, N., P
ısacka, M., Dan- the requirement for antenantal RhD prophylaxis. valent in the Spanish population. Transfusion,
iels, G., van der Schoot, E., Mu~ niz-Diaz, E., Transfusion, 52, 752–758. 46, 616–623.
Madgett, T.E., Storry, J.R., Beiboer, S.H., Maask- Colin, Y., Ch
erif-Zahar, B., Le Van Kim, C., Faas, B.H.W., Beckers, E.A.M., Wildoer, P., Ligt-
ant-van Wijk, P.A., von Zabern, I., Jim
enez, E., Raynal, V., Van Huffel, V. & Cartron, J.-P. hart, P.C., Overbeeke, M.A., Zondervan, H.A.,
Tejedor, D., L
opez, M., Camacho, E., Cheroutre, (1991) Genetic basis of the RhD-positive and von dem Borne, A.E. & van der Schoot, C.E.
G., Hacker, A., Jinoch, P., Svobodova, I. & de RhD-negative blood group polymorphism as (1997) Molecular background of VS and weak C
Haas, M. (2007) The BloodGen project: towards determined by Southern analysis. Blood, 78, expression in blacks. Transfusion, 37, 38–44.
mass-scale comprehensive genotyping of blood 2747–2752. Finning, K., Martin, P., Summers, J., Massey, E.,
donors in the European Union and beyond. Conroy, M.J., Bullough, P.A., Merrick, M. & Poole, G. & Daniels, G. (2008) Effect of high
Transfusion, 47(1S), 40S–46S. Avent, N.D. (2005) Modelling the human rhesus throughput RHD typing of fetal DNA in maternal
Beckers, E.A.M., Faas, B.H.W., Ligthart, P., Simsek, proteins: implications for structure and func- plasma on use of anti-RhD immunoglobulin in
S., Overbeeke, M.A., von dem Borne, A.E., van tion. British Journal of Haematology, 131, 543– RhD negative pregnant women: prospective feasi-
Rhenen, D.J. & van der Schoot, C.E. (1996) 551. bility study. British Medical Journal, 336, 816–818.
Characterization of the hybrid RHD gene lead- Daniels, G. (2013) Human Blood Groups, 3rd edn. Flegel, W.A. (2011) Molecular genetics and clinical
ing to the partial D category IIIc phenotype. Wiley, Oxford. applications for RH. Transfusion & Apheresis
Transfusion, 36, 567–574. Daniels, G.L., Faas, B.H.W., Green, C.A., Smart, E., Science, 44, 81–91.
Beckers, E.A.M., Ligthart, P.C., Overbeeke, M.A. Maaskant-van Wijk, P.A., Avent, N.D., Zonder- Flegel, W.A., Khull, S.R. & Wagner, F.F. (2000)
M., Maaskant, P.A. & van Rhenen, D.J. (2004) van, H.A., von dem Borne, A.E. & van der Schoot, Primary anti-D immunization by weak D type 2
A patient with weak D type 1 and anti-D: auto C.E. (1998) The Rh VS and V blood group poly- RBCs. Transfusion, 40, 428–434.
or allo? Vox Sanguinis, 87(Suppl. 3), 75 morphisms in Africans: a serological and molecu- Flegel, W.A., Denomme, G.A. & Yazer, M.H.
(abstract). lar analysis. Transfusion, 38, 951–958. (2007) On the complexity of D antigen typing: a

8 ª 2013 Crown copyright, British Journal of Haematology

 Review

handy decision tree in the age of molecular K€

orm€ oczi, G.F., Gassner, C., Shao, C.-P., Uchika- stand in daily practice? Transfusion, 47, 1616–
blood group genetics. Journal of Obstetrics & wa, M. & Legler, T.J. (2005) A comprehensive 1620.
Gynaecology Canada, 29, 746–752. analysis of DEL types: partial DEL individuals Okubo, Y., Seno, T., Yamano, H., Yamaguchi, H.,
Flegel, W.A., von Zabern, I. & Wagner, F.E. (2009) are prone to anti-D alloimmunization. Transfu- Lomas, C. & Tippett, P. (1991) Partial D anti-
Six years’ experience performing RHD genotyp- sion, 45, 1561–1567. gens disclosed by a monoclonal anti-D in Japa-
ing to confirm D red blood cell units in Krog, G.R., Clausen, F.B., Berkowicz, A., Jørgen- nese blood donors. Transfusion, 31, 782.
Germany for preventing anti-D immunizations. sen, L., Rieneck, K., Nielsen, L.K. & Dziegiel, M. Okuda, H., Suganuma, H., Kamesaki, T., Kumada,
Transfusion, 49, 465–471. H. (2011) Is current serologic RhD typing of M., Tsudo, N., Omi, T., Iwamoto, S. & Kajii, E.
Frohn, C., D€ umbgen, L., Brand, J.-M., G€ org, S., blood donors sufficient for avoiding immuniza- (2000) The analysis of nucleotide substitutions,
Luhm, J. & Kirchner, H. (2003) Probability of tion of recipients? Transfusion, 51, 2278–2285. gaps, and recombination events between RHD
anti-D development in D– patients receiving D+ Lacey, P.A., Caskey, C.R., Werner, D.J. & Moulds, and RHCE through complete sequencing. Bio-
RBCs. Transfusion, 43, 893–898. J.J. (1983) Fatal hemolytic disease of a newborn chemical Biophysical Research Communications,
Gardener, G.J., Legler, T.J., Hyett, J.A., Liew, Y.-W., due to anti-D in an Rh-positive Du variant 274, 670–683.
Flower, R.L. & Hyland, C.A. (2012) Anti-D in preg- mother. Transfusion, 23, 91–94. Omi, T., Takahashi, J., Tsudo, N., Okuda, H.,
nant women with the RHD(IVS3+1G>A)-associ- Liu, H.-C., Eng, H.-L., Yang, Y.-F., Wang, Y.H., Iwamoto, S., Tanaka, M., Seno, T., Tani, Y. &
ated DEL phenotype. Transfusion, 52, 2016–2019. Lin, K.T., Wu, H.L. & Lin, T.M. (2010) Aber- Kajii, E. (1999) The genomic organization of the
Gassner, C., Doescher, A., Drnovsek, T.D., Roz- rant RNA splicing in RHD 7–9 exons of DEL partial D category DVa: the presence of a new
man, P., Eicher, N.I., Legler, T.J., Lukin, S., Gar- individuals in Taiwan: a mechanism study. partial D associated with the DVa phenotype.
ritsen, H., Kleinrath, T., Egger, B., Ehling, R., Biochimica Biophysica Acta, 1800, 565–573. Biochemical Biophysical Research Communica-
K€orm€ oczi, G.F., Kilga-Nogler, S., Schoenitzer, D. Lowes, B.C.R. (1969) Hemolytic disease of the tions, 254, 786–794.
& Petershofen, E.K. (2005) Presence of RHD in newborn due to anti-Rh14 (RhB). Vox Sanguinis, Østgard, P., Fevang, F. & Kornstad, L. (1986)
serologically D–, C/E+ individuals: a European 16, 231–232. Anti-D in a ‘D positive’ mother giving rise to
mulitcenter study. Transfusion, 45, 527–538. Lubenko, A., Contreras, M. & Habash, J. (1989) severe haemolytic disease of the newborn. Acta
Gonzalez-Porras, J.R., Graciani, I.F., Perez-Simon, J. Should anti-Rh immunoglobulin be given to D Paediatrica Scandinavica, 75, 175–178.
A., Martin-Sanchez, J., Encinas, C., Conde, M.P., variant women? British Journal of Haematology, Pelc-Klopotowska, M., Guz, K., Walaszcyk, A.,
Nieto, M.J. & Corral, M. (2008) Prospective eval- 72, 429–433. Orzinska, A. & Michalewska, B. (2010) Weak D
uation of a transfusion policy of D+ red blood Mayne, K., Bowell, P., Woodward, T., Sibley, C., type 2 patient with allo anti-D. Vox Sanguinis,
cells into D patients. Transfusion, 48, 1318–1324. Lomas, C. & Tippett, P. (1990) Rh immuniza- 99(Suppl. 1), 367 (abstract).
Gruswitz, F., Chaudhary, S., Ho, J.D., Schlessinger, tion by the partial D antigen of category DVa. Pham, B.-N., Peyrard, T., Juszczak, G., Dubeaux,
A., Pezeshki, B., Ho, C.M., Sali, A., Westhoff, C. British Journal of Haematology, 76, 537–539. I., Gien, D., Blancher, A., Cartron, J.P., Rouger,
M. & Stroud, R.M. (2010) Function of human Milkins, C., Berryman, J., Cantwell, C., Elliott, C., P. & Le Pennec, P.Y. (2009) Heterogeneous
Rh based on structure of RhCG at 2.1  A. Pro- Haggas, R., Jones, J., Rowley, M., Williams, M. & molecular background of the weak C, VS+,
ceedings of the National Academy of Sciences Win, N. (2013) Guidelines for pre-transfusion hrB , HrB phenotype in black persons. Trans-
USA, 107, 9638–9643. compatibility procedures in blood transfusion fusion, 49, 495–504.
de Haas, M., van der Ploeg, C.P.B., Scheffer, P.G., laboratories. Transfusion Medicine, 23, 3–35. Also see: Pham, B.-N., Roussel, M., Peyrard, T., Beolet, M.,
Verlinden, D.A., Hirschberg, H., Abbink, F. & van [Link] Jan-Lasserre, V., Gien, D., Ripaux, M., Bourgouin,
der Schoot, C.E. (2012) A nation-wide fetal RHD [Link]. S., Kappler-Gratias, S., Rouger, P. & Le Pennec,
screening programme for targeted antenatal and Mota, M., Fonseca, N.L., Rodrigues, A.,
Kutner, J. P.Y. (2011) Anti-D investigations in individuals
postnatal anti-D. ISBT Science Series, 7, 164–167. M. & Castilho, L. (2005) Anti-D alloimmuniza- expressing weak D type 1 or weak D type 2:
Haer-Wigman, L., Veldhuisen, B., Jonkers, R., tion by weak D type 1 red blood cells with a very allo- or autoantibodies? Transfusion, 51, 2679–
Lod
en, M., Madgett, T.E., Avent, N.D., de Haas, low antigen density. Vox Sanguinis, 88, 130–135. 2685.
M. & van der Schoot, C.E. (2013) RHD and Mouro, I., Le Van Kim, C., Rouillac, C., van Rhe- Rodrigues, M.J., Rodrigues, F., Tilley, L., Poole, J.,
RHCE variant and zygosity genotyping via multi- nen, D.J., Le Pennec, P.Y., Bailly, P., Cartron, J. Chabert, T. & Sousa, G. (2006) Several new
plex ligation-dependent probe amplification. P. & Colin, Y. (1994) Rearrangements of the examples of weak D type 38 in the Portuguese
Transfusion, doi: 10.1111/j.1537-2995.2012.03919. blood group RhD gene associated with the DVI population. Transfusion, 46(Suppl.), 141A–142A
x. [Epub ahead of print]. category phenotype. Blood, 83, 1129–1135. (abstract).
Hemker, M.B., Ligthart, P.C., Berger, L., van Rhe- M€uller, T.H., Wagner, F.F., Trockenbacher, A., Rouillac, C., Colin, Y., Hughes-Jones, N.C., Beolet,
nen, D.J., van der Schoot, C.E. & Maaskant-van Eicher, N.I., Flegel, W.A., Sch€
onitzer, D., Schunter, M., D’Ambrosio, A.M., Cartron, J.P. & Le Van
Wijk, P.A. (1999) DAR, a new RhD variant F. & Gassner, C. (2001) PCR screening for com- Kim, C. (1995) Transcript analysis of D category
involving exons 4, 5, and 7, often in linkage mon weak D types shows different distributions phenotypes predicts hybrid Rh D-CE-D proteins
with ceAR, a new Rhce variant frequently found in three Central European populations. Transfu- associated with alteration of D epitopes. Blood,
in African blacks. Blood, 94, 4337–4342. sion, 41, 45–52. 85, 2937–2944.
Hill, Z., Vacl, J., Kalasov
a, E., Cal
abkov
a, M. & Needs, M., Poole, J., Tilley, L., Stern, S., Preddy, I. Roxby, D., Coloma, M., Flegel, W.A., Poole, J.,
Pintera, J. (1974) Haemolytic disease of new- & Win, N. (2007) Stimulation of anti-D in preg- Martin, P. & Abbott, R. (2004) Observation of
born due to anti-D antibodies in a Du-positive nancy by a weak D type 3 foetus. Transfusion anti-D after D-positive transfusion in an indi-
mother. Vox Sanguinis, 27, 92–94. Medicine, 17(Suppl.1), 40 (abstract). vidual with weak D type-1 phenotype. Vox San-
Hyodo, H., Ishikawa, Y., Kashiwase, K., Kashiwase, Noizat-Pirenne, F. & Tournamille, C. (2011) guinis, 87(Suppl. 3), 77–78.
K., Toyoda, C., Uchikawa, M., Akaza, T., Fujii, Relevance of RH variants in transfusion of sickle Scott, M. (2002) Section 1A. Rh serology. Coordi-
T., Kozuma, S., Taketani, Y. & Juji, T. (2000) cell patients. Transfusion Clinique Biologique, 18, nators report. Transfusion Clinique Biologique, 9,
Polymorphisms of RhDVa and a new RhDVa-like 527–535. 23–29.
variant found in Japanese individuals. Vox San- Noizat-Pirenne, F., Verdier, M., Lejealle, A., Shao, C.-P. (2010) Transfusion of RhD-positive
guinis, 78, 122–125. Mercadier, A., Bonin, P., Peltier-Pujol, F., Fia- blood in “Asia type” DEL recipients. New Eng-
Klein, H.K. & Anstee, D.J. (2005) Mollison’s Blood laire-Legendre, A., Tournamille, C., Bierling, P. land Journal of Medicine, 362, 472–473.
Transfusion in Clinical Medicine, 11th edn. & Ansart-Pirenne, H. (2007) Weak D pheno- Shao, C.-P., Wen, X. & Zhou, Y.-Y. (2006) Multi-
Blackwell Publishing, Oxford, UK. types and transfusion safety: where do we ple isoforms excluding normal RhD mRNA

ª 2013 Crown copyright, British Journal of Haematology 9

Review

detected in Rh blood group Del phenotype with M. (2011) Weak D type 2 and production of anti- associated with the low-incidence antigen Rh32.
RHD 1227A allele. Transfusion and Apheresis D. Transfusion, 51(Suppl. 3), 41A (abstract). Transfusion Medicine, 7, 233–238.
Science, 34, 142–152. Wagner, F.F. & Flegel, W.A. (2000) RHD gene Westhoff, C.M., Vege, S., Halter-Hipsky, C.,
Silvy, M., Di Cristofaro, J., Beley, S., Papa, K., Rits, deletion occurred in the Rhesus box. Blood, 95, Whorley, T., Hue-Roye, K., Lomas-Francis, C. &
M., Richard, P., Chiaroni, J. & Bailly, P. (2011) 3662–3668. Reid, M.E. (2010) DIIIa and DIII Type 5 are
Identification of RHCE and KEL alleles in large Wagner, F.F., Gassner, C., M€ uller, T.H., Sch€
onit- encoded by the same allele and are associated
cohorts of Afro-Caribbean and Comorian zer, D., Schunter, F. & Flegel, W.A. (1998) with altered RHCE*ce alleles: clinical implica-
donors by multiplex SNaPshot and fragment Three molecular structures cause Rhesus D cate- tions. Transfusion, 50, 1303–1311.
assays: a transfusion support for sickle cell dis- gory VI phenotypes with distinct immunohema- White, C.A., Stedman, C.M. & Frank, S. (1983)
ease patients. British Journal of Haematology, tologic features. Blood, 91, 2157–2168. Anti-D antibodies in D– and Du-positive
154, 260–270. Wagner, F.F., Gassner, C., M€ onit-
uller, T.H., Sch€ women: a cause of hemolytic disease of the new-
Singleton, B.K., Green, C.A., Avent, N.D., Martin, zer, D., Schunter, F. & Flegel, W.A. (1999) born. American Journal of Obstetrics & Gynecol-
P.G., Smart, E., Daka, A., Narter-Olaga, E.G., Molecular basis of weak D phenotypes. Blood, ogy, 145, 1069–1075.
Hawthorne, L.M. & Daniels, G. (2000) The pres- 93, 385–393. Wiener, A.S. & Unger, L.J. (1962) Further observa-
ence of an RHD pseudogene containing a 37 Wagner, F.F., Frohmajer, A., Ladewig, B., Eicher, tions on the blood factors RhA, RhB, RhC and
base pair duplication and a nonsense mutation N.I., Lonicer, C.B., M€uller, T.H., Siegel, M.H. & RhD. Transfusion, 2, 230–233.
in most Africans with the Rh D-negative blood Flegel, W.A. (2000) Weak D alleles express dis- Wikman, A.T., Tiblad, E., Karlsson, A., Olsson, M.
group phenotype. Blood, 95, 12–18. tinct phenotypes. Blood, 95, 2699–2708. L., Westgren, M. & Reilly, M. (2012) Noninva-
St-Louis, M., Richard, M., C^

ot
e, M., Ethier, C. & Wagner, F.F., Frohmajer, A. & Flegel, W.A. (2001) sive single-exon fetal RHD determination in a
Long, A. (2011) Weak D type 42 cases found in RHD positive haplotypes in D negative Europe- routine screening program in early pregnancy.
individuals of European descent. Immunohema- ans. BMC Genetics, 2, 10. Obstetrics & Gynecology, 120, 227–234.
tology, 27, 20–24. Wagner, F.F., Moulds, J.M., Tounkara, A., Kouri- Yazdanbakhsh, K., Ware, R.E. & Noizat-Pirenne, F.
Tippett, P. & Sanger, R. (1962) Observations on ba, B. & Flegel, W.A. (2003) RHD allele distri- (2012) Red blood cell alloimmunization in sickle
subdivisions of the Rh antigen D. Vox Sanguinis, bution in Africans of Mali. BMC Genetics, 4, 14. cell disease: pathophysiology, risk factors, and
7, 9–13. Wagner, T., K€ orm€oczi, G.F., Buchta, C., Vadon, transfusion management. Blood, 120, 528–537.
UK Blood Transfusion and Tissue Transplantation M., Lanzer, G., Mayr, W.R. & Legler, T.J. (2005) Yazer, M.H. & Triulzi, D.J. (2007) Detection of
Services. (2005) Guidelines for the Blood Trans- Anti-D immunization by DEL red blood cells. anti-D in D– recipients transfused with D+ red
fusion Service in the UK (7th edn.). [Link] Transfusion, 45, 520–526. blood cells. Transfusion, 47, 2197–2201.
[Link]/[Link]?Pu Wallace, M., Lomas-Francis, C., Beckers, E., Bruce, Yu, X., Wagner, F.F., Witter, B. & Flegel, W.A.
blication=RB. Chapter 13. M., Campbell, G., Chatfield, S., Nagao, N., (2006) Outliers in RhD membrane integration
Vege, S., Fong, C., Lomas-Francis, C., Nickle, P., Okubo, Y., Opalka, A., Overbeeke, M., Scott, M. are explained by variant RH haplotypes. Trans-
Horn, T., Delaney, M., Meny, G. & Westhoff, C. & Voak, D. (1997) DBT: a partial D phenotype fusion, 46, 1343–1351.

10 ª 2013 Crown copyright, British Journal of Haematology