Column agglutination technology: the antiglobulin test

K.J. REIS,R. CHACHOWSKI,

A. CUPIDO,
D. DAVIES,
J. JAKWAY,
AND T.M. SETCAVAGE

A new system for typing and screening blood, based on the sieving effect of lass
bead microparticles, has been developed. The test is performed in a microco umn
in which the red cell ag lutinates are trapped in the glass bead matrix during
9
B
centrifugation, and unagg utinated cells form a pellet at the bottom of the column.
Anti-human globulin reagents were incorporated in the diluent and the new test
system, column agglutination technology, was compared to conventional tube tests
and low-ionic-strengthmethod. Sera and plasmas (228 samples) were screened for
red cell antibodies with two anti-human globulin reagents: one containing only anti-
I G and the other containing both anti-lgG and anti-C3b, -C3d. After initial testing,
f!
t ere was 94-percent agreement between column agglutination technology and tube
tests, and after repeat testing, there was 97-percent agreement. The column
agglutination technology anti-human globulin test eliminates the need to wash red
cells, which decreases the overall test time. The test is easy to perform, and the
results are more objective than those with tube and microplate methods.
TRANSFUSION 1993;33:639-643.

Abbreviations: AHG =anti-human globulin; CAT= column agglutination technology;

HTLA = hlgh-titer, low-ovldlty antibody; NSP = normal serum pool.

THEAGGLUTINATION OF RED CELLS has been used success- the bead column, where agglutinated cells are trapped,
fully by immunohematologists for 90 years, starting with while unagglutinated red cells travel to the bottom of the
the first descriptionsof the ABO system by Landsteiner'-* column, forming a discrete pellet. Antisera can be
and the development by Coombs et aL3in 1945 of a meth- incorporated in the diluent to type red cells or to detect in
od of detecting red cell antibodies by indirect agglutina- vivo and in vitro IgG- or complement-sensitizedred cells.
tion using IgG antibodies. Traditional methods to detect Two CAT AHG tests have recently been developed for
agglutinationreactions are not without problems. Agglu- the detection of IgG- or complement-sensitizedred cells.
tination reactions are generally detected by disrupting a The IgG AHG test contains rabbit anti-human IgG, and
red cell pellet after the centrifugation of a tube or the polyspecific AHG test contains rabbit anti-human
microplate, and they require a skilled technologist for IgG and mouse monoclonal anti-C3b and anti-C3d. An
reliable interpretation. The anti-human globulin (AHG) important feature of the CAT AHG test is the viscosity
test requires that red cells be washed free of excess serum provided by the addition of polymers to the diluent, which
IgG before the addition of AHG reagent and centrifuga- affects the speed at which red cells and serum compo-
tion. A need exists for new tests that are easier to perform nents travel through the column during centrifugation.
and more objective to read and that lend themselves to The difference in the specific gravity of red cells and
automation. One such method is the recently described serum components allows the red cells to pass through the
gel test? which uses gel filtration media impregnated column more quickly than the less dense serum proteins
with blood grouping or AHG reagents to obtain and read do. The need to wash red cells in the traditional manner
agglutination. is eliminated, because red cells are exposed to the AHG
Column agglutination technology (CAT) is a new reagents well before the AHG reagents are exposed to
system for blood grouping and the detection of unex- soluble IgG and complement. In addition, polymers in
pected antibodies based on the sieving effect of glass the diluent potentiate agglutination. In the present study,
bead microparticles. Glass beads are placed in a the CAT AHG system was tested with 228 frozen and
microcolumn along with a diluent. Test red cells, with or fresh antibody-positive and -negative samples, and the
without serum, are placed in the chamber above the results were compared to those of conventionaltube tests.
column. Upon centrifugation,red cells are forced through Additional studies are presented, demonstrating that
~
neutralization of the anti-IgG component of the reagent
From Ortho Diagnostic Systems Inc, Raritan, NJ. does not occur when the test is performed either under
Received for publication October 19, 1992; revision received routine conditions or under conditions designed to stress
February 17, 1993, and accepted February 25, 1993. the system.

639
640 REIS ET AL. TRANSFUSION
Vol. 33. No. 8-1993

Materials and Methods column was prepared by using glass beads and an isotonic
diluent containing high-molecular-weight polymers. The
Samples diluent for the IgG AHG test contained rabbit anti-IgG in an
Frozen sera and plasmas ( 192 samples) were obtained from isotonic buffered solution containing polyethylene glycol.
Morris Dixon, Belle Bonfils Memorial Blood Center (Denver, Diluent for the polyspecific AHG test contained rabbit anti-
CO); Gerald Hoeltge, MD, Cleveland Clinic Foundation (Cleve- IgG and mouse monoclonal anti-C3b and anti-C3d in an
land, OH); Roger Collins, American Red Cross (Rockville, isotonic buffered solution containing dextran. We experimen-
MD); George Garratty, PhD, American Red Cross Blood tally determined antibody concentrations by cross-titrations
Services (Los Angeles, CA); and Ira Shulman, MD, Los with 1-in-2 dilutions of anti-IgG or anti-complement using red
Angeles County-University of Southern California Medical cells sensitized with diminishing amounts of IgG or comple-
Center (Los Angeles, CA). Samples had been stored frozen for ment, respectively. We conducted tube tests by this same
periods ranging from 15 days to 12 years. Most of the antibody- format for comparative analysis. CAT sensitivity determina-
positive samples used were reported to be weakly positive tions by this means were shown to be equal to or slightly more
(defined as c2+ on a scale of 1+ to 4+) by conventional than those of the tube test. Each cassette was sealed to prevent
methods. Random negative samples were included in this evaporation or contamination until time of testing. Reagents
population. Thirty-six fresh samples were obtained from (AHG Anti-IgG [Rabbit] and AHG [Rabbit and Murine Mono-
healthy donors and tested on the day obtained. Both fresh and clonal] BioClone Anti-IgG, C3d; antibody enhancement solu-
frozen samples consisted of sera as well as plasmas collected tion; and red cells (Selectogen Reagent Red Cells and Coombs
in a variety of anticoagulants. Control Cells) were all supplied by Ortho Diagnostic Systems
(Raritan, NJ). We prepared a normal serum pool (NSP) by
combining serum from four to six healthy volunteers found to
Materials be negative when tested for red cell antibodies. Human IgG
Cassettes containing six microcolumns, each column de- was purified from plasma by a modified Cohn fractionation
signed for a single test, were used to perform CAT tests. Each procedure; purity was monitored by immunoelectrophoresis.
For the neutralization experiments, we prepared normal serum
containing excess IgG by adding 15 mg of human IgG per mL
of NSP and filtering it through a 0.22-pn filter (Millex-GV,
Millipore, Bedford, MA). Immunoglobulin-free human serum
was obtained from Cappel (Organon Teknika Corp., West
Chester, PA).

CAT AHG test procedure

ment solution All samples previously frozen were clarified by centrifuga-
Red cells tion to remove fibrin or other particles that might block the
filtering capacity of the beads. We tested samples in CAT anti-
AHG Beads IgG and polyspecific AHG cassettes using a low-ionic-strength
diluent method by adding 40 pL of antibody enhancement solution, 10
pL of 3-percent reagent red cells, and 40 pL of test sample.
Cassettes were incubated at 37°C for 10 minutes and centri-
fuged for 2 minutes at 55 x g and for 3 minutes at 199 x g. A
...

fy
i;o; u Q al )

Positive Negative
FIG. 1. Schematic of the AHG procedure using CAT. The lower
portion of the column contains glass beads and the AHG diluent. Serum,
antibody enhancement solution, and red cells are added to the upper
portion of the chamber. After incubation and centrifugation, agglutin- FIG.2. Agglutination reaction patterns and scoring method in CAT.
ated red cells are trapped above or in the beads, and unagglutinated red A negative reaction (0)is recorded when all red cells pellet to the bottom
cells form a pellet at the bottom of the column. of the tube. Positive reactions are graded on a scale of I + to 4+.
TRANSFUSION
IW3-Vol. 33. No. 8 COLUMN AGGLUTINATION TECHNOLOGY 64 1

schematic of the procedure is shown in Fig. 1. Positive results Table 2. CATand tube antibody screen results
were recorded on a scale of I + to 4+ (Fig. 2). Anti-laG AHG Polvspecific
_ . AHG
(n I2281 (n = 180)
Tube AHG test procedure CAT Tube Initial Repeat Initial Repeat
We performed tube tests according to direction circulars for Positive Positive 81 86 61 65
AHG tube reagents (Ortho), using a low-ionic-strength method. Positive Negative 9 3 10 4
Briefly, 2 drops of serum or plasma and 1 drop of reagent red Negative Positive 3 3 0 1
Negative Negative 135 136 109 110
cells were added to tubes and centrifuged, and the results were
Total agreement 216 222 170 175
read. Then, we added 2 drops of antibody enhancement
Percentage of
solution and read the results after incubation at 37°C for 10 agreement 94.7 97.4 94.4 97.2
minutes. Red cells (RBCs) were washed three times with
saline, and the results were read at antiglobulin phase after the
addition of 2 drops of AHG reagent (anti-IgG or polyspecific).
Positive results were recorded on a scale of I + to 4+. We testing. This sample was not available for repeat testing, but a
confirmed all negative tube test results by using Coombs freshly drawn sample was found negative by both CAT and tube
control cells. methods.

Comparison of CAT and tube polyspecijic AHG tests

Results
A total of 180 samples (144 frozen and 36 fresh) were
Comparison of CAT and tube anti-IgG AHG tests screened as described above with the CAT polyspecific cas-
A total of 228 samples (192 frozen and 36 fresh) were sette and the tube test using AHG (Rabbit and Murine Mono-
screened in the CAT anti-IgG cassette and by conventional clonal) BioClone Anti-IgG,-C3d reagent (Table 1 ). After
tube test using AHG anti-IgG (rabbit) reagent (Table 1). After initial testing, there was 94.4 percent agreement between the
initial testing, 216 samples (94.7%) were in agreement by both test systems. The 10 samples that did not agree after initial
methods. The 12 samples that did not agree in the two tests testing were tested again. After repeat testing, there was 97.2
were tested again. After this, 222 (97.4%) of 228 samples were percent agreement (Table 2). A summary of the discrepancies
in agreement (Table 2). A summary of the initial 12 discrep- and the reaction strengths observed in all testing is shown in
ancies and the reaction strengths observed in all testing is
shown in Table 3. Six of the 12 initially discrepant samples
were positive in CAT on both initial and repeat testing. In Table 3. Summary of discrepant reactions
comparison tube tests, three of these samples were negative CAT Tube
after repeat testing (anti-P,, anti-K, anti-M) and three were Reportedantibody Initial Repeat Initial Repeat
positive after repeat testing (anti-c, anti-M, high-titer low- Anti-lgG AHG tests
avidity [HTLV] antibody). Three of the 12 initially discrepant Anti-P,' 1 1 0 0
samples were consistently positive in tube tests. One sample Anti-K' 1 1 0 0
(containing both anti-E and anti-Fya)was positive in CAT after Anti-M' 2 1 0 0
repeat testing; two samples (anti-P, and a cold agglutinin) were
consistently negative in CAT. Of the three remaining samples Anti-c' 3 3 0 1.5
with initially discrepant tests, one sample, reported to contain Anti-M 1 1 0 0.5
anti-Lea and a cold agglutinin, was detected once by CAT and HTW 1 1 0 1
once by tube test. A second sample, reported to contain
multiple antibodies (HTLA antibody, anti-K, anti-Leb), was Anti-E, -Fya 0 1 1 1
initially positive in CAT but negative in all other testing. One Anti-P, 0 0 1 1
fresh sample was positive in CAT and negative on initial tube Cold agglutinin 0 0 4 1

Anti-Lea+ cold
Table 1 . Frequencies of antibody specificities of samples used agglutinin 1 0 0 0.5
in CAT and tube test comparison HTLA, anti-K, -Leb 1 0 0 0
Unresolved 1 (0)t 0 (0)
Antibody group Anti-lgG AHG Polyspecific AHG
D, C, E, c 42 33 Polyspecific AHG tests
K 34 23 Anti-P,' 1 1 0 0
Fy", Fyb 12 9 Anti-K' 1 1 0 0
Jk", Jkb 6 3 Anti-M' 2 2 0 0
M, S 9 8 Anti-Lea 1 1 0 0
Pl 2 2
Lea,Leb 3 1 Anti-c* 3 3 0 2
Warm autoantibodies 12 8 Anti-Jka 2 1 0 1
Cold 9 6 Anti-K 1 1 0 1
Other' 20 14 Anti-K 1 1 0 1
Negative 136 110
Total 285t 21 7$ HTW 1 0 0 0.5
including HTLA, Bg, U, Tja,Coa,JP. .Anti-Bn
-=
..... 1 n -
n 0
t Of 228 samples tested, 34 contained multiple antibodies. * Samplesdiscrepant in both anti-lgG and polyspecific AHG tests.
*Of 180 samples tested, 22 contained multiple antibodies. t Different sample used for repeat test results.
 TRANSFUSION
REIS ET AL. Vol. 33. No. 8-1993

Table 3. Eight of the 10 initially discrepant samples were We observed no significant titration endpoint differences
consistently positive in CAT. Four of these samples (reported when either IgG-free diluent was used in place of normal serum
to contain anti-P,, anti-K, anti-M, and anti-Lea, respectively) in the anti-IgG AHG test or the polyspecific AHG test (Table
were consistently negative in tube tests. Four samples (1 each 4). All CAT titer endpoints were equal to or better than tube test
anti-c and anti-Jka and 2 anti-K) were positive in the tube test titers in each of the diluents.
after repeat testing. One sample reported to contain an HTLA
antibody was positive once in CAT and tube tests performed Titrations in IgG-containing diluents
twice. One sample reported to contain anti-Bg was initially
positive in CAT and negative on all other tests. Neutralization of anti-IgG activity with samples containing
abnormally high IgG concentrations was investigated experi-
mentally by the titration of three antibodies in NSP or NSP
Neutralization issues specific to CAT
spiked with human IgG. We used the standard CAT AHG
Two experiments were designed to evaluate the potential procedure described in Materials and Methods to determine the
neutralization of the AHG reagent when the test is carried out titer endpoint of each antibody in each of the diluents using
under conditions designed to simulate routine testing and both CAT anti-IgG and CAT polyspecific AHG tests. We
under conditions in which neutralization conceivably could further challenged the system by adding 40 pL of antibody
occur. In one experiment, antibodies were titrated in diluents diluted in NSP, to which was added another 40 pL of NSP with
containing no IgG, and the results were compared to titrations no antibody, to determine whether changes in the serum-to-cell
in normal serum. In the other experiment, titrations were ratio and ionic strength, together with twice the normal level of
carried out in diluents containing higher than normal levels of immunoglobulins, affected the test results. This test simulates
IgG, and the results were compared to titrations in normal the inadvertent double-addition of one antibody-negative se-
serum. rum and one antibody-positive serum to a single column. Tube
tests were performed in parallel with antibody diluted in NSP.
Titrations in IgG-free diluents We observed no significant endpoint titration differences
when additional serum (80 pL instead of 40 pL) was used or
The CAT AHG test was evaluated for possible neutraliza- when serum spiked with additional IgG was used, as compared
tion under normal test conditions. We titrated four antibodies to titrations in NSP in the anti-IgG AHG test or polyspecific
using 1-in-2 dilutions of normal serum and compared the titer AHG tests (Table 5). All CAT titer endpoints in NSP or NSP
endpoints to similar titrations in diluents containing no IgG. spiked with IgG were equal to or better than tube test titers in
Two IgG-free diluents were used, 6-percent bovine serum NSP.
albumin in saline and serum depleted of immunoglobulins. We
used the standard CAT AHG procedure described in Materials
and Methods to determine the titer endpoint of each antibody Discussion
in each of the diluents using both CAT anti-IgG and CAT CAT AHG tests showed close agreement with conven-
polyspecific AHG tests. Tube tests were performed in parallel
for comparison. tional tube tests in a wide variety of plasma and serum
samples. The CAT system was more efficient than the
tube method at detecting antibodies previously reported
Table 4. Reciprocal antibody titers in IgG-free diluents at initial testing. Samples supplied to us for this study
were intentionally selected to contain weakly reacting
Normal Immunoglobulin- Bovine serum
Antibody Method serum depletedserum albumin (6%) antibodies, and many of the samples had been frozen for
CAT and tube anti-lgG AHG a number of years; consequently, only samples that were
-Fya CAT 128 128 128
Tube 128 64 128
Table 5. Reciprocal antibody titers in IgG diluents
-Jkb CAT 256 512 256 ~~

CAT CAT
Tube 256 256 256 Antibody CAT NSP 80pL Tube
specificity NSP +IgG' NSPt NSP
-D CAT 256 512 256
Tube 128 256 256 Anti-lgG AHG
-Fya 512 512$ 256 256
-K CAT 512 1024 512
Tube 512 256 128 -D 12,800 12,800$ 6,400 3,200

CAT and tube polyspecific AHG -K 640 320 320 160

-Fya CAT 256 256 256
Tube 256 64 128 Polyspecific AHG
-Fya 256 256$ 128 256
-Jkb CAT 512 512 256
Tube 256 256 256 -D 12,800 12,800$ 6,400 6,400

-D CAT 256 256 256 -K 640 320 640 320

Tube 128 256 256 * Normal serum pool spiked with IgG.
t 40 pL of antibody in NSP plus additional 40 pL of NSP.
-K CAT 512 512 512 $ Weak-positive reaction in diluent (NSP + IgG) with cells used for
Tube 256 128 128 anti-Fyaand anti-D titrations.
TRANSFUSION
1993-Vol. 33. No. 8 COLUMN AGGLUTINATION TECHNOLOGY 643

positive in either CAT or tube tests at the time of this study levels of IgG would be expected to have been lower than
were considered positive. Some samples previously those in normal serum. The close agreement between
reported to contain weak antibodies were found to be CAT and tube assays in the single-donor, antibody-
negative by both the CAT and tube methods and were positive and -negative studies as well as in the titration
considered to be negative in this study. After initial studies performed with diluents containing various levels
testing, the tube method missed a number of antibodies of IgG demonstrates that significant neutralization does
that were considered to be clinically significant. These not occur.
samples would not have been investigated further in CAT offers a number of advantages over commonly
routine testing. One anti-K and one anti-c were not detec- used hemagglutination assays. Reagents are already
ted in either the anti-IgG AHG or the polyspecific AHG dispensed, with each column containing reagent for a
tube tests. Two additional anti-K samples and one anti- single test. Consequently, variation in techniques used to
Jk” were not detected in the polyspecific AHG tube test. add reagents and resuspend the red cell button in tubes or
The sample containing anti-Jk” was negative with anti- plates has been eliminated. The need to wash red cells in
IgG reagents in both CAT and tube tests. This antibody the AHG test has been eliminated, which decreases the
appears to be an example of a complement-dependent overall test time and increases biosafety. The agglutina-
anti-Jk”. tion reaction in the column is stable for at least 30
Three samples initially not detected in CAT were minutes, which results in a more objective interpretation
detected in the comparable tube test. Two of these of results. The stability of the agglutination permits
antibodies (anti-PI, cold agglutinin) are usually not con- multiple reviewers to observe and comment on the same
sidered to be significant for the purposes of transfusion. reaction. Another advantage we have found in our
One sample containing clinically significant antibodies laboratory is a significant decrease in the time required to
(anti-E and anti-Fy”) was positive in CAT after repeat train technologists to read and interpret results. In
testing. addition to AHG testing, CAT can be used to type red
Equivalent results were achieved when antibodies cells by adding appropriate red cell typing reagents to the
were titrated in diluents containing either normal or diluent. These tests are currently under development.
excess concentrations of soluble IgG as well as in diluents
that contained no soluble IgG. This demonstrated that References
neutralization of the anti-IgG portion of the AHG reagent
1. Landsteiner K. Zur Kenntnis der antifermentativen, lytischen
did not occur when CAT was performed under standard und agglutinierenden Wirkungen des Blutserums und der
conditions or under conditions designed to stress the Lymphe. Zbl Bakt 1900;27:357.
system. The presence of soluble IgG and/or complement 2. Landsteiner K. Uber Agglutinationserscheinungennormalen
menschlichen Blutes. Klin Wochenschr 1901 ;14: 1132.
components can result in neutralization of the AHG 3. Coombs RRA, Mourant AE, Race RR. A new test for the
reagent. In tube tests, neutralization is prevented by red detection of weak and “incomplete” Rh agglutinins. Br J Exp
cell washing with multiple additions of saline and cen- Pathol 1945;26:255.
4. Lapierre Y, Riga1 D, Adam J, et al. The gel test: a new way to
trifugation and decanting steps, which removes the soluble detect red cell antigen-antibody reactions. Transfusion
serum components before the AHG reagent is added. 1990;30:109-13.
Neutralization of the AHG reagent in CAT is prevented
Kathleen J. Reis, PhD, Senior Scientist, Division of Immuno-
by a different mechanism. The diluents for the AHG tests hematology,Research and Development,Ortho Diagnostic Systems,
contain a high-density polymer, polyethylene glycol or Inc., P.O. Box 350, 1001 US HWY 202, Raritan, NJ 08869-0606.
dextran, which, together with the physical design of the [Reprint requests]
small columns and the surface tensions of the serum and Rosemary Chachowski, BSc, Scientist, Division of Immuno-
diluent at the interface, prevents mixing and subsequent hematology,Research and Development,OrthoDiagnostic Systems.
neutralization of the AHG reagent by serum components. Agnes Cupido, BSc, Associate Scientist, Division of Immuno-
hematology,Research and Development,OrthoDiagnosticSystems.
During centrifugation, red cells are removed from the
Donald Davies, MS, Group Leader, Division of Immunohema-
serum and, as they pass through the bead column, are tology, Research and Development, Ortho Diagnostic Systems.
exposed to the AHG reagent contained in the diluent. If Janice L. Jakway, MA, MT(ASCP)SBB, Associate Scientist,
significant neutralization had occurred under normal test Division of Immunohematology, Research and Development,Ortho
conditions, the titration results in the diluents that lack Diagnostic Systems.
IgG (bovine serum albumin or immunoglobulin-depleted Thomas M. Setcavage,PhD, Director, Division of Immunohema-
tology, Research and Development, Ortho Diagnostic Systems.
serum) would be expected to have been higher than the
Disclosure: The column agglutination tests described in this
results in diluents containing normal serum, while titra- paper are currently sold in Europe by Ortho Diagnostic Systems
tion endpoints in the presence of higher than normal under the name of BioVue.