Automated serological compatibility testing using a solid-phase test and standard laboratory equipment

225

Vox Sanguinis

(2001)

80

, 225–229

ORIGINAL PAPER

©

2001 Blackwell Science

Blackwell Science, Ltd

Automated serological compatibility testing using a solid-phase test and standard laboratory equipment

N. Ostendorf, D. Niefhoff, U. Cassens & W. Sibrowski

Department of Transfusion Medicine and Transplantation Immunology, Westfaelische Wilhelms-Universitaet, Münster, Germany

Background and Objectives

To prove the feasibility of a semi-automated cross-matchprocedure using a commercially available solid-phase microplate test and standardlaboratory equipment.

Material and Methods

The new procedure was evaluated against the conventionalspin tube technique and the gel centrifugation system.

Results

The sensitivity of the method and the rate of non-specific reactions were equalto those for the other test systems. The samples taken from the red cell concentratesfor cross-matching remained stable for the shelf-life of the product.

Conclusion

The semi-automated cross-match was successfully introduced in ourroutine laboratory as a means to process large numbers of tests.

Key words

:

laboratory automation, serologic cross-match, solid-phase test.

Received: 31 May 2000, revised 29 August 2000, accepted 10 January 2001

Introduction

Several solid-phase techniques are currently proposed asreplacements for the conventional agglutination methods,such as the spin tube technique, in serological testing ofblood cells against antibody-containing sera [1–5]. Solid-phasetechniques are considered to be more, or at least equally,sensitive than tube techniques and capable of processinglarge amounts of samples [6]. Their use in antibody screen-ing, but not in performing red cell cross-matches, is widelyestablished. We have introduced an automated cross-matchingprocedure into our routine laboratory to test patient seraagainst samples from red cell concentrates. The procedure isbased on a commercially available solid-phase antibody detec-tion test on microtitre plates (Solidscreen II, Biotest, Dreieich,Germany) and runs on commonly available laboratory equip-ment. The aim of this study was to evaluate the sensitivity ofthe procedure and the feasibility of the automatic process ina routine laboratory setting.

Material and methods

In 1999, about 72 100 cross-matches were performed, using20 600 patient sera. The routine method was a conventionalspin tube test in a three-stage assay, comprising a NaCl stageat room temperature, a low-ionic-strength stage at 37

°

C andfinally an indirect antiglobulin test. We performed severaltests with the new automated procedure in comparison tothe routine method and the ID gel centrifugation system(DiaMed, Cressier, Switzerland). After the introduction of theautomated test as a second routine method, we reviewed theserological compatibility tests performed during 1 year withregard to the number of cross-matches found positive byeither of the two methods.

Equipment

The following equipment was used to implement the auto-mated compatibility testing: pipetting robot Tecan RSP MegaFlex ID long with Kubus Sample Processor Software KSP (SLTLabinstruments, Crailsheim, Germany), Columbus microplatewasher (SLT Labinstruments), Spectra II microplate photo-meter (SLT Labinstruments) with Medusa 2000 software, micro-plate centrifuge Labofuge 400e (Heraeus, Hanau, Germany)and a microplate agitator Tinomax 100 (Heidolph, Kelheim,Germany).

Correspondence

: N. Ostendorf, MD, W. Sibrowski, PhD, MD, Department of Transfusion Medicine, Westfälische Wilhelms-Universität, D-48129 Münster, Germany Tel.: + 49 251 83–58506 Fax: + 49 251 83 58505 E-mail: [email protected]

VOX034.fm Page 225 Wednesday, May 23, 2001 2:57 PM

226

N. Ostendorf

et al

.

©

2001 Blackwell Science Ltd.

Vox Sanguinis

(2001)

80

, 225–229

Reagents

Solidscreen II microtitre plates were used, together with thereagents supplied by the manufacturer. This test is describedin detail elsewhere [7]. To perform the cross-match with auto-mated equipment, it was necessary to transfer a sample of thered cell concentrates into barcode-labelled pilot test tubes.The sample was taken from a 10-cm segment of the tubingcontaining the red cell concentrate. The erythrocytes weresuspended in 2 ml of Alsever’s solution and centrifuged at1000

g

for 10 min. This test tube was used for the completeshelf-life of the red cell concentrate. To ensure that the cor-rect segment from the donor unit was placed into a correctlybarcode-labelled test tube, the number of the unit wasscanned with a bar code reader and the label for the test tubewas printed automatically. On sampling, the prelabelled testtube was re-checked against the number of the red cell unit.

During the test procedure, 11

µ

l of the packed red cellsfrom the bottom of the tube was pipetted and diluted with500

µ

l of ‘modified Liss solution’ (Biotest) before transferonto the test plates. The dilutions were prepared and storedbefore analysis by the robot in a specially designed microtitreplate with large (higher) cavities.

A commercially available test panel (P3, Biotest) was usedfor the antibody detection test; 2 ml of the test cells wasdiluted with 4 ml of ‘Modified Liss Solution 2’ (Biotest).

Test procedure

Fifty microlitres of serum or plasma were pipetted into eachwell of the Solidscreen II plate. The prediluted red cells (seeabove) were added, the plates were removed from the pipet-ting robot and incubated for 20 min at 37

°

C. The microtitreplates were then centrifuged for 5 min at 1500

g

and, after-wards, were washed five times with an automated platewasher. Then, 100

µ

l of ‘Anti-Human-Globulin SolidscreenII’ (Biotest) was added, the plate was carefully agitated andcentrifuged for 3 min at 250

g

. The plates were then auto-matically processed by a microtitre plate reader at 492 nmand the results were analysed using the accompanyingsoftware. The instrument was set to measure the difference inlight transmission between the peripheral and the centralparts of the test well, using 20 different detection points. Adifference of less then 20% was considered positive, indicat-ing that erythrocyte coating at the bottom of the test well hadoccurred. The automatic interpretation was visually controlledby a technician. Finally, the test wells were controlled withCoombs-positive cells.

Data management

For data management, the proprietary laboratory softwareProTransmed (Elter computersysteme, Wuerzburg, Germany)

was used. This software system serves the complete function-ing and administration of the blood bank. All patient dataand the data of the stored blood products are managed usingthis software system. The selection of compatible red cellconcentrates to be cross-matched was first performed in thedata bank, with the possibility of selecting compatible products.The identification numbers of the patient samples and theblood products, together with the single tests to be performed,were exported on-line to the K

UBUS

software, which controlsthe pipetting robot. When the tests were completed, the resultswere re-imported into the laboratory information systemfrom the microtitre plate reader. A cross-matching report wasthen generated automatically.

Validation of equipment and tests

After the installation of the equipment, several tests were runto validate the sensitivity, specificity and reproducibility ofthe automated cross-match. Four trials are described indetail.

Sensitivity

Sera of 32 patients with formerly diagnosed specific red-cellantibodies were cross-matched against units of red cells thatcarried the corresponding antigen. With the same samples,an antibody detection test was performed. To obtain a suffi-cient sample volume, the patient samples had to be diluted1 : 2 first, using phosphate-buffered saline. The ID gel cen-trifugation system was used as a reference method and cross-matching and antibody detection were performed in parallel.When the cross-match and the antibody detection test gavenegative results in both systems, the sample was consideredto be no longer reactive. The positive samples were thendiluted further to investigate the sensitivity of both testsystems. Tests were performed according to the manufacturers’specifications.

Stability

Sixteen sera containing known antibodies were cross-matched with samples from incompatible red cell concen-trates directly after preparation of the pilot tubes and after40 days of storage of the pilot tubes at 4

°

C under routineconditions. The patient sera were stored at –40

°

C. To deter-mine the sensitivity at day 0 and day 40, the antibody reac-tions were titred on both days. This procedure, too, wascarried out in parallel with the ID gel centrifugation system.The erythrocytes to be tested in the DiaMed-System werestored in ‘ID-Cellstab’ (solution provided by the manufac-turer for long-term storage) at 4

°

C.

Routine introduction

A total of 955 cross-matches, in combination with 292 auto-logous controls and antibody detection tests using a three-cell


©


Vox Sanguinis

(2001)

80

, 225–229

Automated serological compatibility testing

227

panel, were performed in parallel by the conventional spintube test in a three-stage assay and by the automated solid-phase procedure as described above. The parallel cross-matchwith the automated procedure took place within 24 h of thespin tube test which was carried out as the routine procedure.Positive autologous controls were controlled by direct anti-globulin tests; positive results in the cross-match or antibodydetection test were controlled by re-testing in the ID gel cen-trifugation system. If these controls proved to be negative,the reaction was considered to be non-specific. After somemodification in the washing procedure to prevent accidentalwashing out of the wells, the same type of parallel testing wascarried out again with 403 cross-matches and 127 autologouscontrols, summing up for 1358 cross-matches and 419 auto-logous controls.

Routine performance

Finally, the results of both the manual and the automatedcross-match techniques during 1 year of routine handlingwere reviewed with regard to the overall number of positivetests. In routine testing, to provide red cell concentrates fortransfusion, two more tests were performed from each patientsample: determination of the ABO-blood group and RhD-typing (ABD-test) to reveal any possible misidentificationand an autologous control. For the ABO- and RhD-typing, theErytype ABD-32 kit (Biotest) was used. Moreover, an addi-tional antibody screening test was run parallel to each cross-match. This same policy was carried out with the conventionalspin tube test as well as with the automated cross-matching.

Results

Sensitivity

Of the 32 cross-matches between samples formerly known tocontain a specific antibody and incompatible red cell con-centrates, eight gave negative results in all procedures andwere therefore considered negative. In 13 cases, the antibodywas correctly detected in all test systems with equal sensitiv-ity. The dilution studies showed results differing by not morethan one dilution step. In one sample, both tests showedpositive reactions, but the DiaMed System was more sensitivethan the Solidscreen test (two titre steps). Both test systems failedto recognize two antibodies each that were correctly detectedwith the other system. These reactions were generally weakand near the minimum level of detection, but the Solidscreencross-match missed one Anti-D that gave a titre of 1 : 16with the DiaMed system. The Solidscreen antibody detec-tion test from this sample was positive. Apart from a possibleexplanation, this is a confirmation of our policy to performan accompanying antibody detection test against panelcells together with each cross-match. Six tests showedinconclusive results, with weak positive reactions in only one

system. These included an anti-M reactive at room temper-ature only and a weak anti-D, presumably after prophylacticanti-D injection. The inconclusive results with weak seramay be due to dose effects, since different cell panels wereused for the antibody detection tests in both systems(Table 1).

Stability

The 16 samples tested against pilot tubes stored for 40 daysshowed no significant decrease in sensitivity. On the con-trary, several samples showed an increase in titre, especiallywhen tested in the gel centrifugation system. This may becaused by deterioration of the erythrocyte membrane, givingrise to non-specific agglutination when passing through thegel phase. A drop in titre for more than one step occurring inthe solid-phase test only and not in the ID system was seenin only one case. All samples that were initially positive inboth systems remained positive after 40 days of storage ofthe pilot tube (Table 2).

Table 1 Results of 32 cross-matches and antibody detection tests with sera

from patients formerly diagnosed with a specific antibody (left column). The

eight sera which were found negative in all procedures are omitted

Antibody specifity

ID cross-match (titre)

Antibody detection test

Solidscreen cross-match (titre)

Antibody detection test

Anti-K, -Jkb 1 : 2 pos. 1 : 2 pos.

Anti-M 1 : 2 neg. 1 : 2 neg.

Anti-C, -e 1 : 2 pos. 1 : 4 pos.

Anti-C, -D 1 : 128 pos. 1 : 128 pos.

Anti-E neg. pos. neg. pos.

Anti-D, -E, -Fya 1 : 4 pos. 1 : 4 pos.

Anti-D neg. neg. neg. pos.

Anti-D, -Fya 1 : 2 pos. neg. pos.

Anti-D 1 : 128 pos. 1 : 256 pos.

Anti-Jka, -E 1 : 4 pos. 1 : 4 pos.

Anti-D 1 : 16 pos. neg. pos.

Anti-K, -c 1 : 2 neg. neg. neg.

Anti-K neg. neg. 1 : 4 pos.

Anti-D 1 : 512 pos. 1 : 512 pos.

Anti-D neg. pos. neg. neg.

Anti-K 1 : 16 pos. 1 : 32 pos.

Anti-D > 1 : 512 pos. > 1 : 512 pos.

Anti-Jka neg. pos. 1 : 2 neg.

Anti-E, -c, -Fya 1 : 16 pos. 1 : 4 pos.

Anti-D, -C 1 : 512 pos. 1 : 512 pos.

Anti-D 1 : 128 pos. 1 : 64 pos.

Anti-D, -C 1 : 64 pos. 1 : 64 pos.

Anti-E 1 : 128 pos. 1 : 64 pos.

Anti-Jka neg. neg. 1 : 2 pos.

neg., negative reaction; pos., positive reaction.


228

N. Ostendorf

et al

.

©


Vox Sanguinis

(2001)

80

, 225–229

Routine introduction

During the period of working up the samples with the con-ventional tube technique and the automated cross-match inparallel, the rate of non-specific positive results proved to beslightly higher for the spin tube technique (see Table 3). Therewere no specific positive results in the cross-matches witheither of the procedures. The additional antibody detectiontest was positive in five cases in the automated procedure thatfailed to react in the tube assay (four formerly known anti-bodies and one newly detected Anti-D). There was one anti-body newly detected by the tube technique, that did not reactin the solid-phase test (Anti-P

1

).In a sample run, 23 patient sera were cross-matched with

62 red cell concentrates. The pipetting robot took 31 min toperform this task. Serum and erythrocytes were pipettedsequentially. The latter step, pipetting of the erythrocytesfrom the predilution cavities, took only about 2 min. Therefore,the total incubation time was not significantly prolonged.

Routine performance

After the introduction of the automated cross-match basedon the solid-phase technique as a second routine procedure,the number of cross-matches performed with this methodincreased from 20% initially to 40% in 1999, showing theacceptance of the method. Of these 29 300 automated cross-matches, 1·18% gave a positive result, whereas of the 42 891conventionally tested samples, 1·29% were positive. Whenthe automated procedure showed positive results either in thecross-match or in the antibody detection test against panelerythrocytes, the tests were repeated again in the ID gel sys-tem. If the test was reproducibly positive, the antibody dif-ferentiation was continued, if it was negative, the automatedprocedure was considered a false positive. There was a slighttendency of the solid-phase test to give false-positive resultsnot in the cross-match against samples from erythrocyteconcentrates but in the antibody screening against panelerythrocytes. The number of these false-positive antibodyscreening tests was approximately 5%.

We estimated the time required for an average sample runof 15 patient sera cross-matched against 60 red cell concen-trates. The preparation (searching for the pilot tubes, centri-fugation of the patient samples, preparation of the racks andreagents) took about 60–90 min. If pilot tubes had to be pre-pared first, the procedure was prolonged for about 1 h (1 minper concentrate). The time for pipetting the samples, incuba-tion, centrifugation and reading of the results on the micro-plate reader again took about 60–90 min. Manual control ofthe results and transfer to the laboratory information systemtook about 30 min. Finally, printing of labels and attachingthem to the cross-matched red cell concentrates took an addi-tional 30 min. The approximate training time for the labor-atory technicians to adapt the new method was 2 days.

Discussion

An automated cross-match procedure based on a commer-cially available solid-phase antibody detection test wasintroduced into a routine laboratory programme. We demon-strated that this procedure could be successfully implementedusing standard laboratory equipment and our proprietarylaboratory information system. The automatic procedure hadseveral advantages, compared to the conventional spin tube

Table 2 Antibody titres of sera cross-matched against samples of

incompatible red cell concentrates immediately after the preparation

of the sample and after 40 days of storage of the sample at 4 °C

ID system Solidscreen

Antibody day 0 day 40 day 0 day 40

P1 (room temp.) neg. neg. neg. neg.

D (enzyme only) neg. neg. neg. neg.

C + D 8 4 8 16

C + D 8 4 16 32

D 128 512 128 256

Fya 32 256 32 4

Fya 32 256 32 2

D 4 8 4 256

D 128 256 256 256

E neg. neg. neg. neg.

D 256 512 512 512

c neg. neg. neg. neg.

C + D neg. 2 neg. neg.

C + D 128 512 512 256

M neg. neg. neg. 2

Le(a) neg. neg. 4 neg.

neg., negative reaction.

Spin tube assay results Solid screen test

Total Total positive Specific positive Total positive Specific positive

Cross-matches 1358 16 (1·2%) 0 10 (0·7%) 0

Antibody detections 419 1 (0·2%) 1 (0·2%) 5 (1·2%) 5 (1·2%)

Autologous controls 419 6 (1·4%) 0 22 (5·3%) 12 (2·9%)

Table 3 Results of parallel testing of spin tube

assay vs. solid screen test in the automated

procedure


©


Vox Sanguinis

(2001)

80

, 225–229

Automated serological compatibility testing

229

test. The samples of patients and red cell concentrates arepositively identified by barcode and pipetting errors areavoided, large numbers of cross-matches can be performedin parallel and the interpretation of the results on the micro-titre plates is facilitated. Finally, the automated procedureneeds less sample volume. A larger number of red cell con-centrates can be cross-matched with the amount of serumfrom one patient sample. Errors in the automatic identifica-tion of the samples were not observed. Disadvantages includethe fact that the results still have to be controlled by a tech-nician, because the automatic plate reader may misinterpretweakly positive wells as negative. Coagulation or air bubblesin the sample tubes can induce errors during the automatedpipetting, therefore the samples have to be carefully prepared.

The sensitivity of the automated cross-match was equal tothat of the gel centrifugation technique, comparing the titreof antibody containing sera in both systems. The samples ofthe red cell concentrates taken for the cross-match with theautomated procedure could be used for the whole shelf-lifeof the products. The re-testing after 40 days showed repro-ducible results. The rate of unspecific positive results in thecross-match was 0·7% for the solid-phase technique, com-pared with 1·2% in the spin tube test, as evaluated by paralleltesting of 1358 single cross-matches from 419 patient sera.

The washing procedure can be regarded as a crucial factor.To obtain satisfactory results, it is important that the washingsolution is not injected freely into the cavity but that it shouldrun down the wall. This avoids loss of too many test eryth-rocytes. If the cells are washed out to a certain extent, thewells are interpreted as positive by the microplate reader,because no detectable pellet of erythrocytes is formed in thecentre of the microplate well. The sensitivity of the auto-mated cross-match in routine testing was equal to that of theconventional spin tube method, comparing the overall posit-ive results recorded by the laboratory information system in1999 (1·18% vs. 1·29%). This is in concordance with a studyby Hitzler

et al

. [8], who compared the spin tube method, thegel centrifugation system and three different solid-phase

tests, while other authors found solid-phase assays moresensitive than the conventional spin tube test [6,9]. Duringthe observation period, we recorded no delayed haemolytictransfusion reaction caused by irregular erythrocyte antibodiesand no acute haemolytic incidents on the basis of ABO-incompatibility. Therefore, the safety of the test seemed to besatisfactory. The automated procedure proved to be capableof handling large amounts of samples, but for fast results incases of emergency, the tube technique remained useful.

References

1 Douglas R, Schneider JV, Wilkie D, Harden PA: A solid phaseantiglobulin test.

Transfusion

1987;

27

:378–3832 Plapp FV, Sinor LT, Rachel JM, Beck ML, Coenen WM, Bayer WL:

A solid phase antibody screen.

Am J Clin Pathol

1984;

82

:719–721

3 Rachel JM, Sinor LT, Beck ML, Plapp FV: A solid-phase anti-globulin test.

Transfusion

1985;

25

:24–264 Scott ML: The principles and applications of solid phase blood

group serology.

Transfus Med Rev

1991;

5

:60–725 Sinor LT, Rachel JM, Beck ML, Bayer WL, Coenen WM, Plapp FV:

Solid phase AB0 grouping and Rh typing.

Transfusion

1985;

25

:21–236 Sallander S, Shanwell A, Aqvist M: Evaluation of a solid-phase

test for erythrocyte antibody screening of pregnant women, patientsand blood donors.

Vox Sang

1996;

81

:221–2257 Uthemann H, Poschmann A: Solid-phase antiglobulin test for

screening and identification of red cell antibodies.

Transfusion

1990;

30

:114–1168 Hitzler W, Johanson C, Schömig-Brekner J, Mathias D: Vergleichss-

tudie zur Antikörperindentifizierung im Gelzentrifugationstest(ID-Microtyping-System). Festphasen- Antiglobulintest (Solid-screen, Capture R, Ready ID) und Röhrchentest.

Ärztl Lab

1991;

37

:137–1469 Guignier F, Domy M, Angue M, Richaud P, Chatelain P: Comparison

between a solid-phase low-ionic-strength solution antiglobulintest and conventional low-ionic-strength antiglobulin test: assess-ment for the screening of antierythrocyte antibodies.

Vox Sang

1988;

55

:30–34


Documents

Automated serological compatibility testing using a solid-phase test and standard laboratory equipment