Volume 45, May 2005

TRANSFUSION 689

Blackwell Science, LtdOxford, UKTRFTransfusion0041-11322005 American Association of Blood BanksMay 2005455689693Original Article

SPRCA SYSTEM DETECTION OF WEAK DLAI ET AL.

ABBREVIATIONS:

AGT

=

antiglobulin test; CAT

=

column

agglutination test; MAM(s)

=

microplate agglutination method(s);

SPRCA

=

solid-phase red cell adherence.

From the Transfusion Center, Clinical Hematology, Catholic

University of Sacred Heart, Rome, Italy.

Address reprint requests to:

Marco Lai, MD, Transfusion

Center Immunohematology Laboratory, Catholic University of

Sacred Heart, Largo A. Gemelli 8, 00168 Rome, Italy; e-mail:

marco_lai@fastwebnet.it.

Received for publication June 17, 2004; revision received

September 15, 2004, and accepted September 20, 2004.

doi: 10.1111/j.1537-2995.2005.04250.x

TRANSFUSION

2005;45:689-693.

I M M U N O H E M A T O L O G Y

Detection of weak D with a fully automated solid-phase red cell adherence system

Marco Lai, Gerarda Mavilia, Giuseppe d’Onofrio, and Giuseppe Leone

BACKGROUND:

Microplate agglutination methods (MAMs) and column agglutination technology are widely employed for red cell typing and can be automated. Some tests, however, such as detection of weak D, require manual testing. The possibility of detecting weak D by a solid-phase RBC adherence (SPRCA) test was studied in a fully automated system.

STUDY DESIGN AND METHODS:

The results of 2609 blood samples, characterized as being D– or with a incomplete agglutination reaction, were analyzed for the presence of the weak D phenotype. The 2609 samples were tested by a weak D tube test (antiglobulin method) and a weak D-test with the new SPRCA method. When weak D was detected, which D epitope was involved and whether it was associated with a partial D phenotype were determined.

RESULTS:

Weak D was detected in 60 (2.3%) of the 2609 samples. The 60 samples that were weak D by the tube test were also weak D

+

with the new automated SPRCA test. The sensitivity and specificity for the new weak D typing method were 100 percent, when compared with the standard weak D manual test.

CONCLUSIONS:

These results support the possibility of performing weak D detection with a SPRCA fully automated system. These preliminary results are encouraging, showing good sensitivity and specificity of the test. Detection of weak D will permit full automation of blood typing.

olid-phase techniques for blood group typing wereinitiated in the 1970s and then improved andadapted for different semi- or fully automatedmachines for different immunohematology tests.

At this time, the majority of routine blood typing tests canbe performed in fully automated systems. Some tests,however, such as weak D detection by the standard anti-globulin method (AGT), are still largely performed by themanual tube method, which is the reference method atthe moment.

1

Detection of weak D by a manual methodentails selecting all the samples that are D– or those withweak agglutination reactions and then performing themanual test and recording the results.

Previous reports detailed the possibility of perform-ing weak D typing with a microplate method, includingdirect agglutination in the presence of potentiators, anti-body dilutions, a solid-phase adherence method, or strat-egies that employed a red cell (RBC) monolayer in thebottom of the well.

2-6

These methods, however, did notbecome popular in routine RBC typing.

We analyzed the results of a new method for weak Ddetection with a fully automated solid-phase red celladherence (SPRCA) technology. The aim was to develop amethod for weak D detection that could be integrated intoa fully automated system for blood grouping, and also tohave an acceptable sensitivity and specificity for detectionof weak D, comparable with the current reference test.

S

LAI ET AL.

690 TRANSFUSION

Volume 45, May 2005

MATERIALS AND METHODS

Samples

Ethylenediaminetetraacetate-anticoagulated blood sam-ples were collected from 2609 patients who had no previ-ous blood transfusions in the past year. Samples fromhematologic patients or patients with a positive directAGT were excluded from the study.

D typing by slide or tube test

The D slide test was performed with a polyclonal anti-D(anti-D RH1, Ortho-Clinical Diagnostics, Raritan, NJ), fol-lowing the manufacturer’s instructions. The slide test wasperformed in all cases in which the weak D was detectedwith the standard tube test method.

The D tube test was performed with a monoclonalanti-D (anti-D rapid clone

RUM-1,

Immucor, Norcross,GA) following the manufacturer’s instructions. The D tubetest was performed in all cases when weak D was detectedwith the standard tube test method.

D typing with a column agglutination test

D typing was carried out in a fully automated system(AutoVue System, Ortho-Clinical Diagnostics), employingABO D biocards (clone

D7B8

), following the manufac-turer’s instructions. A visual check was performed beforethe validation of the results.

D typing with microplates

D typing by a microplate agglutination method (MAM),based on a direct agglutination reaction, was carried outin duplicate, with a fully automated system (Galileo,Immucor), employing a monoclonal anti-D (anti-D rapidclone

RUM-1,

Immucor) and a blended anti-D (anti-Dduo, Immucor; IgM clone

TH28

; IgG clone

MS26

), follow-ing the manufacturer’s instructions. The agglutinationreactions were defined incomplete when there wereagglutinates in the center of the well and a large amountof free RBCs in the peripheral zone as shown in Fig. 1. TheGalileo system gave an alarm flag for reactions in a rangeof optical densities of 24 to 56 (in a scale from 1 to 100).

D epitope determination

D epitope detection used two different commercial kits, akit for tube method that determines nine epitopes withthe following anti-D

7-8

, clones

HM 10 (EpD 6.6), HM 16(EpD 6.4), P3

¥

61 (EpD 6.4), P3x 35 (EpD 5.4), P3x 212 11F1 (EpD 8.2), P3x 212 23 B10 (EpD 9.1), P3x 241 (EpD 5.4),P3x 249 (EpD2.1), and P3x 290 (EpD 3.1)

(Diagast, France),and a kit that used the column agglutination test (CAT)method (Diamed, Switzerland), which determines sixepitopes with the following anti-D

7-8

clones and epitopes

detected:

LHM76/55 (EpD 3.1), LHM77/64 (EpD 9.1),LHM70/45 (EpD 1.2), LHM59/19 (EpD 8.1), LHM169/80(EpD 6.3), and LDM-1 (EpD6.5).

The classification ofweak D was performed following the instructions sup-plied from the manufacturers.

Weak D by the standard tube test

This test employed a polyclonal anti-D (Ortho-ClinicalDiagnostics) and anti-human globulin (Immucor), follow-ing the manufacturers’ instructions.

New weak D test with SPRCA

Samples from the 2609 patients were analyzed for theweak D phenotype, with a new fully automated system(Galileo Immucor) employing SPRCA. This method can beconsidered a modification of the SPRCA method for un-expected antibody detection. This weak D-test was per-formed with a blended anti-D (anti-D duo, Immucor; IgMclone

TH28

; IgG clone

MS26

).

New method phases

In the first step, 100

m

L of a 10 percent suspension of RBCs,to be typed for weak D, was added to microplate wells thathad been coated with an IgG anti-glycophorin. In the sec-ond step, the microplate was centrifuged, agitated, andthen washed. In the third and fourth steps, a blend of anti-

Fig. 1.

Different agglutination reactions in microplate D typing

(1) and the new D weak test (2). (A) Complete agglutination with

both the monoclonal (mon) and blended (blend) anti-D in a D

++++

subject; (B and C) incomplete agglutination of weak D RBCs

with the mon anti-D and negative with the blend; (D) absence

of agglutination with both mon and blend in a D– subject; (E

and F) positive and negative reactions, respectively, with RBCs

of a weak D phenotype and with a D– with the new weak D test.

SPRCA SYSTEM DETECTION OF WEAK D

Volume 45, May 2005

TRANSFUSION 691

D was added to the wells and then incubated for10 minutes at 43

∞

C followed by 10 minutes at 37

∞

C. Theanti-D binds to any weak D

+

RBCs in the bottom of thewell. In the fifth step, the indicator RBCs (Capture-R indi-cator cells, Immucor) were added to each well and thencentrifuged. The indicator cells added in this phase wereRBCs coated with anti-IgG; the anti-D was attached to theRBC membrane via the Fc portion of the immunoglobu-lin. In the case of weak D phenotype, the indicator cellsbind to the RBC layer at the well bottom.

Statistical analysis

All the statistical calculations were performed with com-puter software (STATA 8 SE, StataCorp, College Station,TX).

RESULTS

The evaluation of a novel microplate test for the weak Ddetection was performed on 2609 blood samples that werefound to be D– or showed an incomplete agglutinationpattern. Of the 2609 blood samples, tested for D with themicroplate blood typing method, 2571 were D– and 38gave an incomplete agglutination reaction with a mono-clonal anti-D. With the blended anti-D, only 7 D typingtests showed an incomplete agglutination pattern. Theremaining 2602 samples were negative. The D tube testshowed weak agglutination with a strength between 1/2

+

and 1

+

, in 41 of the 60 weak D samples. The slide test waspositive in 1 of the 60 weak D with a delayed reaction (finesand appearance).

New weak D test with SPRCA and D typing with CAT

The new weak D test, with SPRCA, was positive in 60(2.3%) of the 2609 samples tested. These cases includedthe 38 samples showing an incomplete agglutination pat-tern by the MAM and 22 samples that were D– with theMAM (see Table 1).

Rh typing was performed in conjunction with the CATin a fully automated system with the following results: ofthe 2609 samples tested, 2553 samples were D–, 5 sampleswere D

+

(complete agglutination reac-tion strength of 4

+

), 51 samples showedan incomplete agglutination reactionbetween 1/2

+

and 3

+

(see Table 1). Ofthe 2553 samples that were D– with CAT,4 samples proved to be partial D.

Weak D detection with the traditional AGT and D epitope typing

All the D– samples and those showingan incomplete agglutination pattern

with MAM were tested by AGT. Of the 2609 samples tested,60 were weak D. The 60 samples that were weak D by thetraditional AGT were also classified as weak D with theSPRCA method.

All 60 weak D tested by two different commercialkits with gel matrix technology (6 epitopes) and by liq-uid-phase tube test (9 epitopes). Eight of the 60 weak Dwere shown to be category V (3), category VI (3), andDFR (2). In the 8 samples with partial D phenotype, 4 (1DFR that reacted with LHM76/55, LHM77/64, LHM169/80, P3x 212 23 B10, P3x249, and P3x290 and 3 DV thatreacted with LHM76/55, LHM77/64, LHM169/80,LMH59/19, HM10, HM16, P3x 212 11 F1, P3x 212 23 B10,P3x249, and P3x290) gave an incomplete agglutinationpattern in D typing with the CAT methodology. Theremaining 4 (1 DFR reactive for the clones LHM76/55,LHM77/64, LHM169/80, P3x 212 23 B10, P3x249, andP3x290 and 3 DVI reactive for the clones LHM76/55,LHM77/64, and P3x 212 23 B10) were negative in D typ-ing with the CAT. With the MAM technology for D, typing3 partial D (2 DFR and 1 DV) samples gave incompleteagglutination patterns. The remaining 5 (2 DV and 3DVI) were D–.

Sensitivity and specificity of the new SPRCA test in detecting the weak D

The new SPRCA test for weak D demonstrated a sensitivityof 100 percent (confidence interval [CI], 94.04%-100%)and specificity of 100 percent (CI, 99.86%-100%). The pos-itive predictive value was 100 percent (CI, 94.04%-100%).The negative predictive value was 100 percent (CI,99.86%-100%) (Table 2). The median value of the opticaldensities in the microplate wells with the weak D test pos-itive was 93 percent (CI, 79%-94.6%); in the negative casesit was 15.3 percent (CI, 12.6%-28.5). The distribution of allthe optical densities in all the cases analyzed indicated aclear cutoff of the reactions between the weak D and neg-ative tests with the new test as shown in Fig. 2 with thekernel density plot. Figure 1 shows a weak D test positiveand one negative result as they appeared on visualobservation.

TABLE 1. Different results obtained in D typing with the CAT and the microplate method used in fully automated system for blood typing

*

Partial D phenotypes Weak D phenotypes CAT-

incompleteCAT-

positive (4

+

) TotalCAT-

negativeCAT-

incomplete TotalMicroplate-negative 17 0 17 3 2 5Microplate-incomplete 30 5 35 1 2 3Total 47 5 52 4 4 8

* The results are detailed in such manner to give information about the agglutination reaction obtained with the CAT technology and the corresponding result obtained with other microplate technology and vice versa.

LAI ET AL.

692 TRANSFUSION

Volume 45, May 2005

Capacity of the D typing to indicate the presence of the D weak phenotype with MAM or CAT methods

The results of D testing by MAM and by CAT were ana-lyzed on the basis of their capacity to predict the pres-ence of the weak D phenotype by having an incompleteagglutination pattern. The incomplete agglutinationreactions obtained in the microplate testing with themonoclonal anti-D gave a sensitivity (Table 1) of63.33 percent (CI, 49.90%-75.41%) and a specificity of100 percent (CI, 99.86%-100%).

The blended anti-D showed a sensitivity for detectingweak D of 11.67 percent (Table 2) (CI, 4.82%-22.57%) anda specificity of 100 percent (CI, 99.86%-100%). The com-bination of both the results obtained by the monoclonalanti-D and blended anti-D reagents gave a sensitivity anda specificity in predicting the D weak phenotype that wasthe same as that obtained with the monoclonal alone. Dtyping performed with CAT (Table 2) showed a sensitivity

for detecting weak D of 85 percent (CI,73.43%-92.9%) and the specificity was100 percent (CI, 99.86%-100%).

DISCUSSION

New technologies such as the solid-phase and CAT methods have allowedfor automation in blood typing. At themoment, however, some immunohe-matology tests are still performed withmanual tube methods, an example ofwhich is the weak D test. In a medium-or large-sized immunohematology lab-oratory, weak D tests are sufficientlynumerous and manual testing causes aninterruption in the automation line. Weanalyzed the results of a new weak Dtest that employs the SPRCA methodand a fully automated system, the Gali-leo (Immucor).

The results of this evaluation indi-cate that the new weak D test carriedout by the SPRCA method, proving opti-mal accuracy that will have a sensitivityand specificity of 100 percent whencompared with the standard manualmethod. This new weak D test shows apositive predictive value of 100 percentand a negative predictive value of100 percent. The hemagglutinationendpoints of this new test are clear-cutand easily evaluated rendering theinterpretation of the serologic reactionsuniform and producing a standardiza-tion of the results.

The execution of the new weak testin the fully automated system is planned in such a mannerso as to be performed on all the samples that are D– orshow an incomplete agglutination reaction. This permitsthe reduction of the hands-on time to only the evaluationand validation of the results, rendering blood typing to befully automated throughout all its phases without addi-tional manual steps. We also studied the results obtainedin D typing with two different automated methods (micro-plate and CAT).

Microplate D typing showed incomplete agglutina-tion patterns in 37 of the 60 weak D samples, indicatingthe presence of a weak D phenotype with a sensitivity of61.67 percent and a specificity of 100 percent. CAT showeda sensitivity for detecting weak D phenotype with incom-plete agglutination patterns that was 85 percent and aspecificity of 100 percent. In 4 of the 60 weak D samples,however, the CAT gave a complete agglutination reaction(D

+

).

TABLE 2. Sensitivity and specificity of the new test for the weak D detection

Weak D New Test MAM monoclonal MAM blended CATSensitivity (%) 100 63.33 11.67 85

95% CI 94.04-100 49.90-75.41 4.82-22.57 73.43-92.9Specificity (%) 100 100 100 100

95% CI 99.86-100 99.86-100 99.86-100 99.86-100

Fig. 2.

Distribution of the optical densities detected from the microplate reader for the

new weak D test. The optical densities are indicated on the abscissa and number of

events on the ordinate. The negative graph details the optical densities of the D–sam-

ples, the positive graph (rescaled) details the optical densities of the weak D pheno-

types. The total graph details both the negative and positive weak D samples (not

rescaled).

SPRCA SYSTEM DETECTION OF WEAK D

Volume 45, May 2005

TRANSFUSION 693

This new weak D test with the SPRCA technology issuitable for routine blood typing because it is integratedin a fully automated system with good sensitivity andspecificity, allowing for standardization of results. Thismakes automation in ABO and D typing complete andreduces the hands-on time to sample loading. The incom-plete agglutination reactions in the D typing obtainedwith the MAM and the CAT are useful in indicating thepossible presence of the weak D. It is important to empha-size that none of the partial D phenotypes were detectedwith a full agglutination strength in D typing with MAM orCAT. Furthermore, all the partial D phenotypes of categoryVI were negative in D typing with both MAM and CAT.Special care is required with regard to CAT D typing, whichin some cases might not reveal a weak D phenotype.

REFERENCES

1. Brecher ME, ed. Technical manual. 14th ed. Bethesda:

American Association of Blood Banks; 2002.

2. Llopis F, Carbonell-Uberos F, Montero MC, et al. A new

method for phenotyping red blood cells using microplates.

Vox Sang 1997;77:143-8.

3. Llopis F, Carbonell-Uberos F, Plannellesv MD, et al. A

monolayer coagglutination microplate technique for typing

red blood cells. Vox Sang 1997;72:26-30.

4. Glackin KC, Sherwood WC, Hoppe PA, McIntyre K. Weak D

antigen testing using the Groupamatic system. Vox Sang

1984;47:82-7.

5. Sinor LT, Rachel JM, Beck ML, et al. Solid-phase ABO

grouping and RhD typing. Transfusion 1985;25:21-3.

6. Epley KM, Severns ML, Kline LM. A one-step DU test in

microplate. Transfusion 1988;28:233-6.

7. Scott M. Rh serology-coordinator’s report. Transfus Clin Biol

1996;3:333-7.

8. Scott M. Section 1a. Rh serology coordinator’s report.

Transfus Clin Biol 2002;9:23-9.