AABB Antibody Identification

Antibody Identification
By Teresa Y. Harris, MT(ASCP)SBB, CQIA(ASQ)

An excerpted chapter from
Serologic Problem-Solving:
A Systematic Approach
for Improved Practice
Digital
Download
AABB Press ©2005. All rights reserved.
AABB Digital Downloads
System Requirements
• Adobe Reader 6.0 or higher
Download Quantities and Expiration

• The number of times an AABB digital product may be downloaded is specified in the “My Downloads” area
of the online Marketplace and on the product sales invoice.
• The expiration date for downloads is also given in the “My Downloads” area.
Terms and Conditions
Copyright © by AABB. All rights reserved. Reproduction or transmission of text in any form or by any means,
electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system is
prohibited without permission in writing from the Publisher.
The Publisher has made every effort to trace the copyright holders for borrowed material. If any such material has
been inadvertently overlooked, the Publisher will be pleased to make the necessary arrangements at the first
opportunity.
Mention of specific products or equipment by contributors to this AABB publication does not represent an endorsement
of such products by the AABB nor does it necessarily indicate a preference for those products over other similar
competitive products.
AABB authors are requested to comply with a conflict of interest policy that includes disclosure of relationships with
commercial firms. A copy of the policy is located at http://www.aabb.org.
Efforts are made to have AABB publications consistent in regard to acceptable practices. However, for several reasons,
they may not be. First, as new developments in the practice of blood banking occur, changes may be recommended to
the AABB Standards for Blood Banks and Transfusion Services. It is not possible, however, to revise each publication
at the time such a change is adopted. Thus, it is essential that the most recent edition of the Standards be consulted as a
reference in regard to current acceptable practices. Second, the views expressed in this publication represent the
opinions of authors. The publication of this book does not constitute an endorsement by the AABB of any view
expressed herein, and the AABB expressly disclaims any liability arising from any inaccuracy or misstatement.
Customer Service
To purchase books or to inquire about other book services, including chapter reprints and large-quantity sales, please
contact our sales department:
• 866.222.2498 (within the United States)
• +1 301.215.6489 (outside the United States)
• +1 301.951.7150 (fax)
• www.aabb.org>Marketplace
AABB customer service representatives are available by telephone from 8:30 am to 5:00 pm ET, Monday through
Friday, excluding holidays.
AABB
8101 Glenbrook Road
Bethesda, Maryland 20814-2749
Copyright © 2005 by AABB. All rights reserved

In: Rudmann SV, ed.
Serologic Problem-Solving: A Systematic Approach for Improved Practice
Bethesda, MD: AABB Press, 2005
Antibody Identification
TERESA Y. HARRIS, MT(ASCP)SBB, CQIA(ASQ)
1
IN 1996, A SIMILAR VERSION of this text
made several points about antibody identifi-
cation that are still valid today. As in 1996, an-
tibody identification is a complex problem-
solving process. The goal of this process remains to
provide components for transfusion that will benefit—
not harm—the recipient. It is still true that many staff
who have a “gut feeling” about the identity of an anti-
body before completing testing are often right, because Teresa Y. Harris,
developing a hypothesis is an early step in the process MT(ASCP)SBB,
of problem-solving. CQIA(ASQ),
The process of antibody identification today is simi- Immunohematology
lar to what it was a decade ago, but there are now more Reference
tools in the toolbox. Gel, solid phase, and automation Laboratories—
have become much more commonplace. More data Operational Support,
have been published on the clinical significance of American Red Cross,
many antibodies and some new specificities have been Winthrop,
identified. This new information will be incorporated Washington

18 SEROLOGIC PROBLEM-SOLVING
in this discussion where applicable in the antibody

problem-solving process. It is important to note that
this chapter is not intended to be a primer for persons
new to the process of antibody identification. Rather, it
is an attempt to refine understanding of the process for
those experienced in antibody investigations in order
to further develop their problem-solving skills.
Chapter 1 discussed the types of human error en-
countered in complex problem-solving. When dealing
with complex problems such as antibody identifica-
tion, the blood banker must have knowledge of the
tools available that can be applied to the problem. Also,
systematic testing protocols that are carefully designed
to reduce human error must be used to increase the
probability that the solution found through following
the protocols is the correct solution. This chapter re-
views the current tools that can be used in the serologic
investigation of antibody problems and suggests
protocols that will, if applied consistently, increase
efficiency and reduce error.
Overview
The Requirements
The AABB Standards for Blood Banks and Transfusion Ser-

vices does not address the process and content of anti-
body identification. The Standards does require that
“when clinically significant antibodies are detected,
2(p38)
additional testing shall be performed.” There are
additional requirements that address the timing of
samples used in pretransfusion testing in recently
transfused (within the last 3 months) or pregnant indi-
viduals. In these cases, the 3-day rule applies. The
AABB Standards for Immunohematology Reference Labora-
tories contains more detailed requirements for antibody
3
investigation, including :
1. Evaluation of available historical patient informa-
tion.
2. Exclusion of commonly encountered clinically sig-
nificant red cell alloantibodies. When commonly
encountered red cell alloantibodies are not excluded
because of special circumstances, blood released

ANTIBODY IDENTIFICATION 19
for transfusion shall lack the corresponding anti-

gen(s).
3. Reactivity with at least two antigen-positive re-
agent red cell samples and nonreactivity with at
least two antigen-negative reagent red cell samples
(the 2 + 2 rule).
Additional federal regulations for personnel per-
forming the testing also apply.4 Antibody problem-
solving is a complex process. As stated in Chapter 1,
the tendency for human error depends to a large extent
on the context in which the problem-solving takes
place. Protocols to reduce error, therefore, are best de-
veloped for specific environments and allow for appro-
priate institutional variability in antibody problem-
solving procedures.
The Process
Antibody identification is dependent on serologic and

nonserologic components (see Table 2-1). Nonserologic
components include the volume of the sample, the
presence or absence of historical serologic data, the
clinical diagnosis, as well as the type of reagents avail-
able, the routine testing protocols, the experience of the
staff, and so on. These variables are present before any
serologic testing is performed. Serologic components
include the antibody screen and panels, antibody char-
acteristics, testing methods used, and tools available
for problem resolution.
The serologic and nonserologic components overlap
and combine to provide the best means for reaching the
final resolution of the antibody problem and providing
a safe transfusion for the recipient. As they are being
established, institutional protocols can include a num-
ber of queries in order to gather sufficient data for anti-
body resolution, such as the following:
• What is the clinical and serologic history of the patient?
• Where (ie, during what phase of testing, in what re-
agent system, and with which cells) is the reactivity oc-
curring?
• When was the patient last transfused?
• Is the pattern of reactivity more consistent with a single
antibody or multiple antibodies, and with an alloanti-
body or an autoantibody?

Table 2-1. Components of Antibody Identification

Component Elements of the Component
Clinical history Pregnancy Age

Previous transfusion Medications
Stem cell transplant Rate or onset of anemia
Hemolytic disease of the fetus and Diseases
newborn (HDFN) • Autoimmune
Race • Viral
Gender
Serologic history Previously identified antibodies Phenotype
ABO group
Experience and Education Continued competency
competency of staff Continuing education Length and breadth of experience
Training
Sample Identification Age of sample
Appearance (icteric, cold Type (serum vs plasma)
agglutinin, etc)
Amount available
Routine antibody Use of tube, gel, or red cell Commercial or in-house panel
identification adherence (solid phase) Washed or unwashed reagent red
protocols Reagent or method used for the cells
initial testing (LISS, albumin, Use of polyspecific or anti-IgG
PEG, or saline) Criteria for autocontrol setup
Performance of an initial spin step Rule-out expectations
Full crossmatch, initial spin
crossmatch, or computer
crossmatch
Incubation times
Impact of testing Effect of temperature or pH Impact of the AHG used
methods Effects of chemicals or enzymes Tools available (eluate,
on the reactivity adsorptions, neutralization,
etc)
Antibody Phase of reactivity Pattern of reactivity
characteristics Hemolytic property Variability of reactions
Strength of reactivity
Physical appearance of the
reactivity (refractive clumps,
stringy, etc)
Antigen characteristics Number of sites on the red cell Variant structures (point
Interaction with other red cell mutations, etc)
surface structures (steric
presentation)
AHG = antihuman globulin; LISS = low ionic strength saline; PEG = polyethylene glycol

Both the serologic and the nonserologic components

are part of the art and science of antibody identification
or resolution. Some suggested key steps in the process
are listed in Table 2-2. The components and elements in
Table 2-1 are all inherent within these steps.
Nonserologic Components of Antibody Identification
Clinical History
While collecting patient history is the very important

first step in antibody investigation, the history is often
overlooked later in the process. Some factors that
should be considered in the patient history include, but
are not limited to:
• Transfusion and/or transplant history.
• History of (or current) pregnancy.
• Medication history.
• History of infections (eg, recent viral infections).
• Clinical and/or laboratory evidence of hemolysis.
• Age and gender.
• Race and ethnicity.
• Results of previous serologic tests and diagnoses.
• Other pertinent clinical and laboratory data.
Table 2-2. Steps in Antibody Identification

Step Process
1 Gather information.
2 Process and test sample.
3 Interpret test results and information.
4 Look back to make sure resolution is consistent
with data.
5 Use information gained in resolution to select
appropriate blood component for transfusion.

Transfusions, Transplants, and Pregnancies
Obviously, a patient who has a history of pregnancy

and/or transfusion has been exposed to foreign red
cells and is more likely to have a clinically significant
antibody. An antibody found in the serum of a patient
who has not been transfused or pregnant is thus a natu-
rally occurring antibody and more likely to be a clini-
cally insignificant antibody.
Pregnancy can often be accompanied by a weaken-
ing or disappearance of the Lewis antigens. Antibodies
to the corresponding antigens often develop. Thus,
pregnant women commonly have Lewis system anti-
bodies. When evaluating a sample with variable reac-
tivity from a patient who is pregnant, the blood banker
can anticipate that the cause of the reactivity could be
a bH
anti-Le and/or anti-Le . It has been the experience of
some investigators (as well as the author) that anti-IH
5
can also be found often in pregnant women.
The patient’s transfusion and/or transplant history
can provide clues regarding the nature of the signifi-
cant antibody. In general, transfused patients who
form strongly reactive antibodies are more likely to be
forming alloantibodies such as anti-E or anti-K. Weak
reactivity with all cells, including the autocontrol, in a
recently transfused patient is more likely to be an anti-
body such as anti-Ch or anti-Kn (traditionally called
“high-titered, low-avidity-like” antibodies) or a weak
warm autoantibody.
Recent transfusion can also result in simultaneous
6(p961)
production of auto- and alloantibodies. One such
a
report described alloanti-Fy accompanied by a mim-
b 7
icking autoanti-Fy in two Fy(a–b+) patients. Many
stem cell and marrow transplants have shown exam-
ples of ABO discrepancies and, more pertinent to this
chapter, some donor cell-derived antibodies. There have
also been reports of antibody production (such as
anti-D) from passenger B lymphocytes after solid or-
8
gan transplants.
Medication History
Drug or medication history can be very informative in an-

tibody resolution. Recent infusion of IV immunoglobu-
9,10 9,11,12
lin (IVIG), IV Rh Immune Globulin (IVRhIG), Rh
9 13
Immune Globulin (RhIG), antilymphocyte globulin,

14
or antithymocyte globulin can result in unexpected
serologic reactivity caused by passive administration
of blood group antibodies. Table 2-3 lists some of the se-
rologic problems caused by these therapies.
Because both IVIG and IVRhIG may be used to treat
autoimmune thrombocytopenia, an unexplained serum
antibody in a patient with this diagnosis may be ex-
plained by the use of these therapies. Antibiotic admin-
istration, especially second- and third-generation
cephalosporins, can often result in a detectable se-
rum/plasma autoantibody.15 Patients receiving these
antibiotics often show brisk hemolysis. Vaccination
may also have an association with the formation of
16(p99)
warm autoantibodies.
History of Infections or Hemolysis
A history of viral illness may suggest the presence of

cold agglutinins such as anti-I or anti-i. While these are
the most common specificities, other antibodies have
Table 2-3. Possible Serologic Outcomes of Therapies

Outcome
Therapy (Detectable Antibodies)
IV immunoglobulin Anti-A, anti-B, anti-D,

antibodies to other red cell
antibodies, and antibodies
to viral markers
IV Rh immune globulin Anti-D in a D+ patient
Anti-A, anti-B, anti-D,
antibodies to other red cell
antibodies, and antibodies
to viral markers
Rh immune globulin Anti-D in a D– patient
Cephalosporin, procainamide, Warm autoantibodies
alphamethyldopa
Antilymphocyte globulin/ Heterophile antibodies that
antithymocyte globulin may show Lutheran-related
specificities

been formed, reportedly as a result of viral disease. In

addition, autoantibodies, as opposed to alloantibodies,
are often associated with disease processes. Autoimmune
disorders such as Hodgkin’s disease, lymphoma, and
systemic lupus erythematosus are often associated with
warm autoantibody formation. Several reference books
are available that provide information on the effect dis-
17,18
eases may have on both antibodies and antigens.
Hemolytic disease of the fetus and newborn (HDFN)
is by definition associated with a red cell antibody of
maternal origin. If a neonate has an unexplained low
hematocrit or demonstrates serum icterus, the labora-
tory may request a sample in order to screen for red cell
antibodies.
If the clinical picture supports a diagnosis of HDFN,
but the results of an antibody screen are negative, more
work may be needed to determine the etiology of the
disease. Results of the antibody detection test (screen)
and panel can be negative in cases of HDFN due to
ABO incompatibility or antibodies to low-incidence
antigens. Special efforts, such as testing the biological
father ’s red cells with the mother ’s serum/plasma
and/or the neonate’s eluate, may be the only serologic
way to confirm the diagnosis of HDFN. However, if a
neonate has no clinical signs of HDFN or hemolysis, lit-
tle is gained by performing a direct antiglobulin test
(DAT) or elution procedure.
The rate of the onset of anemia can often prompt a
hypothesis regarding the type of serum/plasma anti-
body that will be present. A patient without a history of
recent transfusion who presents with a low hemoglo-
bin, no symptoms of bleeding, and a positive antibody
detection test may have a compensated anemia due to a
warm or cold autoantibody. On the other hand, a pa-
tient who has been recently transfused and has a rap-
idly decreasing hematocrit is likely to have a clinically
significant alloantibody.
A recent transfusion can often complicate the sero-
logic picture in patients with warm autoantibodies. The
problem-solving may be made more difficult and time-
consuming because of the presence of a mixed cell pop-
ulation (patient cells and donor cells), necessitating
more complex testing strategies such as allogeneic
adsorptions. In addition, any decrease in hemoglobin
values may be the result of the autoantibody and/or
the formation of new alloantibodies.

Demographics
Race or ethnic background may be helpful factors in

the identification of alloantibodies, especially those di-
rected against high-incidence antigens. Antibodies to
b a a
the U, Js , Hy, Jo , At , or Rh46 antigens are almost
exclusively found in the Black population. A White pa-
tient is more likely to have an antibody to high-inci-
dence antigens such as Coa, Vel, Lub, k, Kpb, Yta, Sc1, or
a b
Gy . Native Americans may form anti-Di , and Asian
populations may form anti-Jk3.
Race or ethnic background also play a part in what
common antibodies may or may not be formed. Asian
populations rarely form anti-D, because Rh-negative
individuals are uncommon. Black populations are more
b a
likely to form anti-Jk than anti-Jk , based on antigen
19(pp289-360)
frequency [see the AABB Technical Manual for
antigen frequencies]. In addition, Fy(a–b–) Blacks form
anti-Fyb at a much lower rate than Fy(a–b–) Whites, be-
cause of the genetic differences in the development of
the Fy(a–b–) phenotype. Genetically, Blacks often pos-
sess the Fyb gene but do not encode for the insertion of
b b
the red-cell-borne Fy antigen. Thus Fy is not recog-
nized as being a foreign antigen, even when the patient
6(pp444-6)
types Fy(b–).
The age and gender of the patient can have some ef-
fect on certain antibodies, as one might expect. Older
patients are more likely to have had more life experi-
ences that exposed them to immunizing events, such as
transfusion or pregnancy. Autoantibodies are more
common in older patients who may have altered im-
mune systems. One antibody, anti-JMH, is more fre-
quently found in older patients.6(p808)
The patient’s gender plays only a small part in anti-
body identification. In the untransfused patient, a
female is more likely to have a clinically significant an-
tibody than a male patient, because of the possible his-
tory of pregnancy. Due to the frequency differences of
the Xga antigen (X-chromosome controlled), anti-Xga is
more commonly found in men. Approximately 34% of
males are Xg(a–), compared with only about 11% of the
female population.20(p340)
Even before more extensive serologic testing is per-
formed, a good clinical history can supply clues to the
possible causes of a positive antibody screen. Simple

information gathering is a good start for the process of

problem-solving in an antibody investigation.
Serologic History
Occasionally a transfusion reaction results from an un-

detectable antibody that was previously identified.
Such reactions can be caused by a number of human er-
rors, including the following:
• The blood banker failed to check the records.
• The records were incomplete or inaccurate.
• The information was not requested from another trans-
fusion service, blood bank, or reference laboratory.
The best time to determine if the patient had any se-
rologic history is at the time the order for transfusion is
originally received. Clinically significant antibodies,
insignificant antibodies, typing problems, and other
serologic abnormalities can be more efficiently ap-
proached if the nature of the problem is known at the
onset. Initial data gathering (historical serologic re-
sults) can allow for an adequate sample(s) to be col-
lected, if needed, and a physician to be forewarned of
any anticipated delays in providing blood.
The usefulness of historical serologic data is clear in
the following example: When it is learned from a his-
tory check that a patient has an anti-e, the blood banker
sets up an initial selected panel of e-negative cells.
Without this history check, the blood banker would do
an initial panel that would be, for the most part, reac-
tive, and additional cells would need to be tested. A se-
lected panel at the onset would rule out additional
alloantibodies and help release blood to the patient
more efficiently.
ABO type can be an important key in antibody iden-
tification. A group A patient who has a positive anti-
body screen, but has negative crossmatches with group
A donor units, presents the classic case of having either
bH
anti-IH or anti-Le . Low levels of available H antigen
on group A1 individuals will make the autocontrol neg-
ative, as well as most group A units nonreactive, when
either of these antibodies is present. A2 red cells have a
bit more H antigen and may react weakly, but not as
strongly as the group O red cells. On the other hand, if the
group A patient has a negative antibody screen but the
crossmatches are positive, anti-A1 may be the problem.

Some of the other specificities that are associated

with ABO are less common, but include anti-IA or
anti-IB. These antibodies will react only with red
cells that have both the I and A (or B) antigens to-
gether. For example, serum from a group B patient
with anti-IB will react with adult B cells (that have
both I and B) and will be negative or weaker with
group O cells (that have I but not B). So too, these an-
tibodies will react more weakly with cells that have a
diminished expression of the I antigen, such as cord
blood cells.
The Rh type generally affects only whether the corre-
sponding antibody (anti-D) may be present. When an
antibody screen is positive in a patient known to be
Rh-negative, anti-D continues to be the most likely
specificity. However, to determine whether the anti-
body has been actively formed or passively received
(through RhIG administration) a review of the clinical
history is necessary.
Records of complete or partial phenotypes are often
helpful. With a record of the patient’s phenotype, it
may be known what antibodies could be formed by the
patient, based on those antigens the patient lacks.
Phenotyping patients who are likely to be chronically
transfused, such as thalassemia and sickle cell anemia
patients, is very helpful before transfusion. If a pa-
tient’s phenotype indicates a person is positive for cer-
tain antigens, and this is combined with the results of
serum panels, the data may provide converging evi-
dence to increase the probability that a solution is
valid. One limitation of a “complete” phenotype is that
it will not help predict or aid in the resolution of the
more esoteric antibody specificities—such as anti-Doa,
anti-Dob, and anti-Cob—for which phenotyping results
would not be available.
Serologic history and clinical history combined can
often help in efficient resolution of an antibody identi-
fication problem. A complete history provides a great
deal of information about a patient. Inferences can be
made about the nature of an antibody even before the
sample is drawn.
For example, a group A, D-negative, pregnant pa-
tient who has had a previous pregnancy is much more
likely to have antibodies (anti-D, other clinically signif-
icant antibodies, and specificities such as anti-IH and
anti-LebH) than a group O, D-positive male youth who

has never been transfused, but presents with a positive

antibody screen.
An internal record check, consultation with hospi-
tals that have previously seen the patient, or informa-
tion from a local reference laboratory can make a
potentially complex problem much easier to resolve.
Experience and Competency of the Staff
Although many available resources cover the topics of

training and competency assessment, it is important to
mention the importance of good training, on-the-job
experiences, and competency in antibody identifica-
tion. All of these nonserologic variables play a part in
the ability of the staff to arrive at a problem’s correct
resolution most efficiently. Because antibody identifi-
cation is a complex task, the blood banker’s content
knowledge and job competency are very important. Ef-
ficiency, accuracy, and effective troubleshooting in
problem-solving result from the staff member’s solid
understanding of each component. For example,
knowledge about why a certain reagent works, as well
as about the pros, cons, and pitfalls associated with the
use of the reagent, is invaluable in the problem-resolu-
tion process.
A useful competency evaluation tool is a grading or
scoring exercise, in which all staff performing antibody
detection and identification are given several samples
to grade or score. Samples for these types of exercises
are usually prepared by diluting a known antibody-
containing plasma or serum. Once the samples are
coded, all staff should test the samples against the
same one or two red cell samples. This helps eliminate
variation because of differences in the antigen strength
of red cell samples. At least one nonreactive sample
and one weakly positive sample should be included. It
is also desirable to have some samples, if performing
tube testing (as opposed to gel or solid phase testing),
that require an indirect antiglobulin test (IAT) and oth-
ers that do not.
Ideally, staff should obtain results that are within
one grade of each other (eg, 1+ to 2+). This process en-
sures that results are closely reproduceable from one
staff member to another and helps evaluate staff for
proper technique. Even experienced blood bankers can

fine-tune their technique. It can be beneficial to share

the results of successful grading/scoring competency
evaluations among staff so that they can gain confi-
dence in their coworkers’ ability to get similar results.
This easy competency evaluation should be performed
at least once a year. Also, one source of educational
information that is often overlooked is the package
inserts for reagents used in the laboratory.
In addition to on-the-job training, resources such as
local and national conferences, journal clubs, and cur-
rent reference books provide essential continuing edu-
cation that will keep the blood banker ’s problem-
solving skills current and sharp. Facilities that decide
to cut continuing education spending or support
should consider that such action may be short-sighted.
Continuing education enables the staff to strengthen
their grasp of the various elements involved in anti-
body identification. Much time (and hence money) can
be saved when, in a problem-solving task, staff are able
to proceed to the next step in the process more quickly
because of the refined knowledge they possess. A lot of
unnecessary testing and referrals to outside reference
laboratories for problem resolution can be avoided.
Clearly, education, training, and competency are
essential components of the process of antibody identi-
fication.
The Sample
Several steps should be performed each time a sample

is obtained for pretransfusion testing and/or antibody
identification:
• Evaluate the sample for proper labeling.
• Check the date drawn to make sure the correct sample
is being used.
• Examine the sample to be sure it is the correct type of
specimen [with ethylenediaminetetraacetic acid (EDTA),
or clot vs heparin, or other anticoagulant].
• Examine the sample for hemolysis, icteris, clumping,
etc.
Accurate sample identification in any pretransfusion
test is essential for accurate results. Recently, much at-
tention has been given to improving the sample iden-
tification process, including the increased use of
barcodes and barcode readers. Unfortunately, prob-

lems still occur. If a sample is mislabeled or illegible,

it should not be used for pretransfusion testing.
The appearance of the samples can often provide a
clue to the type of problem that may be encountered.
For example, the following should alert the blood
banker to a patient problem that may be related to
transfusion: 1) icteric serum/plasma caused by an ele-
vated concentration of bilirubin, 2) serum/plasma
with a brownish-green tinge produced by the presence
of bilirubin in combination with methemalbumin, or 3)
serum/plasma with a pinkish or red tinge caused by
free hemoglobin. Free hemoglobin usually indicates an
overwhelming and rapid intravascular destruction of
red cells (as in ABO incompatibility), while icterus is
usually associated with slower extravascular destruc-
tion (as in delayed hemolytic reactions).
When plasma or serum shows any such marked dis-
coloration, an effort should be made to demonstrate the
possible cause. Enhancement media and special tech-
niques outside of normal protocols may need to be
used in such cases when the antibody screen is nega-
tive or weak. It has been the author’s experience, for ex-
ample, that a significant number of patients presenting
with brownish-green plasma are experiencing a trans-
fusion reaction due to a Kidd blood group system anti-
body.
Other instances of an unusual appearance of the
blood sample can help in antibody identification. The
presence of a cold agglutinin (with a “lava-lamp” ap-
pearance) in the samples should warn the blood banker
of possible pretransfusion testing problems. The pres-
ence of cryoglobulins may indicate abnormal protein
levels in the serum and may result in rouleaux
formation.
The amount of available sample can have an effect on
the selection of antibody identification procedures.
Current testing protocols that use gel or solid-phase
methods require a much smaller sample. In order to
maximize the sample, a blood banker encountering a
small sample may choose to perform a red cell pheno-
type first. Then, based on the phenotype, a selected
panel may be performed to rule out only those
specificities that the patient may form. Restricted sam-
ple volumes may increase the amount of time required
for the problem-solving process, because the blood
banker must evaluate each result before going to the

next problem-solving technique. Contingency plans

for this situation should be addressed in procedure
manuals. For example, a facility may choose to use one
reagent red cell for the rule-out process instead of the
routine two panel cells. (See more on the rule-out pro-
cess later in this chapter.) Whenever possible, however,
it is better to obtain an additional sample than to resort
to nonroutine testing protocols.
The age of the sample is important because fresher
samples offer a better chance of detecting weak anti-
bodies. Plasma or serum samples that are to be stored
for testing for more than 2 to 3 days should be frozen.
While it is not a significant issue in today’s testing pro-
tocols, older samples will show decreased complement
levels. When performing red cell phenotyping, as part
of the problem-solving process, care must be taken to
test within the manufacturer’s sample requirements; in
some cases this requires using samples less than 2 days
old.
The type of sample used can also affect the type of
reactivity that will be observed. A decade ago, serum
was the most common type of sample used. In the
past, blood bankers believed there were certain vir-
tues in using serum over plasma. Today, plasma is the
sample of choice in the majority of laboratories. Staff
at laboratories that use plasma are not able to observe
cold-reactive antibodies that bind complement in
a serum test environment. EDTA chelates calcium,
preventing complement activation and avoiding ag-
gregation dependent upon low ionic strength saline
6(pp1128-9),21
(LISS).
There are antibodies that will react in plasma but not
serum, and vice versa. One commonly recognized phe-
nomenon is that an apparently benign autoantibody
will react in a clotted sample, but not in an EDTA sam-
ple. This autoantibody is often associated with ulcer-
ative colitis, although the author has seen a few of these
antibodies present in pregnant women as well. It may
be that these benign antibodies are directed toward an-
tigens generated during the clotting process. It is worth
noting that ulcerative colitis is also associated with au-
16(pp84-8)
toimmune hemolytic anemia ; in cases of this dis-
ease, patients present with an autoantibody reactive
in both the plasma and serum and patients will have
a positive DAT (unlike the aforementioned phenome-
non associated with ulcerative colitis). Other anti-

bodies may be reactive only in the presence of

22
EDTA. In these cases, ionized calcium found in the
serum inhibits antibody reactivity.
The sample can play a key role in antibody identifi-
cation. It can forewarn the blood banker of a problem
and affect the choice of approach to a problem. The
type of sample (plasma vs serum) can prevent some
problems from occurring and can occasionally explain
why one facility may find a positive antibody screen
for a patient while another facility finds a negative
reaction.
Serologic Components of Antibody Identification
Antibody Detection and Identification Proto-

cols
There are many standard approaches to antibody de-

tection (antibody screening), one of the first steps in se-
rologic investigations. When there is a negative
antibody screen during initial routine testing, even
when the patient has had a history of an antibody, addi-
tional antibody identification procedures are not war-
ranted. In all cases, whether or not a historical antibody
is currently demonstrable, units for transfusion must
be negative for the antigen that corresponds to the his-
torical antibody.
The protocols that laboratories choose will affect
what they subsequently detect and, thus, what they
need to investigate. To minimize human error and en-
sure optimal efficiency and safety, protocols need to be
cost effective and tailored to the experience of the staff
and the general patient population encountered. Ap-
proaches to positive results on initial antibody screens
and to antibody identification should be reasonably
standardized, but flexible enough to allow for the se-
lection of more efficient protocols, depending on the
clinical and serologic history of the patient.
Many published observations praise or criticize one
or more of the different reagents or methods. However,
by following the guidelines presented in this publica-
tion to reduce error, an experienced blood banker can
use any of the media and usually detect most clinically
significant antibodies that are present.

Currently there are three main ways to investigate

antibodies: test tube, gel, and solid-phase testing. Tube
testing, whether it is saline, albumin, LISS, or polyeth-
ylene glycol (PEG), tends to be more flexible and is still
used by many laboratories (enhancement media is dis-
cussed in more detail later in the chapter). Gel and solid
phase are newer methods that are being used more and
more frequently in the field, because they may be auto-
mated and thus less technique-dependent. Descrip-
tions of the principles of the tests and the pros and cons
23
of each method have been well published. Even those
who use gel or solid phase usually maintain a backup
tube testing process for troubleshooting. Gel and solid
phase eliminate a number of the technique-dependent
variables that are associated with serologic testing by
tube methods. Gel and solid-phase results are stable
and can be held until the blood banker can consult a
more experienced technologist or supervisor when
weak or unusual results are found. As discussed in
Chapter 1, the use of a second person as an independ-
ent judge helps to detect error and decrease the
incidence of a final solution that is incorrect.
A disadvantage of these test methods is that, like the
PEG tube test, they often enhance findings that lack
clinical significance for transfusion, such as clinically
benign autoantibodies or weak reactivity (as in the case
of anti-Ch). This may result in additional work for the
laboratory and cause delays.
As stated earlier, when laboratory management staff
develop protocols, they need to take into account the
experience of the staff in troubleshooting and the type
of patients they see. For example, a hospital with a
large number of special-needs patients—such as oncol-
ogy patients, sickle cell patients, and others that are
transfused frequently—needs to balance the higher in-
cidence of problems because of the increased sensitiv-
ity of gel, solid-phase, or PEG tube testing with staff
competent to use the techniques. Issues of staff exper-
tise, efficiency, and productivity must be considered in
the overall context of ensuring the best clinical out-
comes for patients.
All testing protocols should be reviewed to find
ways of improving efficiency and conserving re-
sources, while not compromising patient safety. It is of-
ten useful to look beyond the reagent that is used to the
amount of technical time spent. The following list in-

cludes some issues that could be considered when

selecting antibody identification methods and devel-
oping problem-solving protocols:
• Making an in-house panel is usually not an effective
use of resources.
• Reading every panel cell that is weak-positive or -nega-
tive microscopically is generally not warranted and
may result in useless troubleshooting for reactivity that
is not clinically significant.
• Many laboratories test quite a few cells to prove anti-
body specificities. In reality, a new antibody can be
proven quite successfully on two cells that are anti-
gen-positive and reactive, and two cells that are anti-
gen-negative and nonreactive. As stated earlier, this is
3
the minimum requirement of AABB IRL Standards.
Testing a third cell offers very little chance of increas-
ing the probability of accurate antibody identification.
• Routine use of an autocontrol is not necessary for an
antibody screen. Even an initial panel may not need an
autocontrol, because it can be set up in later testing
when a panagglutinin is encountered.
• A two-cell screen (with homozygous expression of
Kidd antigens) may be considered instead of a three-
cell screen. If only a small percentage of screens are
positive, each additional gel or tube setup that is per-
24
formed routinely may yield little benefit.
• A DAT need not be performed unless 1) it is ordered by
the physician, 2) the patient is hemolyzing, and/or 3)
the laboratory encounters a panagglutinin.
• Eluates are rarely necessary to improve patient care. If
results of the eluate do not change how a patient is
treated, why perform an eluate? For example, a group
A neonate has a positive DAT and the mother is group
O with a negative antibody screen. Performing an
eluate to show that the infant has anti-A coating his or
her cells will not change the fact that the patient will re-
ceive group O cells, and the high bilirubin will be
treated if present.
In summary, each institution should develop differ-
ent protocols based on its patient population, staff ex-
perience, and the best means of conserving scarce
resources. The optimal protocol should then be se-
lected as the standard approach for patient workups
(see Table 2-4 for an example of a protocol). Safe trans-
fusions are more likely to result from ensuring the staff

Table 2-4. Example of a Standard Antibody Testing Protocol

Situation Procedure
ABO group Test anti-A and anti-B vs patient cells; A1 and B cells vs patient’s
EDTA plasma.
Rh type Perform anti-D immediate spin only.
Antibody screen Use EDTA plasma against a two-cell set, screening cells in gel.
First panel Use EDTA plasma vs gel panel.
DAT Perform if requested by the physician.
When first panel is • If a clear pattern appears, perform the rest of the identification by
positive gel (with selected cells). Phenotype the patient for corresponding
antigens for which the patient has formed antibodies.
• If weak variable reactions or panagglutinin appear, test the same
panel by LISS tube IAT. If the panel is negative by tube test, do not
do further work.
Full phenotype Perform when…
• The patient has not been recently transfused but is expected to
receive many transfusions in the future.
• A patient has formed more than one antibody.
DAT = direct antiglobulin test; EDTA = ethylenediaminetetraacetic acid; IAT = indirect antiglobulin test;
LISS = low ionic strength saline
are well trained than from testing protocols that in-

volve multiple media and extensive testing. More is not
necessarily better.
Antibody Characteristics
An important early step in antibody problem-solving is

the formation of an impression or initial hypothesis
based on the characteristics of various antibodies. This
initial hypothesis is informed by the knowledge and
experience of the blood banker, the nature of the pa-
tient population, the patient history, and the results of
routine serologic typing and screening. While informa-
tion gathered through routine protocol may be differ-
ent from one laboratory to another, it is useful to know
what those protocols will or will not detect. Generally,
evidence for an initial hypothesis is based upon the fol-
lowing observations:

• Stage of testing in which the evidence of an antibody

was observed.
• Phase(s) of reactivity.
• Strength of the reactions.
• Pattern of reactivity.
• Variability in reactions from cell to cell or from phase to
phase.
• Presence or absence of hemolysis.
• Nontypical observations about the agglutinins (eg,
mixed field or weak and easily dislodged).
• Physical appearance.
Testing Stage and Reactivity
The stage (and phase) of testing in which the original

reactivity is detected gives essential information in the
antibody identification process. An incompatible
crossmatch after a negative antibody screen could be
the result of:
• An ABO incompatibility between the recipient and the
donor cells.
• An antibody showing the dosage effect (if the screen-
ing cells do not have a double dose of the antigen, but
the donor cells do).
• Anti-A1.
• A low-incidence antigen present on the donor’s cells
that corresponds to the recipient’s antibody.
• A positive DAT on the donor unit, or a polyagglutin-
able donor unit.
• An antibody reactive only by tube technique, if the
crossmatch is performed by tube test and the screen is
performed by gel or solid phase.
If the reactivity is seen only with the antibody screen
but not with the crossmatches, the source of the prob-
lem could be:
• An antibody that has affected the ABO group, such as
bH
anti-IH or -Le .
• An antibody reacting to an antigen of relatively low
b b
frequency such as anti-K, Yt , or Co .
• An antibody showing the dosage effect (if the screen-
ing cells have a double dose of the antigen but the do-
nor cells do not).

• A preservative-dependent reactivity (preservative in

the reagent red cells) that would affect unwashed re-
agent red cells but not the donor’s red cells suspended
in saline.
• An antibody reactive only by the gel or solid-phase
technique, if the crossmatch is performed by tube test
and the screen is performed by gel or solid phase.
Phase of Reactivity
A summary of antibody serologic reactivity is found in

Table 2-5. Immediate-spin reactivity (reactivity at room
temperature with no incubation) is often seen with
a
cold-reactive allo- or autoantibodies. Anti-P 1 , -Le ,
bH
-Le , -M, -I, -IH, and -A1 can often cause immedi-
ate-spin reactivity in crossmatches, screens, and/or
panels. In gel or solid-phase testing, or when reading
tubes only at 37 C/IAT, these antibodies may not be de-
tected. However, if a negative screen is seen when us-
ing one of these protocols, but subsequently the
immediate-spin crossmatch is positive, one of these
specificities should be suspected. Occasionally, anti-E,
anti-K, and other traditionally clinically significant
specificities can also react during immediate spin.
Whenever immediate-spin reactivity occurs, the pat-
tern should be examined. If no pattern can be discerned
and the antibody reactivity does not carry over to 37 C
incubation and to IAT, no further search for identity is
needed unless the patient exhibits unexplained clinical
signs of hemolysis.
Because immediate-spin-reactive anti-M or anti-P1
can often cause troublesome interference with ABO se-
rum/plasma grouping (reverse grouping), care must
be taken to ensure the reverse ABO group is valid. Use
of neutralizing substances (for anti-P1 or Lewis blood
group antibodies); antigen-negative, reverse-grouping
cells (eg, P1 negative); cold autoadsorption (anti-I); rab-
bit stroma adsorptions; or prewarm techniques may
help resolve reverse-grouping discrepancies.
Most, but not all, antibodies reactive during immedi-
25(pp100-2)
ate spin are IgM. Occasionally, IgM antibodies in
the “window” period of the primary immune response
may be primarily reactive during immediate spin. A
possible pattern of reactivity, such as the reactivity of

38
Table 2-5. Serologic Characteristics of Red Cell Antibodies

Effect of
Enzyme Effect of DTT
1 2
Treatment of Treatment of
Red Cells on Red Cells on
Blood Group or Immediate 37 C Direct Reactivity Reactivity
Collection Antibody Spin Testing IAT (Enz) (DTT) Comments
ABH A1 R OR OR IR NE Anti-A1, BI, or AI may cause

(Note: ABO and H AI, BI R OR OR IR NE incompatible crossmatches
are different HI (or IH) R OR OR IR NE when the antibody screen
blood group H R OR R IR NE results are negative.
SEROLOGIC PROBLEM-SOLVING
systems but Anti-IH may react in the antibody

they interact screen but will not normally
closely.) react with group A1 red cells.
Anti-H can be hemolytic.
a
Cartwright Yt N N R V NE/W
Ytb N N R V NE/W
Chido/Rogers Ch N N R D NE Plasma antigens; antigens reside
Rg N N R D NE on C4d.
a
Colton Co N OR R OIR NE Co(a–b–) cells are the null red
Cob N OR R IR NE cells.
b
Anti-Co is often found in patients
with multiple antibodies.

a
Cromer Cr N N R NE NE/W The Inab phenotype represents the
Tca N N R NE NE/W null red cells.
a
Other high-incidence antigens: Cr ,
a a a b
Tc , Dr , Es , IFC, and WES .
b
Other low-incidence antigens: Tc ,
c a
Tc , WES , UMC, and GUTI.
Diego Dia OR OR R NE NE Other high-incidence antigens: Dib
Dib N OR R NE NE and Wrb.
a
Other low-incidence antigens: Di
(but 11% to 36% in some
a a a
populations), Wr , Wd , Rb ,
a a
WARR, ELO, Wu, Bp , Mo ,
a a a
Hg , Vg , Sw , BOW, NFLD,
a a a
Jn , KREP, Tr , Fr , and SW1.
a
Dombrock Do N N R IR V Gy(a–) cells are the null red cells.
b a b
Do N N R IR V Anti-Do and anti-Do are often
Gy N N R OIR V weak antibodies and found in
Hy N N R OIR V patients with multiple
a
Jo N N R OIR V antibodies.
a
Duffy Fy OR OR R D NE Fy(a–b–) cells are the null red
Fyb RR OR R D NE cells.
Fy3 N N R NE NE
Globoside P R R R IR NE P is a high-incidence antigen.

I I R OR OR IR NE Strong anti-I can often be found in
patients after viral illness and
may be associated with other
disease processes.
Ii (collection) i R OR OR IR NE Anti-i can often be found in
patients after viral illness and
may be associated with other

disease processes.
a
Indian In N N R D D
Inb N N R D D
(Continued)
Note: Antigens, antigen frequencies, and blood group systems/collections not mentioned on this table can be found elsewhere.6,19,20
R = reacts; N = normally nonreactive; OR = occasionally reacts; RR = rarely reacts; IR = increased reactivity; OIR = occasionally increased reactivity; NE = no effect; D = destroyed; W/D = weakened/destroyed;
ANTIBODY IDENTIFICATION
V = variable; NE/W = no effect/weakened

39
40
Table 2-5. Serologic Characteristics of Red Cell Antibodies (Continued)

Effect of
1 2
Kell K OR OR R NE D Ko cells are the null red cells.

k OR OR R NE D Anti-K may not react well in
a
Js N N R NE D LISS-based testing.
Jsb N N R NE D Other high-incidence antigens: Ku,
Kpa N N R NE D K11, K12, K13, K14, K16, K18,

Kpb RR RR R NE D K19, Km, K22, TOU, and RAZ.
a
Other low-incidence antigens: Ul ,
c
K17, Kp , K23, K24, and VLAN.
a
Kidd Jk RR OR R IR NE Jk(a–b–) cells are the null red
Jkb RR OR R IR NE cells.
Jk3 RR OR R IR NE The antibodies in this system
often disappear in the patient’s
serum and may cause severe
delayed transfusion reactions.
a
Knops Kn NR NR R W/D W/D The Helgeson type Kn(a–b–) cells
Yka NR NR R W/D W/D are the null red cells, because
a

McC NR NR R W/D W/D of a low number of CR1
a
Sl NR NR R W/D W/D (complement receptor) on red
cells. Antigens reside on CR1.
In(Lu) cells and some patients
with autoimmune diseases
may show reduced expression
of Knops system antigens.
Lewis Lea R R R IR NE Le(a–b–) cells are null red cells.
bH
Leb/bH R R R IR NE Anti-Le will demonstrate a
positive antibody but negative
group A or AB crossmatches.
The antibodies can often be
hemolytic and are often found
in pregnant women.
a
Lutheran Lu R OR R W/D W/D Lu(a–b–) cells are the null red
Lub RR RR R W/D W/D cells. The In(Lu) gene causes
all Lutheran antigens to be
markedly depressed.
There are multiple high- and
low-incidence antigens in the
blood group system.
k k
MNSs M R R R D NE M M cells are the null red cells
N R OR OR D NE (lacking GYPA, GYPB, and
S OR RR R V NE GYPE genes). MN is
s RR OR R V NE associated with GYPA; Ss is
U RR RR R NE NE associated with GYPB.
Some examples of anti-s will react
best at IAT after a room-
temperature or 4 C incubation.
There are multiple low-incidence
antigens in this system
because of single nucleotide
mutations and/or hybrid

glycophorins.
P P1 R R R IR NE
(Continued)
Note: Antigens, antigen frequencies, and blood group systems/collections not mentioned on this table can be found elsewhere.6,19,20

41
42
Table 2-5. Serologic Characteristics of Red Cell Antibodies (Continued)

Effect of
1 2
Rh D RR OR R IR NE Rhnull cells are the null red cells.

C RR OR R IR NE There are many compound
E RR OR R IR NE antigens [eg, f(ce), RH7(Ce),
s
c RR OR R IR NE V(ce ), RH22(CE), and

e RR OR R IR NE RH27(cE)]; variant antigens
W variant
C RR OR R IR NE (eg, D , evariant, and EW);
G RR OR R IR NE antigen deletions (eg, D– –);
low-incidence antigens (eg,
X a
C , V, Go , and Rh32); and
high-incidence antigens (eg,
Nou, See, and Dav) in this
blood group system.
Scianna Sc1 NR NR R NE W/D Sc:-1,-2,-3 cells are the null red
Sc2 NR RR R NE W/D cells.
Other antigens in the system: Sc3
and STAR (high incidence)

and Rd (low incidence).
Others AnWj NR NR R NE V Anti-AnWj is nonreactive or only
weakly reactive with In(Lu) red
cells.
a
At NR RR R NE NE All At(a–) red cells have been
found in Blacks.
Others (Continued) Bg RR RR R The Bg expression on the red cells
produced by an individual can
vary over time.
a
Cs NR NR R R V
Ge RR RR R W/D(Ge2) V(Ge2) Ge:-2,-3,-4 cells (Leach type) are
NE(Ge3) NE(Ge3) the null red cells.
W/D(Ge4) NE(Ge4)
JMH NR NR R D D Anti-JMH is often found as a
transient antibody in older
patients.
a
Jr OR OR R IR NE A higher frequency of Jr(a–) red
cells exists in the Japanese
population.
Lan RR RR R NE NE
a
LW OR RR R NE W/D LW(a–b–) cells are the null red
cells.
b
LW is a low-incidence antigen.
a
Anti-LW can look like anti-D.
a
Sd R OR R IR NE Agglutination can be refractile
under the microscope.
CAD+ red cells show strong

a
expression of Sd .
Vel R R R IR NE Anti-Vel is often hemolytic.
a
Xg R R W/D NE
6,19,20
Note: Antigens, antigen frequencies, and blood group systems/collections not mentioned on this table can be found elsewhere.
43
a
anti-K, -E, or even -Jk during immediate spin, should
be cause for further investigation.
Reactivity after the 37 C incubation (37 C spin) but
before the washing and adding of antihuman globulin
(AHG) is suggestive of a clinically significant antibody.
Current gel and solid-phase methods, as well as PEG
tube testing, do not incorporate this phase of testing. In
addition, many facilities have dropped this phase of
testing from other tube testing protocols. In prob-
lem-solving, positive reactions during the 37 C spin
phase can be very useful in identifying or sorting out
different patterns of reactivity, especially in those cases
of complex multiple antibodies. Several specificities
a bH
such as anti-E, -K, -Le , -Le , or -P1 can show very
clear-cut patterns during this phase. Inclusion of this
phase of testing in extended antibody investigations
should be considered.
Reactivity during the IAT phase of testing is usually
associated with a clinically significant antibody. Allo-
and autoantibodies that are clinically significant are al-
most always detectable during this phase. Thus, reac-
tivity at this phase needs to be investigated until the
antibody can be identified or sufficient data can be col-
lected to ensure the safe and efficacious choice of blood
components for the patient.
If an antibody is classified as clinically insignificant
(eg, cold antibody, reagent-dependent reactivity),
techniques known to avoid the reactivity should be
carefully followed: prewarm technique or adsorption,
and changing or eliminating enhancement media,
respectively. Such techniques will allow for the demon-
stration of any underlying clinically significant allo-
antibodies that may have been masked by the
reactivity of the cold-reactive antibody.
However, as always, there are exceptions to the rule.
Normally clinically insignificant antibodies, such as
anti-IH, can react during IAT and in rare instances can
26 k
be clinically significant. Similarly, anti-Vel, -PP1P , -H,
or an autoantibody of broad thermal range may be re-
active at all phases and with all panel cells, yet be
clinically significant.
Strength of Reactivity
The strength (as well as the pattern) of reactivity is an

important clue to the identity of the antibody. Strong

reactivity with some reagent red cells and clear-cut

negative reactions with others is the usual scenario as-
sociated with a single alloantibody. However, reactiv-
ity with all panel cells or weak or variable reactivity
with some or all cells causes the most difficulty. In these
cases, the blood banker often faces the dilemma of de-
ciding whether to try to desensitize the testing system
by eliminating the enhancement media (to eliminate re-
activity of a clinically insignificant antibody) or to try
to enhance the reactivity (to increase the sensitivity of
the system to detect and identify a weak, clinically sig-
nificant antibody). Much time is spent (and wasted)
pursuing clinically insignificant, reagent-dependent
reactivity or autoantibodies that would not be detected
in another media or in saline tube technique.
A very strongly reactive antibody is more likely to be
clinically significant; however, weakly reactive anti-
bodies cannot be dismissed as clinically insignificant.
Strongly hemolytic, warm autoantibodies are more of-
ten seen in patients with rapid hemolysis due to the
autoantibody. Most Rh antibodies, such as anti-D,
anti-E, and anti-e, tend to be strongly reactive. How-
ever, other clinically significant Rh antibodies tend to
be weakly reactive (as in many examples of anti-c).
a
Anti-K and anti-Fy tend to be strongly reactive,
b
whereas anti-Fy and Kidd blood group system anti-
bodies tend to be weakly reactive. Clinically insignifi-
cant antibodies, such as anti-Kn and anti-Yk, tend to be
weak and variable in reactivity. It has been the author’s
a
experience, however, that anti-Cs is often very strong
(2+ or greater). Other antibodies that tend to react
weakly include antibodies to Lutheran, Cartwright,
and Dombrock blood group antigens. Antibodies to
Colton blood group antigens, on the other hand, tend
to be quite strong.
Pattern of Reactivity
A pattern shown by an antibody panel in which most or

all cells are reactive except the autocontrol would be
consistent with an antibody to a high-incidence anti-
gen or a multiple-antibody mixture. If the strength of
reactivity is essentially all the same, a single antibody
to a high-incidence antigen is more likely. Strong reac-
tions with all cells except the autocontrol during imme-
diate spin through the addition of AHG, may indicate

k
the presence of anti-IH, anti-Vel, anti-PP1P , an auto-
antibody of broad thermal range, or a mix of IgM and
IgG autoantibodies. If the reactivity is variable, multi-
ple antibodies are the more likely cause. An antibody
reacting with only one reagent red cell or one donor
crossmatch is most likely an antibody to a low-inci-
dence antigen. Variable reactivity with most cells, ac-
companied by a 37 C spin or immediate-spin reactivity,
is indicative of anti-Lea with -LebH or anti-P1. Antibodies
that react with some or most of the red cells tested are
either single or multiple specificities.
While most specificities can be determined by inclu-
sion and/or rule-out of common specificities, antibod-
ies such as anti-Doa, anti-Dob, or anti-Ytb may not be
identifiable by examining the antigen profiles. Weak
and often variable reactivity with most or all of the red
cells (except the autocontrol) may indicate the presence
b a a
of anti-Ch, anti-Rg, anti-Lu , anti-Yt , or anti-Sd . Many
a a
other specificities, such as anti-Yk , anti-Kn , or
a
anti-McC , may also show this pattern of reactivity, and
antibodies in the Cromer, Lutheran, and Dombrock
blood group systems may all be weakly reactive. Reac-
tivity with all cells including the autocontrol (usually
the pattern for an autoantibody) can be indicative of re-
agent-dependent reactivity or a delayed transfusion re-
action.
An autocontrol can be essential in differentiating an
autoantibody from an alloantibody. Along the same
line, a DAT, in conjunction with an autocontrol, can
help differentiate an autoantibody or delayed transfu-
sion reaction from a reagent-dependent reactivity. LISS
or PEG panagglutinins will show a positive auto-
control but a negative DAT. An autoantibody or
delayed transfusion reaction will show a positive auto-
control and DAT; however, it is important to be aware
that a negative DAT and/or autocontrol do not abso-
lutely rule out the presence of an autoantibody. Tran-
sient antigen depression can occur with the concurrent
formation of the corresponding antibody in a patient
with autoimmune hemolytic anemia.
Variable Reactivity
Variable reactivity can be the result of an antibody

showing the dosage effect, multiple antibodies, or an

antigen showing variable expression from one panel

cell to another.
The dosage effect is shown when antibodies react
with cells from a person who is homozygous for a gene,
but the antibodies are nonreactive or much weaker in
reactivity with cells from a person who is genetically
heterozygous. It is assumed that the homozygous indi-
vidual has more of the derived antigens expressed on
the red cell surface—hence, the stronger reactivity seen
with the corresponding antibody. An antibody that
a a
typically shows the dosage effect is anti-Jk . Anti-Jk
might react 2+ with Jk(a+b–) red cells, but it would be
nonreactive with red cells that were Jk(a+b+). Antibod-
ies that commonly show the dosage effect correspond
a a
to the following antigens: c, E, K, k, M, N, S, Fy , Jk , and
b
Jk .
Certain antigens show a wide range of concentration
and number on red cells from different people. Anti-
gens that commonly show a wide variety of expression,
as well as this variable reactivity with their correspond-
a bH a
ing antibodies, include P1, Le , Le , Bg, and Sd . The P1
expression on red cells from some Black individuals is
much stronger than that from most White individuals.
The following situation is often a clue pointing to
anti-P 1 : A R 0 Fy(a–b–) P 1 + panel cell demonstrates
strong reactivity, while the rest of the panel cells show
weak variable reactions with some, but perhaps not all,
a
P1+ red cells. Sd expression can vary widely, and Cad+
a
red cells show very strong reactivity with anti-Sd .
Hemolysis and Agglutinins
Partial or total hemolysis of the reagent or donor red

cells can occur with antibodies to the following anti-
a b/bH
gens: A, B, H, Vel, Le , Le , and PP1Pk, and occasional
I, IH, and Kidd blood group antigens. Some autoanti-
bodies may also show hemolysis of the test cells. It has
been the author’s experience that some clinically insig-
nificant LISS-dependent panagglutinins can show
hemolysis. Hemolysis is always a result that should be
investigated and explained, since some of the specifi-
cities can cause brisk hemolysis in a patient who re-
ceives incompatible units.

Physical Appearance
The physical appearance of the reactivity often pres-

ents clues that make the process of antibody identifica-
tion easier. The “coin stack” appearance of rouleaux
when viewed under the microscope is easy to recog-
nize. Anti-Sda shows refractile and mixed-field type of
a
agglutinins when evaluated microscopically. Anti-Lu ,
a
anti-Xg , and some autoantibodies can also show a
mixed-field appearance. Additionally, some Lewis and
Lutheran antibodies can appear “stringy.”
Impact of Testing Methods
The pH of the media or additive and the temperature at

which the test is performed can have profound effects
27
on some antibodies. Many examples of anti-M have
been shown to react more strongly in an acidic environ-
ment. In some techniques, however, such as solid-
phase red cell adherence, the use of acidic or un-
28
buffered saline may cause antibodies to be missed.
Some autoantibodies that are pH-dependent may
cause ABO typing discrepancies when various mono-
29
clonal typing reagents are used.
The temperature at which an antibody reacts (also a
clue in the antibody identification process) is usually
indicative of whether an antibody is potentially clini-
cally significant. An antibody reactive only at room
temperature or lower temperature is probably not clini-
cally significant because it is not reactive at body
temperature. The exceptions are very potent cold
autoantibodies or antibodies in patients who will un-
dergo reduction in core temperature in the course of
treatment. The use of cold serum/plasma or reagents
or routine, room-temperature incubations can often
cause cold-reactive antibodies to bind to red cells, re-
sulting in a possible “carry-over” reactivity during 37
C or IAT. Care should be taken to avoid the use of cold
samples and reagents, as well as routine or inadvertent,
extended, room-temperature incubations.
The effect of chemicals and enzymes on target red
cell antigens can provide very useful information (see
Table 2-6). Many antigenic structures are destroyed
19,20,25,30-38
and some are enhanced. An early problem-solv-
ing step can be the chemical modification of red cells

Table 2-6. Effects of Chemicals on Antigen Expression
Antigens Depressed or
Chemical Weakened Antigen Expression Enhanced
a b a
Enzymes: papain and M, N, S (variable), Fy , Fy , Yt , Rh, Kidd, and Colton blood
a g
ficin Ch, Rg, Pr, Xg , Pr, Tn, M , group antigens; P1; Lewis
a a a
Mi/Vw, Cl , Je , Ny , JMH, antigens; and I/i
a a
En TS, En FS, Lu8, Fy6, Ge2,
b
Ge4, and In
0.2M dithiothreitol Kell blood group antigens Kx
(DTT) (except Kx), LW, Indian and
Scianna blood group
antigens, Kn, McC, Kn/Mc,
a
JMH, Yt , Gy, and Hy
Those mainly just weakened:
Cromer, Dombrock, and
Lutheran blood group
systems; AnWj; and MER2
a b
0.01M DTT Weakened: Js and Js None reported
ZZAP (combination of All the antigens listed next to None reported
enzyme and DTT) “Enzymes” and 0.2M DTT
Chloroquine Bg None reported
diphosphate Reports of other antigens being
weakened with extended
treatment, most notably Rh,
exist.
Hypochlorite S None reported
(“bleach”)
a a b
Aminoethyliso- Yt , Hy, Kn, Yk, McC, LW , Lu , None reported
thiouronium Kell blood group antigens
bromide (AET) (except Kx), Lu3, Lu4, Lu5,
Lu6, Lu8, Lu12, Lu13, and
Lu17
Those mainly just weakened:
Cromer blood group
systems, Gy, and Hy
EDTA/glycine/acid Kell blood group system None reported
(EGA) antigens, the Er collection
antigens, and Bg

using, for example, 0.2 M dithiothreitol (DTT) or en-

zymes. If a multiple-antibody problem arises and there
is a suspected antibody to a high- or low-frequency
antigen, enzyme treatment of reagent red cells is proba-
bly one of the most efficient tools for antibody identifi-
cation.
Reagent anti-IgG is derived from different sources
(rabbit, monoclonal). Some current monoclonal sources
of anti-IgG lack recognition of IgG4. Those occasional
antibodies that are pure IgG4 are generally clinically
insignificant and are usually specificities such as
anti-Kn, anti-Yk, anti-McC, or anti-JMH; some benign
autoantibodies are also primarily IgG4. A monoclonal
reagent, lacking the IgG4 detection, can be a useful tool
in avoiding this clinically insignificant reactivity.
AHGs used routinely in the blood bank do not con-
tain anti-IgA. In rare instances, IgA antibodies may
39
be involved in red cell destruction. Currently, many
of the polyspecific AHG reagents are made up of
monoclonal source material. In the polyclonal
sources that are not combinations of monoclonal
sources, some reactivity can be directed toward the
light chains (kappa or lambda) and these reagents
could potentially react with antibodies that are par-
tially or primarily IgM or IgA. If unexplained
hemolysis occurs in a patient, it may help to use AHG
that is IgM, IgA, and/or polyclonal in order to detect
the rare IgM and/or IgA antibodies. It is essential
that blood bankers be aware of the characteristics of
the antiglobulin reagent used in their facility so that
they can make informed choices when solving com-
plex antibody problems.
Antigen Characteristics
The number of antigen sites, along with the avidity and

titer of the corresponding antibody, is usually equated
with the strength of reactivity. This is certainly true of
the antibodies in the ABO blood group system. Discus-
sion of the number of antigen sites for different anti-
6,20,25
gens can be found in several sources. Very few
antibodies outside the ABO and Rh blood group sys-
tems show 4+ reactivity. Some antibodies, such as
anti-P1, demonstrate variable reactivity because of the
diverse number of antigens present on the different re-

agent red cells. Antibodies such as anti-Yk and anti-Ch

demonstrate weak reactivity due, in part, to fewer anti-
gens on the red cell surface.
An interesting area of study is the variant antigenic
structures. The discovery of apparent anti-D in a
D-positive patient has led to the classification of many
structural differences in the D antigen. Currently, many
molecular-level evaluations of these variants are lead-
ing to an understanding of the nature of the structural
variations that explain earlier serologic findings.
Monoclonal antibodies have also been used to further
define these variant structures.
Variant structures are seen in other Rh blood group
S B
antigens such as the e+, hr –, and/or hr – variants, or
the EW antigen. If an alloantibody is detected in a pa-
tient that apparently types positive for the correspond-
ing antigen, a variant antigen may be the cause.
Patients with variant structures lack certain epitopes
that are commonly present on the structure in question.
When the patient is exposed to the common epitope
that he or she lacks, an antibody is formed that recog-
nizes that common structural epitope. Additionally,
some antisera may not recognize variant structures.
For instance, some examples of anti-S do not recognize
the S antigen when it is accompanied by the low-fre-
quency TSEN antigen.40
Steric presentation of antigens may be important in
antibody identification. As more and more is known
about blood group antigens—including their structure
on the surface and in the red cell membrane, their pre-
sentation on the surface of cells, and their interaction
with neighboring structures—it has become clear that
their variations can have an effect on antibody recogni-
tion of antigens and the strength of reactivity. An anti-
body may not react—or the reaction may be much
weaker—with a red cell having a certain antigen that
seems to “block” or somehow prevent or reduce bind-
ing of the antibody to its target antigen.
The interaction between the antibody and the anti-
gen is often affected by the antigen itself and the
epitope that the antibody recognizes. Often that
epitope may be a combination of structures such as the
combination recognized by anti-A,B, -IH, or -IB. The in-
teraction can result in unexpected reactivity, such as
anti-D in a D-positive patient, or an unexpected lack of
reactivity, as in the case of anti-S that does not react
with S+TSEN+ red cells. This diversity is a source of

never-ending challenges in the art and science of anti-

body identification.
Tools Available for Identification
Rule-Out
“Cross-out” or “rule-out” is a process by which the

blood banker looks at panel cells that are nonreactive
and eliminates possible specificities that correspond to
antigens present on the nonreactive cells. This is a tool
in the process and is not infallible. It is important for
blood bankers to understand that rule-out is a process
that helps to select specificities with a low probability
of being present in the serum tested. Further evidence
should be collected to confirm the solution. Other tech-
niques and reagents outlined in this chapter also play a
part in the process of broader antibody identification
and should be used in concert with this tool. That aside,
rule-out is a major player in determining probable anti-
body specificities.
The broad array of institutional variability in rule-
out protocols must be taken into account when results
from one institution are compared with those from
another. For example, some blood bankers rule out a
specificity based on one nonreactive cell without con-
sideration of zygosity. Others go to the other extreme,
requiring three cells, two of which must be a dou-
ble-dose (homozygous) expression. Most laboratories
fall somewhere in between; procedures are established
at each facility. Following the conservative, yet practi-
cal rule-out protocols suggested below can increase the
likelihood of reaching the correct solution, without
“breaking the bank” by performing unnecessary
testing.
Rule 1. It is always preferable to rule out an antibody
specificity on a cell with a double-dose expression, es-
pecially a specificity that can be affected by dosage.
a
This is particularly true when ruling out anti-Jk or
b
anti-Jk . The blood banker should be cautioned, how-
ever, that there are silent alleles in some blood group
systems and an apparent double-dose expression
might just be a single-dose expression. For example,
red cells from a Black donor whose phenotype may be
Ro, Le(a–b–), most likely will not have a double dose of
a
the Fy antigen, even though the red cells type

Fy(a+b–). A Fy(a+b–) cell in this population is most

likely FyaFy genetically. The silent Fy gene makes the
antigen a single-dose expression instead of a dou-
ble-dose expression. Any problems this causes may be
overcome, to some extent, by following Rule 2 below.
Rule 2. It is better to rule out specificities with two
unique cells rather than one. Variant antigen struc-
tures, lack of antigen recognition by the antibody,
weakened antigen expression, unusually rapid antigen
degradation, and the possibility for technical error
make the two-cell rule-out preferable because it allows
for an internal control in the process itself.
However, exceptions can be made. If, for instance, a
patient’s anti-D is strong, the rule-out of anti-C on one
r′r (D–C+E–c+e+) cell and of anti-E on one r′′r
(D–C–E+c+e+) cell is probably sufficient. A patient
who forms strong anti-D usually demonstrates equally
strong anti-C and/or anti-E. Because almost all D-neg-
ative cells are negative for both C and E (rr), not identi-
fying weak anti-E or -C in a patient with an anti-D is
unlikely to result in the infusion of incompatible blood.
In this case, the risk associated with using one cell for
rule-out is reduced. Another exception to this rule
could be that clinically insignificant specificities, such
bH
as anti-Le , may be ruled out with one cell, because the
risk of not detecting the antibody has little or no chance
of causing harm to the recipient.
There is no reason to routinely rule out antibodies to
low-incidence antigens, and even screening cells rou-
tinely may not detect these antibodies in the first
41
place. However, when a normally low-frequency anti-
gen is of higher frequency in certain populations and
the antibody has potential for harm, the normally
low-frequency antigen may then need to be part of the
rule-out. For example, the Dib antigen is rare in Blacks
and Whites, but it is found in 10% of the Asian popula-
tion and 36% of the South American Indian popula-
tion.20(p301) In these cases, Dib has a higher frequency than
the K antigen, and anti-Dib is a potentially clinically sig-
nificant antibody; thus, it makes sense to routinely try
to rule out these specificities when dealing with such
populations. Because of the complexity of this process
and the number of variables, each institution should
standardize its protocol and balance efficiency with
safety.

Rule 3. In those instances when an antibody shows

extreme variability or reacts in more than one phase,
a b
the rule-out of anti-P1, anti-Le , and anti-Le might be
best carried out by a “reverse of rule-out.” In this case,
the question should be asked: Are all the P1– cells or all
the Le(a–b–) cells nonreactive? An affirmative answer
suggests that one of these specificities may be present.
It is important to note that this should not be a rou-
tine procedure for rule-out, and that the data collected
in this way can be error-prone and should be used only
in the context of a set of converging data consistent
with the final solution.
Rule 4 (“The Golden Rule”). As mentioned previ-
ously, rule-out is a tool. It does not, by itself, define
which antibodies are or are not present. Even if a speci-
ficity is ruled out by the laboratory’s standard practice,
it does not mean the antibody is not present. An anti-
body may be detectable through other media or in
other phases, or it may have been falsely ruled out be-
cause of technical error or lack of epitope recognition.
Whenever dealing with a patient demonstrating clini-
cal-evidence hemolysis with no explainable cause
(based on initial serologic findings), the blood banker
should not forget to employ the other tools in the tool-
box.
A general rule relating to all antibody problem-solv-
ing is that converging data should be collected until
there is a high probability that the final conclusion is
correct. Inconsistent data should be investigated in or-
der to explain why the pattern is not typical of the anti-
body hypothesized to be present. Selected protocols
should be such that they minimize human error, while
at the same time maximizing efficiency.
Confidence in the Validity of an Antibody Solution

19(p399)
The p value is a calculation of the number of anti-
gen-positive cells that react and the number of anti-
gen-negative cells that do not react. In the past, most
people considered the p value necessary to prove the
presence of an antibody. In reality, the p value was orig-
inally used when antibodies were newly discovered in
order to determine their probable specificity. Because it
is now known which cells are positive and which are
negative (ie, E+ and E–), the p value is not really appli-
42
cable or of any value.

3
As mentioned previously, AABB IRL Standards re-
quire that laboratories find two antigen-positive cells
that are reactive and two antigen-negative cells that are
nonreactive before concluding that an antibody is pres-
ent. This is a reasonable approach.
In addition, the overall picture of the antibody inves-
tigation plays an important part in decision-making.
For example, if an antibody is suspected to be anti-Jsb
but the facility has only one Js(b–) cell to test, the labo-
ratory could test 0.2M DTT-treated red cells (with Kell
blood group system antigens destroyed) to see if they
are nonreactive. If the treated cells are nonreactive, the
cells could be used along with the one Js(b–) red cell
sample to rule out additional specificities. Typing the
b
patient’s red cells for common antigens when anti-Js is
suspected could help further determine the antibody
specificity, along with which additional antibodies need
to be ruled out. If, in addition, the patient is Black and
the antibody is IgG, converging evidence indicates that
b
the antibody is anti-Js . If antibodies to common blood
group antigens cannot be ruled out, it would be neces-
sary to consult a reference laboratory.
In the event of an emergency, when additional
specificities cannot be ruled out (yet could be formed
based on the patient’s phenotype), a crossmatch of
compatible units negative for the antigens not ruled
out should be performed. For example, a patient has
b a
anti-Js and -K1; anti-Jk cannot be ruled out, and the
b
patient is typed as Js –, K–, and Jk(a+b–), because the
a
patient could make anti-K1 (but not anti-Jk ). Blood se-
lected for an emergency transfusion in this case should
b
be negative for Js and K1 and crossmatch compatible.
Chemicals
Different chemicals can be used in the problem-solving

process. (Chemical effects are listed in Table 2-6.) When
chemicals are being used, care should be taken to un-
derstand their effect and to use them appropriately.
Appropriate controls should be used with all reagent/
chemical testing.
Enzymes, DTT, EDTA/glycine/acid (EGA), and
chloroquine-diphosphate have multiple uses (see Table
2-7). Some less frequently used chemicals, 2-mercapto-
ethanol (2-ME) and 2-aminoethylisothiouronium bro-
mide (AET), have effects similar to DTT. By choosing

Table 2-7. Uses of Various Chemicals

Chemical Uses
Enzymes (papain • Enhancing warm or cold adsorptions

or ficin) • Destroying or enhancing antigens to
aid antibody resolution
• As a component of ZZAP for warm
adsorptions
Dithiothreitol • Destroying antigens to aid in
(DTT) antibody resolution (0.2M)
• As a component of ZZAP for warm
adsorptions (0.2M)
• Destroying IgM antibodies in the
serum/plasma (0.01M)
• Treating IgM-coated cells to remove
the IgM as an aid in red cell typing
(0.01M)
EDTA/glycine/acid • Destroying antigens to aid in
(EGA) antibody resolution
• Removing IgG from red cell surfaces
for phenotyping and for adsorptions
Chloroquine- • Removing IgG from red cell surfaces
diphosphate for phenotyping and for adsorptions
• Weakening or destroying Bg
antigens to aid in antibody
identification
one or two of these chemicals, most laboratories can in-

crease their ability to perform more complex antibody
resolutions.
However, it is important to become well informed
about the effects of the chemicals as well as their pit-
falls. Treatment with 0.2M DTT or EGA can make cells
somewhat “sticky,” and microscopic reading should be
avoided. Over-treatment of red cells with enzymes can
cause the red cells to spontaneously agglutinate or
hemolyze. Treatment with chloroquine-diphosphate
may weaken blood group antigens, so very potent-typ-
ing antisera should be used when typing the treated
cells.
Many of the methods for use of these chemicals are
given in the AABB Technical Manual.19(pp731-70) In general,
it is better to train staff well in the use of one or two

chemicals than to have an extensive battery of chemi-

cals on hand.
Enhancement Media
The selection and use of enhancement media (poten-

tiators) has always evoked debate among serologists.
Many articles, lectures, and Web-based conversations
have covered the pros and cons of the various available
media. Everyone has his or her favorite medium, but
none of them will detect all of the clinically significant
antibodies and avoid all the clinically insignificant
antibodies; none are without their fallibilities.
As the medium’s sensitivity is increased (to detect
weaker antibodies), the test system loses specificity (ie,
reaction with clinically insignificant entities). Whether
it is beneficial either to increase or to decrease sensitiv-
ity of the test system is dependent upon the nature of
the antibody present.
Again, protocols should be developed that provide
guidelines for which media to use and in what circum-
stances to use them. Guidelines should assist the blood
banker in the selection and interpretation of nonrou-
tine tests used in complex antibody identification. As
with chemicals, it is much more advantageous to have
staff well trained in the use of one or two media than to
have a multitude of available media or the “newest and
hottest” technique.
Enhancement media include albumin, LISS, and
PEG. Polybrene, once commercially available, is used
by a few immunohematology reference laboratories
(IRLs) but its use is very limited. Gel methods operate
in a LISS-based environment. Enzyme treatment of red
cells enhances reactivity of some antibodies but is con-
sidered a chemical treatment rather than an enhance-
ment medium.
Whatever medium is used, it is important for the
blood banker to know the advantages and disadvan-
tages of the reagent. The examples given below are not
at all an inclusive list:
• Some examples of anti-K do not react well in LISS.
• Albumin may greatly enhance Rh antibodies and
anti-P1 during the 37 C spin phase.
• A newly forming, primarily IgM antibody in a primary
response may not react well in PEG.

• Clinically benign autoantibodies are more likely to be

detected in PEG (or with enzyme-treated red cells).
• PEG (or enzyme-treated red cells) can greatly enhance
weak Kidd blood group antibody reactivity.
• Approximately 25% of normal donors have antibodies
43
directed toward PEG.
• PEG may cause problems in patients with hypergamma-
globulinemia resulting from red cell “entrapment.”44
• Gel methods can enhance antibodies, such as anti-Ch,
that are not clinically significant.
• Gel and solid-phase methods can increase detection of
clinically benign autoantibodies.
• Gel and solid-phase methods have more stable end-
points than the traditional tube methods.
• Even though gel is an IgG test, IgM cold antibodies,
such as anti-IH, can agglutinate cells directly and cause
a strong positive reaction. The reaction occurs as the se-
rum/plasma are mixed and these agglutinates do not
travel through the gel.
The bottom line is that every enhancement medium
can have its problems. In addition to the enhancement
medium itself, some chemical additives may be in the
medium that can cause their own set of problems. Sub-
stances such as thimerosol in the medium can cause re-
activity with the patient’s serum/plasma.
Garratty has published an excellent review45 of these
various “other” additives and their problems. Prob-
lems related to enhancement media will result in posi-
tive reactions in all tubes where the media are present
(eg, all cells on a panel and the autocontrol) and nega-
tive reactions in test environments where the enhance-
ment solution is not present (eg, the DAT). In these
cases, eliminating or changing the enhancement me-
dium may resolve the problem. Laboratories often per-
form unnecessary testing, such as adsorptions, and
delay transfusion while investigating reactivity caused
by enhancement media.
The saline technique (with no additives or enhance-
ment media) is generally underused by laboratories in
antibody problem-solving. Many limit the use of saline
to the prewarming technique. When room temperature
saline (as opposed to 37 C saline) is used, the technique
is sensitive enough to detect most red cell antibodies.
The reactivity may be more fragile or a bit weaker, but
most antibodies will react adequately in the saline tube

technique such that trained blood bankers can detect

them. When problems with additives or enhancement
media are suspected, saline testing is a useful, inexpen-
sive, and technically simple option.
It is important to keep in mind that the manufac-
turer’s circular is an excellent source of information re-
garding the characteristics of any product.
Panels / Selected Panels
Each commercial company has a selection of reagent

cells in a panel that detects most common and some
uncommon red cell antibodies. When encountering a
multiple-antibody sample, the blood banker must of-
ten select panel cells to rule out and rule in antibodies.
Select panels are just that—select; cells are chosen
based on the original reactivity and initial hypothesis.
Too many blood bankers will continue to run addi-
tional full panels or test cells that do not provide addi-
tional useful data. Often only one or two additional
cells are sufficient to provide information that supports
elimination of possible antibodies. As always, care
should be taken not to perform unnecessary testing.
Currently, computer programs are available that aid in
performing selected panels; panel sheets may be
printed with the selected cell information and an area
to record the test results.
Another essential but frequently overlooked tool is
the special/additional types provided on individual
panel cells. The additional typings from the panels are
useful for the identification of less common specifi-
b b a
cities, such as anti-Yt , anti-Bg, anti-Co , anti-Do , and
b
anti-Do . It is helpful, as well, to look at the actual indi-
vidual panel donor cell code. An R1R1-112 donor on one
panel is the same donor as the same R1R1-112 found on a
later panel. These are not two different panel red cells
but red cells from the same donor source. Testing these
two cells with a given serum sample will not meet the
intention of the 2 + 2 rule discussed earlier.
When encountering a weak panagglutinin with a
negative DAT, washing the reagent red cell may be a
prudent step to take. Patients often have antibodies to
antibiotics (such as neomycin), or sugars that are
45
added to the reagent red cell suspension medium.
When blood bankers take the panel cells or the selected
cell directly from the vial, there is a good chance they

will encounter reactivity with some patient samples be-

cause of the antibiotics or sugars. Washing the reagent
red cells will eliminate this reactivity.
It is common practice for laboratories to retain old
panel cells for selected cell panels. These outdated pan-
els can be essential for complex problem-solving. Much
debate has focused on the appropriate use and quality
control of these expired reagents. Typing these cells
a
with potent, commercial antisera (eg, anti-Fy ) for the
target antigen does not indicate adequate antigen
strength. A weakened antigen may react well with the
potent antisera but may not react with the weaker pa-
tient antibody. A more prudent and reasonable ap-
proach is to follow Rule 2 under the rule-out section.
The chance of two separate cells showing equally accel-
46,47
erated deterioration is slim. Recent studies have
shown that antigen strength on the expired cells re-
mains relatively stable; however, panel cells should not
be kept indefinitely. A reasonable rule is to use panels
only two months postexpiration and only when
unexpired panel cells are not available.
Neutralizing or Inhibitory Substances
Neutralizing or inhibitory substances are readily avail-

able. The two more useful substances, P1 and Lewis, are
available commercially. Fresh plasma or serum may
also be used to inhibit anti-Ch or anti-Rg. These simple
tools are often overlooked in complex problem-solv-
ing. If a patient has a complex multiple-antibody mix-
ture that possibly includes anti-P 1 , it is easier and
quicker to neutralize the anti-P1 than to try to put to-
gether a P1-negative selected cell panel.
Lewis substance can be used to neutralize reactivity
resulting from antibodies in the Lewis blood group sys-
tem when performing prenatal screens. If the original
panel is positive and suggestive of Lewis specificity, it
can be repeated with a Lewis neutralized sample. If the
neutralized sample is nonreactive, Lewis antibodies
are confirmed and no further testing is needed. If it is
positive, further investigation is needed to determine
the specificity of the antibody. Lewis antibodies will
not harm the neonate [neonates are all Le (a–b–)], and
they are rarely clinically significant in transfusion. In
addition, the antibodies are often transiently formed in
the prenatal patient.

Laboratories without large inventories of rare red

cells or special reagents may perform simple steps to
resolve antibody identification if the group of antibodies
formerly called “high-titered, low-avidity” antibodies
19(pp702-3)
is suspected. A serum/plasma neutralization us-
ing a pool of fresh antibody-free serum/plasma can be
used to neutralize anti-Ch or anti-Rg. The neutralized
sample can be used to rule out underlying antibodies.
Neutralization procedures carry a very slight risk of
missing extremely weak alloantibodies because of dilu-
tion by inhibition or neutralizing substances; therefore,
controls are important when these procedures are per-
formed. A dilution control should always be performed
in parallel. This usually involves adding the neutraliz-
ing or inhibiting substance to the patient serum and, in
another tube, adding an equal amount of saline or albu-
min to the patient serum to serve as a control. Thus, if
the reactivity remains in the control tube but is gone
from the test serum, the reactivity was not diluted to
the point that it was nonreactive. If the control and the
test sample are both negative, it is not known whether
the lack of reactivity is due to dilution or neutraliza-
tion/inhibition or both. If the test and control tubes
both remain reactive, then neutralization was not ac-
complished and additional investigation is needed.
The possible causes and solutions are listed below:
• The antibody was not directed toward the expected an-
tigen (eg, P1 or Lewis).
• The neutralizing substance may not be working. It
should be tested using an antibody known to be di-
rected toward the target antigen to ensure it is
working.
• The neutralizing/inhibitory substance is reacting with
the test cells. (This is more common when human sera
is the substance.) The substance should be tested di-
rectly against the test cell without the patient’s sample.
Phenotyping Reagents and Lectins
Phenotyping reagents and some lectins are tools for

collecting converging data and increasing the probabil-
ity that the antibody solution is correct. These reagents
are also useful in the identification of complex multiple
antibodies or antibodies to high-incidence antigens,

because they allow the blood banker to predict the anti-

bodies the patient can form.
However, both human and monoclonal sources of
antisera are occasionally associated with serologic
problems. Human sources may have antibodies to low-
incidence antigens that have the potential to cause
false-positive antigen-typing results.
In one case the author is aware of, a patient had
b
formed anti-Jk , but typed Jk(b+). The patient had a
negative DAT and autocontrol. The laboratory contin-
ued to test the cells with a total of four commercial
sources and several strongly reactive patient and do-
nor sources. Three of the four commercial sources
typed the cells as Jk(b+); one commercial source and
all the single-source antisera were nonreactive. It was
determined through discussion with the commercial
companies that three of the commercial sources used
one of the same donors in their pool of anti-Jk b. At that
time, the specificity of the low-incidence antigen
“contaminating” the antisera could not be identified.
Adsorbing the reactivity antisera onto the patient’s
cells and performing the eluate recovered the anti-
body to the low-incidence antigen that was not
anti-Jkb. This example illustrates the possible error
that can result from ruling out an antibody specificity
using the phenotype alone, and again emphasizes the
importance of collecting converging evidence and
considering all data when interpreting antibody iden-
tification tests.
As discussed earlier, antisera may also lack epitope
recognition. For example, not all anti-S reagents react
with S+TSEN+ cells. These variations are not isolated
to human (polyclonal) sources of antisera but may
occur with monoclonal sources as well. Monoclonal
sources may even recognize structures not normally
detected by polyclonal sources.
In one case, the anti-A produced by one clone re-
sulted in reactivity with B(A)-phenotype individuals,
causing some mistyping until the problem was identi-
fied and the clone was diluted or removed from the an-
tisera.
Typing reagents may also give false-positive or false-
negative results with red cells heavily coated with im-
munoglobulins. This can occur even when the antisera
48
are monoclonal and not being read at the IAT phase.
Lectins, while not literally phenotyping reagents,
have many uses in blood bank problem-solving.

Anti-A1 (Dolichos biflorus) is an essential tool in the in-

vestigation of an ABO discrepancy between the red cell
and the serum/plasma grouping. If red cells type posi-
tive with anti-A but negative with anti-A1 lectin, the
patient is an A2 (or other A subgroup) and could poten-
tially form anti-A1.
Soybean lectin (Glycine soya) is useful in determining
decreased sialic acid content on the red cell surface. Pa-
tients or donors having missing or altered glycophorin
A or B (GPA/GPB) may react with this lectin. Red cells
that are U– or En(a–), as well as those cells that are Tn,
T, or Cad polyagglutinable, may react with this soy-
bean lectin. Lectins are useful when investigating a do-
nor unit that is incompatible with most or all patient
samples and the unit has a negative DAT. A quick
screen with soybean lectin will determine if the unit
demonstrates one of the more common forms of poly-
agglutination. Other lectins will help classify the type
of polyagglutination present.
Elution Techniques
In the past, many articles praised one elution technique

over another. Currently, with the availability of com-
mercial eluate kits, and because many of the other
elution techniques require toxic and/or carcinogenic
agents, this controversy is no longer active.
Most laboratories perform eluates for ABO antibod-
ies using the Lui freeze or 56 C heat techniques. The
commercial kits are then used for IgG antibody recov-
ery. For the most part, only advanced IRLs use organic
solvents (dichloromethane, ether, or xylene) for an
eluate, when needed. Both the acid-based techniques
(on which all commercial kits are based) and the or-
ganic solvent techniques may miss some specificities,
but for the most part they are very successful in recov-
ering IgG antibodies.
Elutions provide useful serologic evidence but are
warranted only in certain clinical and serologic situa-
tions. An elution may be needed when a patient has a
positive DAT or, occasionally, when investigating a pa-
tient who has a negative DAT and appears to have au-
toimmune hemolytic anemia. In the author’s opinion,
there are really few reasons to perform an eluate.
One reason is to gather further evidence when a pa-
tient who has been transfused within the last 2 to 3

weeks presents with a positive DAT and/or unex-

plained hemolysis. Many institutions specify a longer
time (up to 3 months) to determine when an elution
49,50
should be performed. However, data and the au-
thor’s experience suggest that the longer time span is
unnecessary and unlikely to improve patient outcome.
It may be that even the 2-week window is not neces-
sary. Many of the more sensitive methods (PEG, solid
phase, and gel) make it easier to detect antibodies at
lower levels.
Another reason to perform an elution might be to as-
sist in the investigation of unexplained hemolysis, such
as in a suspected drug-induced anemia. Elutions per-
formed on red cells from patients with known auto-
antibodies have very little benefit in patient care and
do not alter transfusion recommendations. If a patient
has a classic warm autoantibody in the serum, it will
most likely be recovered in the eluate. Differential
adsorptions performed on these eluates are, in the au-
thor’s opinion, warranted only in rare cases when the
patient seems to be undergoing a hemolytic transfu-
sion reaction, even though the patient’s serum or ad-
sorbed serum demonstrates no alloantibodies.
In many institutions, an eluate is performed for all
cases of suspected HDFN, but the need for this should
be questioned. The antibodies that could be causing the
HDFN are in the mother’s sample; if there are multiple
antibodies, it matters little which of them is/are the
causative specificity. The treatment in any case would
be to transfuse blood compatible with all the mother’s
antibodies, not just with the causative antibody. One
could argue that some specificities are more likely to
cause a stormy course for the neonate, but this informa-
tion is minimally useful because the treatment is based
on the physiologic symptoms and outcomes in the neo-
nate, not on the causative antibody.
Many eluates are performed on neonates or cord
blood samples of group A or B neonates when the
mother is group O and has a negative antibody screen.
In this case, when the DAT is positive, the eluate is per-
formed; however, more than 99% of the time, the recov-
ered antibody will be anti-A, -B, and/or -A,B. This
eluate does not improve patient care because, if AABB
2(p42)
Standards are followed, the neonate will receive
group O cells either way.

Adsorption Techniques
Adsorption techniques fall into two basic categories:

autoadsorption and alloadsorption. The most common
use of adsorptions is to remove autoantibodies from
the serum/plasma to allow the detection of underlying
potentially clinically significant alloantibodies masked
by the autoantibody. When dealing with a cold auto-
antibody, this procedure can allow the laboratory to
perform a serum/plasma (reverse) ABO grouping
without the interference of the cold-reacting autoanti-
body.
Autoadsorption is preferable to alloadsorption, be-
cause it removes only the autoantibody and subse-
quently allows for the detection of any alloantibody the
patient may have formed.
The red cells for either type of adsorption are usually
treated to remove bound IgG in order to free up the
sites for the adsorption and/or to enhance antigens in
order to increase autoantibody uptake. These treat-
ment methods include the use of ZZAP, enzyme,
glycine-HCl, chloroquine diphosphate, and/or com-
mercial kits. A more recent adsorption procedure uses
PEG.19(pp720-1)
The key to using auto- or alloadsorptions is to know
the risks involved. Alloadsorption always carries a small
but significant risk of removing an antibody to a high-
incidence antigen that the patient has formed. Even a
correctly performed autoadsorption, however, will
have a slight dilution factor for each adsorption per-
formed, because of intercellular trapped saline.
Alloadsorptions are usually performed on a pheno-
typically similar cell (if the phenotype is known), or on
a battery of two or three red cell samples of known phe-
notype. The latter, often termed a triple or differential
adsorption, can be a complex puzzle to solve. Rule-outs
need to take into account not only the positive and neg-
ative postadsorption reactions, but also the phenotype
of the adsorbing red cells. The blood banker needs to
know what would have been removed and what would
have been left behind, based on the phenotype of the
adsorbing cells.
In addition, many adsorption procedures use ZZAP
or enzyme, and the blood banker needs to take into ac-
count what antigens could have been destroyed on the
adsorbing red cells. This can add to the complexity of

the task, so these procedures require concentration and

attention to detail. Such complex problems are apt to be
associated with human error.
Adsorptions may also be used to identify complex
multiple antibodies and/or to rule out underlying
specificities in the presence of an antibody to a high-
incidence antigen. If, for example, a patient has
anti-Coa and the available Co(a–) cells will not allow for
rule-out of anti-Jk a , an alloadsorption may be per-
formed. The cells used for adsorption would be Co(a+),
Jk(a–) so that they would adsorb the anti-Coa but not
anti-Jka. After adsorption, the presence of the anti-Jka
could be ruled out or confirmed in the adsorbed se-
rum/plasma.
Several things need to be kept in mind when per-
forming this type of adsorption procedure. First, the
phenotype of the adsorbing cells should be selected
carefully in order to leave the appropriate specificities
behind. Second, it is better to adsorb the antibodies to
antigens of higher incidence in order to remove this re-
activity and allow more freedom in the choice of red
cells for the postadsorption testing. Eluates may be per-
formed from the adsorbing red cells to provide further
evidence to support a case solution.
There are many reasons adsorptions might not be
successful in removing antibodies. Treatment of the ad-
sorbing cells with enzymes and/or 0.2M DTT destroys
many antigens (see Table 2-5). If the allo- or auto-
adsorption is directed toward any one of these de-
stroyed antigens, the adsorption will not remove the
antibody. For example, if 0.2M DTT-treated autologous
or allogeneic cells are used, they would not remove any
Kell blood group autoantibody or antibodies to high-
incidence antigens in the same system. In addition, if
the serum contains one of the specificities formerly
called “high-titered, low avidity antibodies”—such as
anti-JMH, anti-Kn a, or anti-Csa—the antibodies will
probably not adsorb from the serum/plasma very
well, because of the high titers and possible low bind-
ing affinity and/or excess antibodies to antigen-bind-
ing sites.
Most autoantibodies are of relatively low titer, even
when strongly reactive. Titers are commonly eight or
less. If an apparent autoantibody is not adsorbed or sig-
nificantly reduced after three adsorption procedures,
the blood banker should consider the possibilities men-

tioned above to make sure she/he is not “on the wrong

track.”
Testing of the ZZAP- or enzyme-treated autocontrol
against the raw or neat serum can show whether the
problem is a result of antigen destruction. Testing a
chloroquine- or EGA-treated autocontrol or a reticulo-
cyte-enriched fraction may tell the blood banker
whether the antibody is an auto- or alloantibody as
well.
Questions that should be asked when encountering
an apparent warm autoantibody include:
• Are the positive DAT and the serum/plasma antibody
unrelated?
• Is the reactivity autoantibody- or alloantibody-related,
or a combination?
After asking these questions, the blood banker will
need to support or refute the original hypothesis that
the antibody is an autoantibody. Other tools, such as
phenotyping, chemical treatments, rare or null cells,
and titrations, could be used to help support or refute
the hypothesis as well.
Special Reagents
Special reagents such as human platelet concentrate

(which removes anti-Bg from the serum or plasma) and
rabbit red cell stroma (which removes cold agglutinins
from the serum or plasma) are commercially available.
The reagents are made to help simplify the antibody
identification process and their use varies throughout
the United States. The rabbit stroma can adsorb other
antibodies such as anti-B and anti-Vel and should be
used with caution. The manufacturer ’s instructions
should be consulted to determine limitations of the re-
agents. Protocols in the laboratory should define when
it is appropriate to use these products.
Rare Red Cell Inventory
Cells of unusual, helpful, or rare phenotypes can come

from many sources. They can come from commercial
panels, red cells from patients and donors, or even cells
from employees. The international Serum, Cells, and
Rare Fluids Exchange and other trades between labora-
tories can supply a broader and more esoteric range of

rare cells and antisera for problem-solving. Also,

laboratories can freeze small aliquots of red cells
and antisera for use in problem-solving. Most refer-
ence laboratories use liquid nitrogen procedures;
however, some laboratories do use conventional
freezers for this process. The procedures for provid-
ing protection for the freezing cells is different for
each method.
A laboratory’s approach to a problem is often deter-
mined by its resources. For example, to resolve a
serum/plasma ABO grouping problem and rule out
underlying antibodies in a patient with strong anti-I, a
laboratory could perform a prewarming technique, a
cold autoadsorption, or an adsorption with rabbit
stroma, or a laboratory with many rare cells could per-
form a select I– red cell panel.
Among the more useful red cells to freeze or main-
tain for future use are cells that can rule out or prove
specificities in cases of complex multiple antibodies.
Many blood bankers think that only cells negative for
high-incidence antigens need to be frozen, but multi-
ple-antigen-negative red cells are probably more useful
on a day-to-day basis. For example, an R2R2 cell that is
Fy(a+b–), Jk(a+b–), K+, M–N+S+s–, and P1– would be
very useful for a patient with anti-e, -Jkb, -Fyb, and -P1,
who could also make anti-S and anti-K. However, these
“ideal” cells for a particular patient may not be on
in-use commercial panels or in-house panels.
Another often overlooked group of red cells that
may be useful if frozen are those used to resolve se-
rum/plasma grouping in patients with a strong imme-
diate-spin antibody. A1 and B cells that are Vel– are very
helpful for a patient with anti-Vel reacting during im-
mediate spin.
Null cells, by definition, lack multiple antigens in
one blood group system. Cells such as Rh null ; hh;
Co(a–b–); Fy(a–b–); Gy– (the null cell in the Dombrock
blood group system); K o ; Sc:-1,-2; Jk(a–b–); and
Lu(a–b–)/In(Lu) cells are very useful in resolving diffi-
cult serologic problems. These cells can quickly direct
the blood banker to the blood group system involved if
one of these cells is nonreactive. An “artificial” null cell
in the Kell blood group system can be made by chemi-
cally treating a cell with 0.2M DTT, AET, or EGA. This is
the preferred screen for this blood group instead of the
rare, genetic Ko.

It is important not to “squander” rare cells by identi-

fying autoantibody specificities such as anti-dl and
anti-pdl in the Rh blood group system. It is also essen-
tial to know that null cells may lack other non-
system-related antigens, such as in the case of In(Lu)
individuals, whose red cells have not only depressed
Lutheran antigens (including Aua and Aub) but also de-
pressed AnWj, P1, I, and CR1-related (Kn, McC, Yk) an-
tigens. Similarly, Rh null cells lack LW and Duclos
antigens and express a depressed U antigen.
Other Technical Procedures
The selection and use of many procedures continues to

generate controversy among serologists. The pre-
warming technique is a prime example of how opin-
ions differ.51,52 [One prewarming procedure may be
19(pp754-5)
found in the AABB Technical Manual. ] A recent
53
study shows that prewarming can decrease or elimi-
nate clinically significant reactivity, and that enhance-
ment media may not improve the ability to detect
antibodies when using the prewarming technique.
Prewarming is useful to determine the potential clini-
cal significance of traditionally cold alloantibodies,
such as anti-M, -P1, and -A1. It is a useful tool, also,
when it is clear a laboratory is dealing with a cold anti-
body, and no other protocol will allow for serologic res-
olution of pretransfusion and compatibility problems
resulting from the cold antibody.
A n o t h e r proc e d u re , t h e D o n a t h - L a n d s t e i n e r
19(pp784-5)
test, is a relatively simple test, useful in the detec-
tion of a biphasic IgG hemolysin that is seen in paroxys-
mal cold hemoglobinuria (PCH). PCH should be
suspected when a young child with a recent viral infec-
tion presents with hemolytic anemia.
Titration, as well, may prove useful in antibody
problem resolutions. Titrations performed on an anti-
body can provide a clue to whether the antibody is
likely to be one that exhibits high titer and low avidity.
These antibodies may show relatively weak reactivity
(1+) but will show a quite high titer (greater than 64).
Titrations can demonstrate whether an antibody is
more likely to be an autoantibody (low titer) than an
alloantibody (titers medium-high to high). In some
emergency cases, a dilution of an autoantibody in order

to evaluate for underlying alloantibodies is based on

54
the generally lower titer of autoantibodies.
Other methods used in the blood bank setting in-
clude reticulocyte separation, monocyte-monolayer
testing, and molecular genotyping. When selecting any
method used in the laboratory, usability, validity, and
process controls should be considered. Blood bankers
should question whether the results give useful infor-
mation that can be tied to diagnostic information (such
as thermal amplitude studies) or selection of safer com-
ponents, or that otherwise improves patient outcome.
The usefulness of special techniques and procedures
depends on competent and well-trained blood bank-
ers, clear directions, understanding of the results
achieved, and the ability to use the information gained
to improve patient care.
Conclusion of the Process
All the components of antibody identification listed in

Table 2-2 and incorporated in the foregoing discussion
are part of the process. The gathering of information,
testing of samples, and interpretation of test results all
lead to selection of the appropriate blood component.
The process may include screening the blood for the ap-
propriate antigen (eg, K– or S–), using the crossmatch
alone to determine acceptability (eg, Lewis blood
group system antibodies), or choosing units that are in-
a
compatible (eg, anti-Yk or a warm autoantibody).
The goal is to provide the best component to improve
patient outcome and not to cause harm, but the process
to achieve this goal needs to be completed accurately
and efficiently. Too often the blood banker agonizes
over serologic incompatibility when there may not be
any adverse outcome from that incompatibility (eg, a
cold autoantibody). Very few antibodies outside the
ABO blood group system have serious life-threatening
potential. Most antibodies that are clinically significant
reduce the survival of the incompatible, transfused red
cells, but the extent of that destruction is dependent on
a multitude of factors. Occasionally, rapid destruction
can occur because of antibodies (eg, Kidd blood group
system) that cannot be detected in the pretransfusion
sample.

The process of antibody identification and selection

of the appropriate blood product is best served by an
overall look at all of the components of that process.
This chapter provides an overview of the components
and processes of antibody investigation and identifica-
tion. Many other available sources give detailed, exten-
19,55,56
sive procedures for problem-solving ; discussions
on all the different blood group antigens, antibodies,
6,19,20
and antigen frequencies ; and other clinical and sci-
entific aspects that define antigens and antibodies.18,19,25
Problem-Solving: A Case Study
A 39-year-old White female with colon cancer was ad-

mitted to a small hospital with a hematocrit of 25%. She
had received a transfusion 5 years before and had been
pregnant twice. Units for transfusion were requested
for surgery. The patient was group O, Rh positive. The
hospital performed an antibody screen and obtained
the results listed below (see Table 2-8).
The hospital crossmatched two units, which were
both incompatible during LISS 37 C spin and the IAT
phase of testing. Surgery was postponed and the pa-
tient was transferred to another facility. This facility
performed a panel (see Fig 2-1) and typed the patient to
be D+C–E+c+e+. The panel was performed using sa-
line techniques with no enhancement.
Table 2-8. Case Study: Results of Antibody Screen

IS LISS 37 C IAT
SC I 0 1+ 2+
SC II 0 0 W+
SC III 0 0 1+
IS = immediate spin; LISS = low ionic strength saline; IAT = indirect antiglobulin
test; SC = screening cell

Global View
The patient appears to have formed red cell antibodies.

She could have clinically significant red cell antibodies
because she has a history of both transfusion and preg-
nancy. Her low hematocrit, the positive autocontrol,
and her diagnosis could also indicate the presence of an
autoantibody. The negative result with cell number 3
could be consistent with a patient with autoanti-e
and/or multiple red cell antibodies.
The original screening results from the hospital are
consistent with this hypothesis. The difference in the
reactivity with screening cell 2 (usually R2R2 or e–) from
the original hospital and cell 3 on the second facility’s
panel could be the result of one of the multiple antibod-
ies or an autoantibody showing stronger reactivity
with the enhancement media.
The fact that the patient has not been recently trans-
fused indicates that the positive autocontrol is unlikely
to be associated with a delayed transfusion reaction.
There is no indication of a cold antibody. The reactions
seen during the 37 C spin phase show a pattern consis-
tent with anti-C (and the patient types C–).
When rule-out is performed, it is seen that the pa-
tient could have a combination of the following
a b
specificities: anti-C, anti-S, anti-Fy , anti-Jk , anti-K,
b
and anti-Le . Anti-e is not ruled out; however, the Rh
phenotype indicates the patient is e+. Alloanti-e in this
variant
e+ patient is unlikely. Alloanti-e in e is more com-
mon in the Black population, but this patient is White.
All the reactivity detected could be explained by the
multiple-antibody specificities that are not ruled out.
Weaker reactions are seen with Jk(b–) cells and may in-
b
dicate that anti-Jk is present.
Plan for Solution
Based on all the information gathered, several steps

could be taken next. A select panel of red cells, includ-
ing e–, C–, and Jk(b–) cells, may help eliminate other
specificities. A full phenotype using saline-reactive
reagents, such as monoclonal sources and/or chloro-
quine diphosphate-treated cells, may help determine
which specificities the patient can develop. A DAT may
or may not help in problem resolution because reagent-

Figure 2-1. Case study: Results of initial antibody panel (saline). IAT = indirect antiglobulin test; IS = immediate spin; ✓ = check cells positive
73
dependent reactivity seems unlikely with the presence

of a negative panel cell. Testing an enzyme-treated red
cell panel may help resolve the problem by denaturing
some of the target antigens.
Further testing is performed in the laboratory. A
DAT yields a positive result (1+ with anti-IgG). A full
p h e n o t y p e u s i n g D AT- n e g a t i v e , c h l o ro q u i n e -
diphosphate-treated red cells and controls is per-
formed. The patient’s phenotype is M+N+S–s+, P1+,
Le(a–b+), K+, Fy(a+b–), and Jk(a+b–). A papain-treated
panel showed the same results as the saline panel with
only a slight increase in reactivity on some red cells. A
selected panel is set up (see Fig 2-2).
Converging Evidence
The phenotype and second selected panel have ruled

a
out the possibility of anti-S, anti-Fy , anti-K, anti-e (auto
b
or allo), and anti-Le . Anti-C would explain the reac-
b
tion with cell 3 on the second panel. Anti-Jk would ex-
plain the reactions with cells 2 and 6. However, cells 1,
7, and 8 on this panel and cells 4 and 9 on the original
panel cannot be explained by the presence of only anti-C
b
and anti-Jk . An additional antibody must be present.
The two cells that are nonreactive on the second
a
panel are both Do(a–). The Do antigen is present in
19(p330)
about 65% of the population. Unexplained reac-
tions occur with nearly the same frequency.
Additional Do(a–), C–, and Jk(b–) cells were tested
and all were nonreactive. To prove the presence of
b
anti-C and anti-Jk , additional Do(a–) cells were tested
and they confirmed the presence of the two specifi-
cities. The patient’s chloroquine-treated cells type
Do(a–).
Look Back, Then Look Forward
The patient appears to have made three alloantibodies:

b a
anti-C, anti-Jk , and anti-Do . The patient also had a
positive DAT reaction. This result is unrelated to the
detected serum antibody. Although the patient had a
low hematocrit, there were no other signs of hemolysis
and the patient was not taking any medications known
to cause hemolysis. An elution was not performed be-

Figure 2-2. Case study: Results of select antibody panel (saline). IAT = indirect antiglobulin test; ✓ = check cells positive
75
cause the information gained by the procedure would

not affect her care. The original Do(a–) cell on the first
panel reacted because it was Jk(b+). The results and
conclusions are consistent with the original hospital’s
results.
Future blood needs for this patient will be difficult.
C–, Jk(b–), and Do(a–) units come from less than 2% of
Rh-positive donors and less than 10% of Rh-negative
b
donors. While potent commercial anti-C and anti-Jk
are available, there are no readily available sources of
a
anti-Do . The patient’s serum samples should be frozen
a
for future screening of red cells because anti-Do is of-
ten weakly reactive and often becomes undetectable in
the serum.
This patient has the potential to make additional
alloantibodies. Communication with the patient’s phy-
sician should include discussion of the possible prob-
lems that lie ahead concerning the patient’s current and
future needs.
References
1. Johnson ST, Rudmann SV, Wilson SM, eds. Serologic

problem-solving strategies: A systematic approach.
Bethesda, MD: AABB, 1996.
2. Silva MA, ed. Standards for blood banks and trans-
fusion services. 23rd ed. Bethesda, MD: AABB, 2004.
3. Harris T, ed. Standards for immunohematology ref-
erence laboratories. 3rd ed. Bethesda, MD: AABB,
2003:19.
4. Gregory K, Biehl R. Blood bank regulations: A to z.
6th ed. Bethesda, MD: AABB Press, 2005:220-5.
5. Voak D, Lodge TW, Hopkins J, Bowley C. A study of
antibodies of the HOI-B complex with special refer-
ence to their occurrence and notation. Vox Sang
1968;15:353-66.
6. Issitt PD, Anstee DJ. Applied blood group serology.
4th ed. Durham, NC: Montgomery Scientific, 1998.
7. Harris T. Two cases of autoantibodies that demon-
strate mimicking specificity in the Duffy blood
group system. Immunohematol 1990;6:87-91.
8. Lee JH, Mintz PD. Graft versus host anti-Rho(D) fol-
lowing minor Rh-incompatible orthotopic liver trans-
plantation. Am J Hematol 1993;44:168-71.

9. Garratty G. Problems associated with passively

transfused blood group alloantibodies. Am J Clin
Pathol 1998;109:769-77.
10. Boshkov LK, Kelton JG. Use of intravenous gamma-
globulin as an immune replacement and an immune
suppressant (review). Transfus Med Rev 1989;2:82-
120.
11. Salama A, Mueller-Eckhardt C. Use of Rh antibodies
in the treatment of autoimmune thrombocytopenia.
Transfus Med Rev 1992:6:17-25.
12. Rushin J, Rumsey DH, Ewing CA, Sandler SG. De-
tection of multiple passively acquired alloantibodies
following infusions of IV Rh immune globulin.
Transfusion 2000;40:551-4.
13. Anderson HJ, AuBuchon JP, Draper EK, et al. Trans-
fusion problems in renal allograft recipients: Anti-
lymphocyte globulin showing Lutheran specificity.
Transfusion 1985;25:47-50.
14. Swanson JL, Issitt CH, Mann EW, et al. Resolution of
red cell compatibility problems in patients receiving
anti-lymphoblast or anti-thymocyte globulin. Trans-
fusion 1984;24:141-3.
15. Arndt PA, Leger RM, Garratty G. Serology of anti-
bodies to second- and third-generation cephalo-
sporins associated with immune hemolytic anemia
and/or positive direct antiglobulin tests. Transfu-
sion 1999;39:1239-46.
16. Petz LD, Garratty G. Immune hemolytic anemia.
Philadelphia, PA: Churchill Livingstone, 2004.
17. Garratty G, ed. Blood group antigens and disease.
Arlington, VA: AABB, 1983.
18. Garratty G, ed. Immunobiology of transfusion medi-
cine. New York: Marcel Dekker, 1994.
19. Brecher ME, ed. Technical manual. 15th ed. Bethesda,
MD: AABB, 2005.
20. Reid ME, Lomas-Francis C. The blood group antigen
factsbook. 2nd ed. San Diego, CA: Elsevier Academic
Press, 2004.
21. Koepke JA. Specimen collection for cellular hematol-
ogy. In: Koepke JA, ed. Practical laboratory hematol-
ogy. New York: Churchill Livingstone, 1991:31-40.
22. Gunson HH. A serum agglutinin inhibited by ion-
ized calcium. Vox Sang 1969;17:511-24.
23. Rumsey DH, Ciesielski DJ. New protocols in sero-
logic testing: A review of techniques to meet today’s
challenges. Immunohematol 2000;16:131-7.

24. Judd WJ. Commentary: Testing for unexpected red

cell antibodies—two or three reagent red cell sam-
ples? Immunohematol 1997;13:90-2.
25. Mollison PL, Engelfriet CP, Contreras M. Blood
transfusion in clinical medicine. 10th ed. Oxford,
England: Blackwell Scientific Publications, 1997.
26. Campbell SA, Shirey RS, King KE, Ness PM. An
acute hemolytic transfusion reaction due to anti-IH
in a patient with sickle cell disease. Transfusion
2000;40:828-31.
27. Beattie KM, Zuelzer WW. The frequency and prop-
erties of pH-dependent anti-M. Transfusion 1965;
5:322-6.
28. Rolih S, Thomas R, Fisher F, Talbot J. Antibody de-
tection errors due to acidic or unbuffered saline.
Immunohematol 1993;9:15-18.
29. Kennedy MS, Waheed A, Moore J. ABO discrepancy
with monoclonal ABO reagents caused by a pH-
dependent autoantibody. Immunohematology 1995;
11:71-3.
30. Wilkerson SL. Serological approaches to transfusion
of patients with allo- or autoantibodies. In: Nance SJ,
ed. Immune destruction of red cells. Arlington, VA:
AABB, 1989:227-61.
31. Daniels G. Effects of enzymes on and chemical mod-
ifications of high frequency antigens. Immunohema-
tol 1992;8:53-7.
32. Eckrich RJ. Inactivation of Holley (HY) and Gregory
(GY) antigens by dithiothreitol (DTT). Immunohe-
matol 1988;4:1-4.
33. Toy EM. Inactivation of high-incidence antigens by
dithiothreitol. Immunohematol 1986;2:57-8.
34. Swanson JL, Sastamoinen R. Chloroquine diphos-
phate stripping of HLA, A,B antigens from red cells
(letter). Transfusion 1985;25:439-40.
35. McShane K, Cornwell S. Chloroquine reduces anti-
gen strength (letter). Transfusion 1985;25:83.
36. Mallory D, Reid M. Misleading effects of chloro-
quine (letter). Transfusion 1984;24:412.
37. Sassetti R, Nicholls D. Decreased antigen reactivity
caused by chloroquine (letter). Transfusion 1982;22:
537-8.
38. Rygiel SA, Issitt CH, Fruitstone MJ. Destruction of
the S antigen by Clorox (abstract). Transfusion 1983;
23:410.

39. Kerr MA. The structure and function of human IgA.

Biochem J 1990;271:285-96.
40. Reid ME, Moore BP, Poole J, et al. TSEN: A novel
MNS-related blood group antigen. Vox Sang 1992;
63:122-8.
41. Garratty G. Screening for RBC antibodies: What
should we expect from antibody detection RBCs?
Immunohematol 2002;18:71-7.
42. Kanter MH, Poole G, Garratty G. Misinterpretation
and misapplication of p values in antibody identifi-
cation: The lack of value of a p value. Transfusion
1997;37:816-22.
43. Armstrong JK, Leger R, Wenby RB, et al. Occurrence
of an antibody to polyethylene glycol in normal do-
nors (abstract). Blood 2003;102(Suppl):556a.
44. Hoffer J, Koslosky WP, Gloster ES, et al. Precipita-
tion of serum proteins by polyethylene in pretrans-
fusion testing. Immunohematol 1999;15:105-7.
45. Garratty G. In vitro reactions with red blood cells
that are not due to blood group antibodies: A re-
view. Immunohematol 1998;14:1-11.
46. Hamilton JR, McCraney S, McGuire D, et al. Valida-
tion of expired red cells for use in antibody tube tests
(abstract). Transfusion 2004;44(Suppl):20A-21A.
47. Hamilton JR, Dickstein B, McCraney S, Kosanke J.
Validation of expired 0.8% red cells for use in anti-
body gel tests (abstract). Transfusion 2004;44 (Suppl):
123A.
48. Rodberg K, Tsuneta R, Garratty G. Discrepant Rh
phenotyping results when testing IgG-sensitized RBS
with monoclonal Rh reagents (abstract). Transfusion
1995;35(Suppl):67S.
49. Judd WJ, Barnes BA, Steiner EA, et al. The evaluation
of a positive direct antiglobulin test (autocontrol) in
pretransfusion testing revisited. Transfusion 1986;
26:220-4.
50. Judd WJ, Butch SH. Cost containment in the blood
bank: Eliminating unnecessary serologic testing. J
Med Technol 1984;1:485-95.
51. Judd WJ. Prewarmed tests: Con. Transfusion 1995;
35:271-5.
52. Mallory D. Prewarmed tests: Pro-why, when, and
how—not if. Transfusion 1995;35:268-70.
53. Leger RM, Garratty G. Weakening or loss of anti-
body reactivity after prewarm technique. Transfu-
sion 2003;43:1611-14.

54. Leger RM, Garratty G. Evaluation of methods for de-

tecting alloantibodies underlying warm autoanti-
bodies. Transfusion 1999;39:11-16.
55. Judd WJ. Methods in immunohematology. 2nd ed.
Durham, NC: Montgomery Scientific, 1994.
56. Mallory D, ed. Immunohematology methods.
Rockville, MD: American Red Cross, 1993.

AABB Antibody Identification

Uploaded by

Copyright:

Available Formats

AABB Antibody Identification

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AABB Antibody Identification

Uploaded by

Copyright:

Available Formats

What are some of the tools that have become more commonplace in antibody identification?

What are some of the tools that have become more commonplace in antibody identification?

How has our understanding of the clinical significance of some antibodies changed?

How has our understanding of the clinical significance of some antibodies changed?

Antibody Identification

By Teresa Y. Harris, MT(ASCP)SBB, CQIA(ASQ)

• Adobe Reader 6.0 or higher

Download Quantities and Expiration

Terms and Conditions

Copyright © 2005 by AABB. All rights reserved

Copyright © 2005 by AABB. All rights reserved

in this discussion where applicable in the antibody

The AABB Standards for Blood Banks and Transfusion Ser-

Copyright © 2005 by AABB. All rights reserved

for transfusion shall lack the corresponding anti-

Antibody identification is dependent on serologic and

Copyright © 2005 by AABB. All rights reserved

Table 2-1. Components of Antibody Identification

Clinical history Pregnancy Age

Copyright © 2005 by AABB. All rights reserved

Both the serologic and the nonserologic components

Nonserologic Components of Antibody Identification

While collecting patient history is the very important

Table 2-2. Steps in Antibody Identification

Copyright © 2005 by AABB. All rights reserved

Transfusions, Transplants, and Pregnancies

Obviously, a patient who has a history of pregnancy

Drug or medication history can be very informative in an-

Copyright © 2005 by AABB. All rights reserved

History of Infections or Hemolysis

A history of viral illness may suggest the presence of

Table 2-3. Possible Serologic Outcomes of Therapies

IV immunoglobulin Anti-A, anti-B, anti-D,

Copyright © 2005 by AABB. All rights reserved

been formed, reportedly as a result of viral disease. In

Copyright © 2005 by AABB. All rights reserved

Race or ethnic background may be helpful factors in

Copyright © 2005 by AABB. All rights reserved

information gathering is a good start for the process of

Occasionally a transfusion reaction results from an un-

Copyright © 2005 by AABB. All rights reserved

Some of the other specificities that are associated

Copyright © 2005 by AABB. All rights reserved

has never been transfused, but presents with a positive

Experience and Competency of the Staff

Although many available resources cover the topics of

Copyright © 2005 by AABB. All rights reserved

fine-tune their technique. It can be beneficial to share

Several steps should be performed each time a sample

Copyright © 2005 by AABB. All rights reserved

lems still occur. If a sample is mislabeled or illegible,

Copyright © 2005 by AABB. All rights reserved

next problem-solving technique. Contingency plans

Copyright © 2005 by AABB. All rights reserved

bodies may be reactive only in the presence of

Serologic Components of Antibody Identification

Antibody Detection and Identification Proto-

There are many standard approaches to antibody de-

Copyright © 2005 by AABB. All rights reserved

Currently there are three main ways to investigate