Original Article

Strategies to identify candidates for D variant genotyping

Xunda Luo1, Margaret A. Keller2, Ian James1, Michelle Grant1, Shiguang Liu1, Kellie Simmons Massey1,
Andrew Czulewicz3, Sandra Nance2,4, Yanhua Li5

1
Department of Pathology and Laboratory Medicine, Temple University Hospital; 2American Red Cross;
3
Department of Pathology, Jeanes Hospital; 4University of Pennsylvania, Department of Pathology and Laboratory
Medicine, Division of Transfusion Medicine; 5Department of Pathology and Laboratory Medicine, Lewis Katz
School of Medicine at Temple University, Philadelphia, PA, United States of America

Background. RhD variants have altered D epitopes and/or decreased antigen copies per red cell.
Individuals carrying these variants may test antigen negative, weakly positive, or positive by serology,
and may or may not be at risk of alloimmunisation after exposure. There have been recommendations
to perform RHD genotyping of patients, pregnant women and females of childbearing potential with
serological weak D phenotype, to guide prophylactic use of Rh immune globulin (RhIG), and better

l
conserve D-negative blood products. The purpose of this study was to evaluate the performance of

Sr
a set of empirical criteria to identify such patients.
Materials and methods. A two-method strategy of gel testing (GT) and tube testing (TT) was
used for Rh typing of patients with no historical blood type in the present institution. A monoclonal-

zi
polyclonal blend anti-D was used for Rh typing by TT at immediate spin. Three empirical criteria
were used to identify candidates for genotyping: C1: discrepancy between the two test methods and

i
a GT reaction strength >2+ stronger than TT; C2: weak serological reaction, defined as reaction
rv
strength ≤2+ regardless of testing method if both GT and TT were performed or reaction strength
≤2+ if only GT was performed, or reaction strength ≤1+ if only TT was performed; C3: presence of
Se
anti-D in D-positive patients with no history of RhIG use in the preceding 3 months and in whom
alloanti-D is suspected.
Results. Overall, 50 patients, ranging from newly born to 93 years old, were identified. Genomic
testing confirmed D variants in 49/50 cases with a positive predictive value of 98%.
TI

Discussion. This two-method strategy is a powerful screening tool for identifying candidates for
RHD genotyping. This strategy meets the current requirements of two blood type determinations/two
specimens in pre-transfusion testing while simultaneously identifying candidates for RHD genotyping
M

with a minimal increase in work load and cost.

SI

Keywords: RhD phenotyping, RHD genotyping, Rh immunoglobulin, weak D, partial D.

©

Introduction agglutination with antihuman globulin testing5-9. Weak

Rh is the second most important clinically significant D variants express amino acid substitutions in the
blood group system in transfusion medicine after the intramembranous and/or cytoplasmic regions of RhD
ABO system. Since the initial discovery of RhD1, which affect antigen expression and/or exposure on
more than 50 antigens within the Rh system have the red blood cell surface7,11. Of the more than 70 weak
been identified2. RHD and RHCE are closely linked D variants, there are sufficient data showing lack of
homologous genes that encode these Rh group antigens3. alloanti-D formation for the three most common D
More than 200 RHD variants have been recognised to variants, weak D types 1, 2, and 3, to support not giving
date4,5. carriers prophylaxis12,13, whereas this is not the case for
D variants are sorted into three major categories, certain other D variants11,12,14-16.
weak D, partial D, and DEL, based on genotypes Partial D is defined as incomplete expression
and potential for making alloanti-D. D variants can of the multiple RhD epitopes on the red blood cell
be typed positive, negative or weak. A serological surface. Variants in this category include those with
weak D phenotype is defined as no or weak (≤2+) red amino acid and exon substitutions in the extracellular
blood cell reactivity to an anti-D reagent with non- region of the RhD protein which destroy one or more
antihuman globulin testing but moderate to strong epitopes11. The strength of red cell reactivity to an

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

© SIMTI Servizi Srl
All rights reserved - For personal use only 293
No other use without premission
 Luo X et al

anti-D reagent depends on the epitope specificity of Materials and methods

the reagent, which is the rationale behind partial D Institution and subjects
phenotyping using monoclonal anti-D reagent panels. This study was conducted at a tertiary medical
It is possible for partial D recipients to be D-positive centre with approximately 15,000 red blood cell
in serological testing yet still produce alloanti-D unit transfusions, 2,300 births, and 14,500 surgical
following transfusion of D-positive red blood cells17-21, operations annually. The main hospital does not serve
presumably by mounting a reaction against D epitopes paediatric patients; however, the Transfusion Service
that they lack. provides its services to a children's hospital with about
A recent College of American Pathologists (CAP) 1,200 operations per year. In addition, annually, the
Transfusion Medicine Resource Committee survey Transfusion Service serves about 10,000 new patients
found that in the United States, most transfusion with no prior ABO/Rh type results in the electronic
service laboratories do not use indirect antihuman record system. All new patients met inclusion criteria
globulin testing on potential blood product recipients for this study.
and women of childbearing age in order to avoid For the retrospective phase of this study, all available
detecting serological weak D individuals in this RHD genotype reports on file since 2006 were reviewed
population and identifying them as D-negative22. to identify qualified cases that met at least one of the
For blood donors and neonates, however, the AABB screening criteria for RHD genotyping (14 cases).

l
standards require that laboratories test for serological Since January 2015, all cases that met one or more of

Sr
weak D and interpret positive tests as D-positive in the criteria (36 cases) were enrolled in the prospective
order to minimise exposure of D antigen-negative phase of the present study, and the specimens were sent
subjects to weak D-positive red blood cells via for RHD genotyping.

zi
transfusion, and to administer Rho(D) immune
globulin (RhIG) to D-negative mothers of weak D Serological RhD phenotyping
babies23.
i In order to comply with AABB Standards for
rv
Based on the results of this survey and a systemic Blood Banks and Transfusion Services 23, ABO/Rh
review of literature, a Work Group formed by the AABB typing is performed twice on the first specimen for
Se
and CAP, with scientific consultants knowledgeable in all new patients. To comply with CAP requirements,
RHD genetics, developed the following recommendations a second specimen is required prior to transfusion
aimed to change clinical management of serological for non-emergency patients. Prior to late 2011, all
weak D patients: (i) RHD genotyping should be Rh testing was performed using a monoclonal-
TI

performed when an inconsistency in RhD typing polyclonal blend anti-D (BioClone; Ortho Clinical
results and/or a serological weak D phenotype is Diagnostics, Raritan, NJ, USA) by tube testing (TT),
identified; (ii) recipients with weak D type 1, 2, and 3 with additional weak D testing (anti-globulin phase)
M

alleles should be reported and managed as D-positive as needed. In late 2011, gel testing ([GT]; Ortho
individuals; (iii) women of childbearing age with weak ProVue; Ortho Clinical Diagnostics) was implemented
SI

D phenotypes who are not type 1, 2, or 3 should receive as the primary method for ABO/Rh typing and antibody
RhIG prophylaxis. The Work Group posited that by screening. Since June 2012, both methods (GT and
implementing these recommendations, RhIG use would TT) have been used for ABO/Rh typing on the first
©

be minimised and D-negative units better preserved specimens from new patients except for neonates
without increasing the incidence of alloimmunisation (see text below about neonatal testing). The first
to the D antigen 10. The multi-organisational Joint ABO/Rh typing is performed by GT, which is
Statement recently published on the AABB website also followed by a second ABO front typing and Rh
recommends RHD genotyping for pregnant women and typing (immediate spin phase only) by TT. Cases
other females of childbearing potential with a serological that met one or more of the criteria below were sent
weak D phenotype24. for RHD genotyping.
It is of primary importance to establish practical According to the manufacturer's instructions,
criteria to identify candidates for RHD genotyping to EDTA plasma is required for GT. In our institution,
implement these recommendations. Although the Work all cord blood specimens are collected into red-top
Group suggests using a discordant RhD typing result as a tubes without anticoagulant. In addition, heel-stick
screening test10, consistent criteria to identify discordant specimens are usually of small quantity (~0.3 mL
serological types for RHD genotyping are currently on average) and not sufficient for GT by ProVue.
lacking. The purpose of the present study is to evaluate Therefore, for neonates, only the TT method is used for
the performance of a set of empirical criteria to identify both the initial and the second ABO/Rh typing, using
candidates for RHD genotyping. the same specimen.

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

294 All rights reserved - For personal use only
No other use without premission
Candidates for RHD genotyping

Screening criteria for RHD genotyping prepared using Superscript III (ThermoFisher, Waltham,
A previous study showed that GT performs as well MA, USA), RHD cDNA was amplified and Sanger
as TT in terms of serological RhD phenotyping 25. sequencing was performed (P40).
Unpublished data from the present institution also
suggest that red blood cells carrying D variants tend Results
to react more strongly with GT than with TT. Based A total of 50 cases (32 women and 18 men), ranging
on this observation, the following empirical criteria from a neonate (P35) to a 93-year old (mean ± standard
were developed and used to identify candidates for deviation [SD]: 40±19 years) were identified. Fourteen
RHD genotyping in the present study: (criterion 1: of these were identified through retrospective review,
C1) discrepancy between the two testing methods: and the remaining 36 cases prospectively (Table I).
GT reaction strength at least 2+ stronger than TT; Twenty-seven cases were African-Americans, 16 were
(criterion 2: C2) serological weak reaction defined as: Hispanics, six were Caucasians, and one was of unknown
agglutination observed (i.e. reactive), however reaction race/ethnicity. This race/ethnicity constitution was not
strength ≤2+ regardless of testing method if both GT statistically significantly different from the estimated
and TT were performed, or reaction strength ≤2+ if only race/ethnicity constitution of the patient population at
GT was performed, or reaction strength ≤1+ if only TT the institution (based on 2015 data) (χ2=10.3, degrees
was performed, which is different from the Work Group of freedom [DF]=5, p=0.066; Table II). Twenty-three

l
recommendations; (criterion 3: C3) presence of anti-D cases met C1 only, nine met C2 only, three met C3

Sr
in D-positive patients with no history of RhIG use in the only, 13 met both C1 and C2, one met both C1 and C3,
preceding 3 months and in whom alloanti-D is suspected. and one met all three criteria. Genotyping confirmed
For C2, the reaction strength cut-off was established at D variants in 49 cases (10 [20%] with weak D and 39

zi
≤2+ if only GT was performed because it was presumed [78%] with partial D), with a positive predictive value
that the TT reaction would not be stronger than the GT of 98% (49/50) (Table I).
one, whereas the cut-off was set at ≤1+ if only TT was
i Among the 39 partial D subjects, 20 (51.3%) met
rv
performed so if a second GT were to be performed with C1 only, four (10.3%) met C2 only, two (5.1%) met C3
the same specimen, C2 would still be satisfied if the GT only, 11 (28.2%) met both C1 and C2, one (2.6%) met
Se
reaction strength were ≤2+, or C1 would be satisfied if both C1 and C3, and one (2.6%) met all three criteria.
the GT reaction strength were >2+. By definition, C2 is GT reactivity among those with partial D variants ranged
not met if only TT is performed and no agglutination is from 1+ to 4+, whereas TT reactivity was mostly 1+ or
observed (i.e. non-reactive), since a negative TT by itself non-reactive, with one 2+ (P1, who met C1) and one
TI

does not favour weak D/partial D over a true D-negative. 4+ only (P38, who met C3). The greatest discrepancy
in reactivity seen in this group of patients was 4+ in GT
RHD genotyping and non-reactive (-) in TT (P5). Within the ten weak D
M

All RHD genotyping was performed by a reference subjects, two (20%) met C1 only, five (50%) met C2
molecular laboratory. Genomic DNA was isolated only, none met C3 only, three (30%) met C1 and C2,
SI

from peripheral blood using standard methodology. none met C1 and C3, and none met all three criteria. GT
RHD genotyping included testing by RHD BeadChip reactivity among these subjects ranged from 1+ to 3+,
(Immucor, Norcross, GA, USA)26, multiplex polymerase whereas TT reactivity was mostly 1+ or non-reactive.
©

chain reaction (PCR) for C, c, RHD exon 4, exon 7 The greatest discrepancy in reactivity seen in this group
and RHD pseudogene27, hybrid Rhesus Box PCR28 of patients was a case that was 3+ in GT and non-reactive
and PCR-restriction fragment length polymorphism in TT. There was a statistically significant difference
with PstI for RHD deletion29. PCR-restriction fragment between partial D and weak D cases in percentage
length polymorphism for RHD c.667, c.697, c.1136 and of criteria met (χ2=18.3, DF=5, p=0.003; Table III).
sequence-specific primer PCR for c.455 were performed Reaction strength discrepancy (C1) was more likely to be
as described previously30 with restriction enzymes from observed among partial D subjects (84.6% in partial D vs
New England Biolabs (Ipswich, MA, USA). Sanger 50% in weak D), whereas serological weak D phenotype
sequencing of RHD exons was performed if necessary (C2) was more prevalent among weak D subjects (41.0%
to assign the alleles and to resolve indeterminate calls in partial D vs 80% in weak D), as the nomenclature of
generated by RHD BeadChip (P14, P20, P31, P33, P39) weak D would suggest (Tables I and III).
using primers as described30.
When no variants were detected by targeted genomic Strong reactivity
testing and a D variant was strongly suspected, total Table I shows that 25 (P1-P21, P33, P39, P41, P42)
RNA was isolated from red cells using an RNeasy out of 49 D variant cases demonstrated strong reactivity
Minikit (Qiagen, Carlsbad, CA, USA), cDNA was (3+ or 4+) in GT, but weak reactivity in TT except for

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

All rights reserved - For personal use only 295
No other use without premission
 Luo X et al

Table I - Patients' characteristics.

Patients Age Gender Race/ GT TT Reason for Rh RHD genotype
(years) ethnicity genotyping phenotype
P1 52 F B 4+ 2+ C1 Partial D RHD*DAU/RHD*pseudogene
P2 31 F B 4+ 1+ C1 Partial D RHD* weak partial D 4.0 (hemizygous or
homozygous)
P3 55 F B 4+ 1+ C1 Partial D RHD*DAU5 / RHD*pseudogene (Ψ)
P4 27 F H 4+ 1+ C1 Partial D RHD*DAU5 (hemizygous or homozygous)
P5 52 M B 4+ 0 C1 Partial D RHD*weak partial D 4.0 / RHD*01N.01
P6 24 F B 3+ 1+ C1 Partial D RHD*DAU4 (hemizygous or homozygous)
P7 23 F B 3+ 1+ C1 Partial D RHD*DAU5 (hemizygous or homozygous)
P8 22 M B 3+ 1+ C1 Partial D RHD*DAU5 / RHD*DV type 1
P9 33 M B 3+ 1+ C1 Partial D RHD*DAU5 / RHD*ᴪ
P10 57 F B 3+ 1+ C1 Partial D RHD*pseudogene / RHD*weak partial D 4.0
P11 49 M B 3+ 1+ C1 Partial D RHD*weak partial D 4.0 / RHD*DIIIa-CE(4-7)-D

l
P12 25 F H 3+ 1+ C1 Partial D RHD*DAR (hemizygous or homozygous)

Sr
P13 50 M H 3+ 1+ C1 Partial D RHD*DAR (hemizygous or homozygous)
P14 49 F H 3+ 1+ C1 Partial D RHD*DOL1 or RHD*DOL2 (hemizygous or
homozygous)

zi
P15 59 F H 3+ 1+ C1 Partial D RHD*weak partial D 4.0 (hemizygous or
homozygous)
P16 43 M U 3+ 1+
i C1 Partial D RHD*DAU5 (hemizygous or homozygous)
rv
P17 42 F B 3+ 0 C1 Partial D RHD*DAR (hemizygous or homozygous)
P18 19 M B 3+ 0 C1 Partial D RHD*DAU5
Se
P19 63 M H 3+ 0 C1 Partial D RHD*DAR / RHD*01N.01
P20 23 F H 3+ 0 C1 Partial D RHD*DAU4 / RHD*01N.01
P21 42 F W 3+ 0 C1 Partial D RHD*weak D Type 4.1 / RHD*01N.01
TI

P22 37 F B 2+ 0 C1, C2 Partial D RHD*DIIIa-CE(4-7)-D / RHD*DAR

P23 85 M B 2+ 0 C1, C2 Partial D RHD*DAR
M

P24 21 F B 2+ 0 C1, C2 Partial D RHD*DAR (hemizygous or homozygous)

P25 19 F B 2+ 0 C1, C2 Partial D RHD*DAR / RHD*DIIIa CE(4-7)
SI

P26 64 M B 2+ 0 C1, C2 Partial D RHD*DAR1 (hemizygous or homozygous)

P27 60 M B 2+ 0 C1, C2 Partial D RHD*DAU5 (hemizygous or homozygous)
P28 33 F B 2+ 0 C1, C2 Partial D RHD*DIIIa-CE(4-7)-D / RHD*DAR
©

P29 35 F B 2+ 0 C1, C2 Partial D RHD*IIIa/RHD*weak partial D 4.0 or RHDIIIa-

CE(4-7)-D/RHD*weak partial D 4.0
P30 28 F H 2+ 0 C1, C2 Partial D RHD*DAR1 (hemizygous or homozygous)
P31 24 F H 2+ 0 C1, C2 Partial D RHD*DSC1 (hemizygous or homozygous)
P32 42 F B 2+ 0 C1, C2, C3 Partial D RHD*DAR (hemizygous or homozygous)
P33 62 M H 4+ 1+ C1, C3 Partial D RHD*DSC1 or RHD*DFV
P34 25 M B NP 1+ C2 Partial D RHD*DAU5 (hemizygous or homozygous)
P35 NB F B NP 1+ C2 Partial D RHD*DIII*weak partial D 4.0
P36 47 M B 2+ 1+ C2 Partial D RHD*DAR (hemizygous or homozygous)
P37 29 F B 1+ 0 C2 Partial D RHD*DAU5 / RHD*DIIIa-CE(4-7)-D
P38 22 F H NP 4+ C3 Partial D RHD*DIVa / RHD*01N.01
P39 29 F H 3+ NP C3 Partial D RHD*DIVa / RHD*01N.01
P40 21 F H 1+ 0 C2 Weak D RHD*weak D type 61
Continued on next page

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

296 All rights reserved - For personal use only
No other use without premission
Candidates for RHD genotyping

Table I - Patients' characteristics (continued from previous page).

Patients Age Gender Race/ GT TT Reason for Rh RHD genotype
(years) ethnicity genotyping phenotype
P41 53 F H 3+ 1+ C1 Weak D RHD*weak D type 3 (hemizygous or homozygous)
P42 40 F B 3+ 0 C1 Weak D RHD*weak D type 1
P43 54 M H 2+ 0 C1, C2 Weak D RHD*weak D type 2 (hemizygous or homozygous)
P44 64 F W 2+ 0 C1, C2 Weak D RHD*weak D type 1
P45 27 F W 2+ 0 C1, C2 Weak D RHD*weak D type 2 (hemizygous or homozygous)
P46 26 F H NP 1+ C2 Weak D RHD*weak D type 3 / RHD*01N.01
P47 57 M W 2+ NP C2 Weak D RHD*Weak D type 2 / RHD*01N.01
P48 18 F B 1+ 0 C2 Weak D RHD*weak D type 2 / RHD*01N.01
P49 55 M W 1+ 0 C2 Weak D RHD*weak D type 2 (hemizygous or homozygous)
P50 93 M W 4+ 4+ C3 D-positive RHD*weak D type 3 / RHD*01
F: female; M: male; GT: gel testing; TT: tube testing; A: Asian/Pacific Islander; B: Africa-American/Black; H: Hispanic; W: White; U: unknown; NB: newborn;
NP: not performed; 0: no agglutination observed.

l
Sr
Table II - Race/ethnicity of the 50 cases in the present study.
Race/ Number of cases Percentage of cases Percentage in the patient population Statistics
ethnicity who met the criteria who met the criteria of the present institution (2015 data)

zi
A 0 0.00 1.40

B 27 54.00 39.72

i
H 16
rv 32.00 25.16 χ2=10.3
DF=5
W 6 12.00 25.03 p=0.066
Se
U 1 2.00 4.85

O 0 0.00 3.84

Total 50 100 100

TI

A: Asian/Pacific Islander; B: Africa-American/Black; H: Hispanic; W: White; U: unknown; O: others; DF: degrees of freedom.

Table III - Criteria (C1-C3) met by the two different Negative tube testing results
M

categories of D variant carriers. In the present study, 25 cases (P5, P17-P32, P37,
Partial D Weak D Statistics P40, P42-P45, P48, P49) out of 50 (50%) were non-
SI

reactive with TT, and would have been missed and

C1 51.28 20.00
managed as D negative, had GT not been performed
C2 10.26 50.00 as a second method. Seven of these cases (7/24, 29%)
©

C3 5.13 0.00 χ2=18.3 were weak D. Among these weak D cases, six cases (6/7,
DF=5
C1, C2 28.21 30.00 p=0.003
85.71%; P42-P45, P48, P49) were type 1, 2 or 3, and
the remaining one (P40) was weak D type 61 (Table I).
C1, C3 2.56 0.00

C1, C2, C3 2.56 0.00 Cases with anti-D

DF: degrees of freedom. RHD genotyping was performed in five cases (P32,
P33, P38, P39, P50) because C3 was met (alone or in
P39 (TT not performed). These cases would have been combination with other criteria), i.e., anti-D was present
categorised as D positive if only GT had been used for in a D-positive patient with no history of RhIG use in the
typing, unless an allo-anti-D had been identified. Partial preceding 3 months in whom allo-anti-D was suspected
D variants were identified in 23 of these cases (23/25, (Table IV). The auto control was weakly positive (1+) by
92%), whereas weak D variants were identified in two GT in two cases (P32, P50). The direct antiglobulin test
cases (2/25, 8%; P41, P42) only (Table I). Twenty-three (DAT) by TT was negative and, therefore, no eluate or
of these cases (23/25, 92%; P1-P21, P41, P42) were autoadsorption was performed on these cases according
identified because two methods were used and C1 was to laboratory procedures. However, we recognise that for
met. a case of positive auto control by GT, the DAT should be

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

All rights reserved - For personal use only 297
No other use without premission
 Luo X et al

Table IV - Characteristics of the patients expressing anti-D.

Patient Age (years)/ Interpretation Alloanti-D Obstetric
Transfusion RhIG within Anti-D Autocontrol DAT Eluate/auto-
gender propensity history within preceding status by adsorption
preceding 3 months TT
3 months
P32 42/F Partial D DAR G3/P2/A1 Yes, 5 weeks No Present GT: 1+ 0 NP/NP
previously ago PEG: 0
reported7
P33 62/M Partial D DFV NA No No Present 0 0 NP/NP
previously
reported7
P38 22/F Partial D DIVa G4/P2/A2 Yes No Present 0 0 NP/NP
previously
reported7
P39 29/F Partial D DIVa G8/P4/A3 No No Present 0 0 NP/NP
previously
reported7
P50 93/M D-positive NA NA No No Present GT: 1+ 0 NP/NP
DAT: direct antiglobulin test; TT: tube testing; GT: gel testing; NA: not applicable; NP: not performed; 0: not reactive; G/P/A: gravida/para/abortus; PEG:
polyethylene glycol.

l
Sr
performed in GT and should also be positive. However, whereas to date, no regulations require the use of two
TT is the only method validated in our laboratory for different methods for RhD typing. In addition, all current
DAT; GT is not validated for DAT in our laboratory and RhD serology typing reagents/methods on the US market

zi
DAT by GT was not, therefore, performed. Genotyping are approved for independent clinical use, and hence
categorised four of the five cases as partial D (P32, P33, confirmation using a second reagent is not required.
P38, P39) and one as D-positive (P50). Of the four partial
i Most hospitals do not, therefore, mandate RhD typing
rv
D subjects who demonstrated alloanti-D antibodies, with two testing methods. In most studies published to
three (P32, P38, P39) were multiparous females (P38 date32,33, RHD genotyping was performed because of
Se
also had a recent history of transfusion with Rh-positive either weak reactivity or Rh type discrepancy noticed
red cells at the present institution) and one was a male incidentally in practice such as historical type vs current
(P33) with no known transfusion history. P38 tested type, donor type vs recipient type, or between facilities.
strongly D positive (4+) with TT (GT not performed), In our hospital, since June 2012, ABO front typing and
TI

whereas P39 tested moderately D positive (3+) with GT Rh typing have been performed twice by two different
(TT not performed). Three cases (P33, P38, P39) tested methods (GT and TT) on the first specimen from all
M

negative in the auto control, which is typical for partial patients without historical blood type information.
D subjects serologically demonstrating alloanti-D. P32 In addition, a second specimen has been required
tested weakly positive for the auto control (1+) with GT. for ABO/Rh group confirmation for non-emergency
SI

Although an autoanti-D component could contribute transfusion. In the present study, we report on the use
to the weak anti-D reactivity in P32, the possible role of these two methods for Rh typing and the discrepancy
of alloanti-D cannot be completely ruled out, given in reactivity between them, as a complementary
©

the partial D phenotype in this patient. The D variants criterion (C1) in addition to weak reactivity (C2) and
in these subjects, DAR (P32), DFV (P33), and DIVa the presence of anti-D in D-positive subjects (C3), to
(P38 and P39), have been associated with alloanti-D screen individuals for RHD genotyping.
production8. The fifth subject in this group (P50) is a As early as 1995, a strategy of using two monoclonal
93-year old male with no known transfusion history. The antibodies to identify DVI variants was proposed
patient's red blood cells demonstrated strong serological by Wagner and Colleagues 33. Subsequently, other
reactivity (4+) with both GT and TT, and tested weakly researchers also noted that a two-monoclonal anti-D
positive for auto control (1+) with GT. The DAT by TT strategy could be useful in identifying D variants34.
was negative and, therefore, elution was not performed There are some similarities and differences between a
according to laboratory procedures. This patient's two-monoclonal antibody strategy and a two-method
genotype was a normal D allele and a weak D type 3 strategy. The basic principle of both strategies is the
allele such that his phenotype is D-positive. same, i.e., using discordance between two different
test results to identify samples that may express a D
Discussion variant. A two-monoclonal anti-D strategy is efficient at
Both the AABB23 and CAP30 require two specimens identifying specific D variants of interest, for example
be tested for non-emergency pre-transfusion testing, DVI, which is common in individuals of European

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

298 All rights reserved - For personal use only
No other use without premission
Candidates for RHD genotyping

descent and clinically important due to the risk of facilities. Specimens from 23 partial D patients (P1-P21,
alloimmunisation. Our approach, which is not limited P33, P39) reacted strongly (3+ or 4+) with GT. Had it
to the use of monoclonal anti-D reagents, was designed not been for the reactivity discrepancy identified by the
to detect more variants, which would be expected in a two-method strategy, these partial D patients would have
multi-ethnic population. In addition, both anti-D reagents been managed as D-positive patients putting them at risk
used in the two testing methods (TT and GT) in this study of developing allo-anti-D antibodies after exposure to
are approved by the US Food and Drug Administration D-positive blood.
for independent use and both can detect normal D and Similarly, a total of ten weak D cases (weak D type
most variant types. Our approach uses both TT and GT. 1, 2 or 3) were identified by targeted genotyping in the
TT is performed with a monoclonal-polyclonal blend present study. Of these ten cases, six (P42-P45, P48, P49)
anti-D and GT is performed using a monoclonal human were D-negative by TT. These cases would have been
IgM anti-D secreted by a mouse/human hybridoma. improperly categorised had the GT not been performed.
The differences in both test mechanisms and antibodies However, if all specimens with negative TT were sent for
used are contributors to the sensitivity in detecting RHD genotyping, the yield (i.e. specificity) of detecting
discrepancies between the two methods. D variants would be low, since negative TT by itself is
not able to differentiate true D-negative from D variant
The criteria are highly predictive for RHD variants cases. All these six weak D patients (P42-P45, P48, P49)

l
Among all the 50 patients who met at least one of the were either weak D type 1 or type 2 patients (Table I),

Sr
criteria, D variants were detected by targeted genomic and could be managed safely as D-positive recipients.
testing in 48 cases (10 [20%] with weak D and 38 [78%] RhIG prophylaxis and Rh-negative red blood cell
with partial D) and by cDNA sequencing in one case products are not required (since the chance of producing

zi
(P40, weak D) with a positive predictive value of 98% alloanti-D is low12,13), which helps to save D-negative
(49/50) (Table I). blood products and RhIG resources.
When RHD genotyping was limited to targeted
i These results demonstrate that the two-method
rv
testing, which included use of a commercial array (RHD reactivity discrepancy criterion (C1) can help to
BeadChip) and several laboratory-developed tests for identify candidates for RHD genotyping. This is not
Se
variants not interrogated on the array, D variants were only because individuals with weak reactivity are
identified in 48 of 50 patients. When the patient with no more likely to be identified by two methods rather than
variants identified by targeted testing (P40) was tested one, but more importantly because the discrepancy in
using the higher resolution cDNA analysis, a D variant reaction strength can be appreciated only by using two
TI

(weak D type 61) was detected. This case illustrates methods. Since both the AABB and CAP require two
that when targeted genomic testing fails to detect a blood typing determinations in pre-transfusion testing,
variant, the negative test result does not fully exclude and performing two blood typing tests by two different
M

the possibility of a D variant. The commercial RHD methods can be done with minimal additional labour
genotyping products test for many variants commonly or expense, and useful information is obtained from
SI

found in individuals of Caucasian and African ancestry. possible discrepancies in reactivity between the two
Given the Hispanic ethnicity of this female patient methods, this strategy should be considered by hospital
and her GT 1+, TT negative serology reactivity, it was transfusion services for testing of all new patients.
©

strongly suspected that she carried an RhD variant that Reactivity discrepancy was more likely to be seen
was not detectable by the targeted genomic testing used. with partial D than weak D specimens (Results and
cDNA analysis is useful to detect other variants in such Table III). This is not surprising, as partial D variants
scenarios. have altered or missing RhD epitope(s) such that use
of different reagents and different methods may give
The two-method strategy is more sensitive disparate results. If a particular epitope detectable
The criteria described here are more sensitive at by the anti-D reagent used is lacking in a partial D
identifying subjects likely to express weak or partial D specimen, the reactivity observed with this anti-D
variants than weak reactivity of <2+ (C2) using only reagent is expected to be weak or negative. However,
one method such as GT or TT. As mentioned above, the the epitope detectable by another anti-D reagent
cases with strong reactivity (≥3+) by GT, and those cases may be present in this partial D specimen, and the
negative by TT would not have been identified without reactivity would be strong if the other anti-D reagent
using a second method (Table I). is used. Thus, the discrepancy between GT and TT
Some partial D individuals in the present study would in reactivity is a useful screening method to identify
have been typed as D-positive if only one method had specimens requiring RHD genotyping, especially for
been used, which is the current common practice in many partial D variants.

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

All rights reserved - For personal use only 299
No other use without premission
 Luo X et al

Criterion 3 (presence of anti-D in a presumed renal disease and had a past medical history of diabetes
D-positive patient) is complementary to criterion 1 mellitus, hypertension, and childhood asthma. He did
(discrepancy) and criterion 2 (weak reaction) not speak English, and an interpreter was present for
Five cases (P32, P33, P38, P39, P50) met C3, all of the entire encounter. He was admitted to an outside
whom were serologically D-positive by history and/or hospital 3 months previously and did not recall any
typing in the present institution and/or other facility, and history of blood transfusion or RhIG administration.
demonstrated anti-D antibodies with no concurrent non- He was strongly D-positive (4+) with GT and weakly
specific warm autoantibodies. Genotyping identified D-positive (1+) with TT. No agglutination was
four of these five as partial D (P32, P33, P38, P39) and observed with the auto control, suggesting that the
one as D-positive (P50). Three cases (P38, P39, P50) anti-D antibodies were the result of allo-immunisation,
met C3 only, and two cases (P32, P33, both partial D) most probably due to a previous, remote exposure to a
met C3 and at least one other criterion (Table IV). D-positive blood product.
Case P38 (partial D, RHD*DIVa) was strongly
D-positive (4+) by TT (GT not performed). This More D variant cases, especially partial D, identified
case would not have been considered a candidate in the current study
for RHD genotyping without the identification of In the present study, the D variants identified
anti-D. P39 (partial D, RHD*DIVa) tested moderately using our screening criteria consisted of more partial

l
D-positive (3+) with GT (TT not performed). Case D cases (39/50; 78%) than weak D cases (10/50;

Sr
P50 demonstrated strong serological reactivity (4+) 20%). These results are different from published
with both GT and TT. This case would not have been data26,27. In a study by Van Sandt et al.27 weak D type
considered as a candidate for RHD genotyping, had 1, 2, or 3 was identified in 424/628 (67.5%) cases,

zi
anti-D not been identified. RHD genotyping showed type 4.0/4.1/4.3, 4.2, 5, 11, 15, or 17 was identified
that this patient carries an apparently conventional D in 22/628 (3.5%) cases, and partial D variants were

i
allele suggesting that the anti-D in this patient is likely identified in 49/628 (7.8%) cases (mainly DVI types,
rv
to be an autoantibody. n=27). In a study published by Haspel et al.26, weak
RHD genotyping can identify variants in patients in D type 1, 2 or 3 was identified in 27/36 cases (75%),
Se
whom anti-D is identified in serologically D-positive whereas partial D variants were identified in only 9/36
individuals. Both P32 (female, partial D) and P50 cases (25%). It is known that types of the D variants
(male, D positive) tested weakly positive for auto and their frequencies in the population vary with race/
control (1+) with GT. Patient P32 typed D-negative ethnicity. A difference in race/ethnicity between this
TI

and received RhIG during her pregnancy at an outside and prior studies may contribute to, but is not likely to
hospital about 10 years before the current specimen be the only reason behind, the significant difference in
was taken. However, 5 weeks prior to the date of the constitution of partial D between our study (78%) and
M

current specimen, she typed D-positive at another other studies (7.8-25%). Using two typing methods
outside hospital and received two units of D-positive and the discrepancy criterion (C1) for screening is
SI

red blood cells. Our workup showed Rh type 2+ with probably a prominent contributor to the higher partial
GT, negative with TT, positive for anti-D antibody with D variant constitution observed in the present study.
positive auto control (1+). The DAT was negative and We hypothesise that the difference in sensitivity to D
©

eluate tests were not performed. Auto-adsorption was variants between the two typing methods is the key
not performed because of the recent transfusion. The contributor to the difference in the constitutions of D
patient had not received RhIG in the 3 months prior to variants between our study and other published studies.
the current specimen being taken. Genotyping detected
RHD*DAR (hemizygous or homozygous), which has Conclusions
been reported to produce alloanti-D antibody following The three-criteria approach proposed in the
exposure to D-positive blood. This patient's anti-D is, present study is a useful screening tool for identifying
therefore, probably an alloantibody candidates for RHD genotyping, with a high positive
Patient P50 was strongly positive (4+) with both GT predictive value of 98%. This approach is more
and TT. He had no recent history of transfusion or RhIG sensitive than using a single method and is more
use within the 3 months prior to the current specimen sensitive for partial D. Incorporating the two-method
being taken. Auto-adsorption was not performed. RHD discrepancy strategy into the current AABB and CAP
genotyping confirmed the presence of a normal D allele requirements for two blood type determinations/
and D-positive phenotype. specimens in pre-transfusion testing will help to
P33, a 62-year old Hispanic man, was seen initially identify more candidates for RHD genotyping with a
in mid-2015 for pre-transplant evaluation for end-stage minimal increase in work load and cost.

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

300 All rights reserved - For personal use only
No other use without premission
Candidates for RHD genotyping

Acknowledgements 17) Cannon M, Pierce R, Taber EB, Schucker J. Fatal hydrops fetalis
The Authors acknowledge the work of the staff of the caused by anti-D in a mother with partial D. Obstet Gynecol 2003;
102: 1143-5.
American Red Cross National Molecular Laboratory, 18) Prasad MR, Krugh D, Rossi KQ, O'Shaughnessy RW. Anti-D in
who performed the genotyping, and Trina Horn and Rh positive pregnancies. Am J Obstet Gynecol 2006; 195: 1158-62.
Jessica Keller, who aided in interpretation of the results. 19) Yazer MH, Triulzi DJ. Detection of anti-D in D-negative recipients
The Authors would like to thank Jen H. Lin for his transfused with D-positive red blood cells. Transfusion 2007; 47:
2197-201.
critical reviews and Jeffrey J. Long for collecting data. 20) Shao CP, Maas JH, Su YQ, et al. Molecular background of Rh
D-positive, D-negative, D(el) and weak D phenotypes in Chinese.
Authorship contributions Vox Sang 2002; 83: 156-61.
XL, IJ, MG, SL and YL designed the study. IJ and 21) Kormoczi GF, Forstemann E, Gabriel C, et al. Novel weak D
types 31 and 32: adsorption-elution-supported D antigen analysis
KSM collected the data. XL performed the data analysis. and comparison to prevalent weak D types. Transfusion 2005;
XL, MAK, AC, SN and YL analysed and interpreted 45: 1574-80.
the data. XL and YL wrote the manuscript. MAK wrote 22) Sandler SG, Roseff SD, Domen RE, et al. Policies and procedures
the genotyping section. All Authors critically reviewed, related to testing for weak D phenotypes and administration
of Rh immune globulin: results and recommendations related
edited and approved the manuscript. to supplemental questions in the Comprehensive Transfusion
Medicine survey of the College of American Pathologists. Arch
The Authors declare no conflicts of interest. Pathol Lab Med 2014; 138: 620-5.

l
23) AABB. Standards for Blood Banks and Transfusion Services. 30th

Sr
ed. Bethesda, MD: AABB; 2016.
References 24) AABB. Joint Statement on Phasing-In RHD genotyping for
1) Levine P, Stetson RE. An unusual case of intragroup agglutination. pregnant women and other females of childbearing potential with
JAMA 1939; 113: 126. a serologic weak D phenotype. Available at: http://www.aabb.
2) International Society of Blood Transfusion. Table of blood group

zi
org/advocacy/statements/Pages/statement150722.asp. Accessed
antigens v4.0 141124; transfusion ISoB. Available at: http:// on 01/12/2016.
www.isbtweb.org/fileadmin/user_upload/files-2015/red%20 25) Langston MM, Procter JL, Cipolone KM, Stroncek DF. Evaluation
cells/links%20tables%20in%20introduction%20text/Table%20
i of the gel system for ABO grouping and D typing. Transfusion
blood%20group%20antigens%20within%20systems%20v4.0%20
rv
1999; 39: 300-5.
141124.pdf. Accessed on 28/12/2015. 26) Horn T, Crowley JA, Keller MA. RhD typing discrepancies
3) Okuda H, Suganuma H, Kamesaki T, et al. The analysis of resolved through RHD genotyping: a patient perspective.
Se
nucleotide substitutions, gaps, and recombination events between Transfusion 2014; 54: 29A.
RHD and RHCE genes through complete sequencing. Biochem 27) Wagner FF, Flegel WA. RHD gene deletion occurred in the Rhesus
Biophys Res Commun 2000; 274: 670-83. box. Blood 2000; 95: 3662-8.
4) RhesusBase; 2015. Available at: http://www.rhesusbase.info/. 28) Chiu RW, Murphy MF, Fidler C, et al. Determination of RhD
Accessed on 14/03/2017.
TI

zygosity: comparison of a double amplification refractory mutation

5) International Society of Blood Transfusion. Table of blood system approach and a multiplex real-time quantitative PCR
group antigens v.6_170205. Available at: http://www.isbtweb. approach. Clin Chem 2001; 47: 667-72.
org/fileadmin/user_upload/Working_parties/WP_on_Red_Cell_ 29) Singleton BK, Green CA, Avent ND, et al. The presence of an
M

Immunogenetics_and/Table_of_blood_group_antigens_within_ RHD pseudogene containing a 37 base pair duplication and a

systems_v6_170205.pdf. Accessed on 28/12/2016. nonsense mutation in Africans with the Rh D-negative blood group
6) AABB. AABB Technical Manual. 18th ed. Bethesda, MD: AABB; phenotype. Blood 2000; 95: 12-8.
SI

2014. 30) Westhoff CM, Vege S, Wylie D, et al. The JAL antigen (RH48)
7) Agre PC, Davies DM, Issitt PD, et al. A proposal to standardize is the result of a change in RHCE that encodes Arg114Trp.
terminology for weak D antigen. Transfusion 1992; 32: 86-7. Transfusion 2009; 49: 725-32.
8) Daniels G. Variants of RhD--current testing and clinical 31) College of American Pathologists. Transfusion Medicine Checklist;
©

consequences. Br J Haematol 2013; 161: 461-70. 2011. Available at: https://pdfs.semanticscholar.org/458e/a201925

9) Jenkins CM, Johnson ST, Bellissimo DB, Gottschall JL. Incidence af537de9b476a9e961056be95cc0c.pdf. Accessed on 14/03/2017.
of weak D in blood donors typed as D positive by the Olympus PK 32) Haspel RL, Westhoff CM. How do I manage Rh typing in obstetric
7200. Immunohematology 2005; 21: 152-4. patients? Transfusion 2015; 55: 470-4.
10) Sandler SG, Flegel WA, Westhoff CM, et al. It's time to phase in 33) Van Sandt VS, Gassner C, Emonds MP, et al. RHD variants in
RHD genotyping for patients with a serologic weak D phenotype. Flanders, Belgium. Transfusion 2015; 55: 1411-7.
Transfusion 2015; 55: 680-9. 34) Wagner FF, Kasulke D, Kerowgan M, Flegel WA. Frequencies
11) Wagner FF, Gassner C, Muller TH, et al. Molecular basis of weak of the blood groups ABO, Rhesus, D category VI, Kell, and of
D phenotypes. Blood 1999; 93: 385-93. clinically relevant high-frequency antigens in south-western
12) Flegel WA. How I manage donors and patients with a weak D Germany. Infusionsther Transfusionsmed 1995; 22: 285-90.
phenotype. Curr Opin Hematol 2006; 13: 476-83. 35) Von Zabern I, Wagner FF, Moulds JM, et al. D category IV: a
13) Pham BN, Roussel M, Peyrard T, et al. Anti-D investigations in group of clinically relevant and phylogenetically diverse partial
individuals expressing weak D type 1 or weak D type 2: allo- or D. Transfusion 2013; 53: 2960-73.
autoantibodies? Transfusion 2011; 51: 2679-85.
14) Le Marechal C, Guerry C, Benech C, et al. Identification of 12 novel
Arrived: 28 September 2016 - Revision accepted: 21 December 2016
RHD alleles in western France by denaturing high-performance Correspondence: Yanhua Li
liquid chromatography analysis. Transfusion 2007; 47: 858-63. Department of Pathology and Laboratory Medicine
15) McGann H, Wenk RE. Alloimmunization to the D antigen by a Lewis Katz School of Medicine at Temple University
patient with weak D type 21. Immunohematology 2010; 26: 27-9. 3401 North Broad Street
16) Wagner FF, Frohmajer A, Ladewig B, et al. Weak D alleles express Philadelphia, PA 19140, USA
distinct phenotypes. Blood 2000; 95: 2699-708. e-mail: yhli02478@yahoo.com

Blood Transfus 2018; 16: 293-301 DOI 10.2450/2017.0274-16

All rights reserved - For personal use only 301
No other use without premission