Interferences occurring in blood testing can result in potentially fatal errors or cause a delay in supplying blood for transfusion. The sources of interference in testing are generally categorized as pre-analytical, analytical, and post-analytical. In general, a large pro- portion of errors occur in the pre-analytical phase of the tests. Pre-analytic variables that can influence the test results can be grouped broadly into two categories:
(1) those related to specimen collection, handling, and processing and (2) those related to physiologic factors and patient’s endogenous variables, including diseases, circulating antibodies, and drug therapy.
Physiologic variables include the effects of age, sex, time, season, altitude, conditions such as menstruation and pregnancy, and lifestyle. Among these, age, sex, and pregnancy status are very important when making recommendations in transfusion medicine. Other patient-related variables, such as clinical diagnosis, problem list, past and current medication, and medical or surgical procedures, are important and may be directly related to the testing results. For this reason, knowledge of medical history is required for correct interpretation of test results and release of the right type of blood. Unlike for other laboratory testing, the fasting or postprandial status, prior exercise, posture during blood drawing, or specific timing of blood sam- ple collection in regard to circadian or physiologic cycles are generally not known to interfere with trans- fusion medicine assays. Factors related to specimen collection, such as effects of the duration of tourniquet application, the anticoagulant:blood ratio, specimen handling, and processing steps in the preparation of serum/plasma and RBC separation, can introduce an important pre-analytic variable. If they result in hemo- lysis or fibrin clots being present in the sample, this may be misinterpreted as a positive reaction.
At the initial time of interpretation, the true cause for test results is not known; thus, interferences and errors may look alike. Hemolysis resulting from incorrect drawing looks like hemolysis due to immune causes. If suspected, potential errors, including technical and clerical errors, have to be 281
INTERFERENCES IN BASIC TRANSFUSION MEDICINE TESTS
ruled out for correct test interpretation. Interferences should be distinguished from errors occurring in the process of blood testing and administration. A major potential source of error can occur when the proto- cols for patient identification or blood drawing are not respected and the patient is not properly identi- fied (blood is drawn from a patient different from the one intended) or when the blood sample is col- lected from the right patient but mislabeled or blood samples are mixed up. In these instances, the wrong blood in the tube (WBIT) is tested. Although uncom- mon, WBIT does not represent a true interference but, rather, an error in the process that requires early recognition and further investigation to prevent such events from reoccurring. Two blood samples from the same patient are required for ABO/Rh determi- nation prior to transfusion in order to prevent cleri- cal error or detect possible WBIT. If WBIT is suspected, the test should be repeated on a different sample collected from the same patient. Case Study 1 demonstrates why a high level of alertness should be maintained concerning this rare but still possible source of error in blood bank testing. It also high- lights that interpretation of blood tests is part of a process with multiple sources of potential errors and how one of the most critical errors can be initially presented to the clinical pathologist as a discrepancy.
Methodology used for testing plays a definite role in test interpretation. In general, microcolumn gel tech- nology is known to be more sensitive than tube testing and less sensitive than solid-phase adherence technol- ogy for detection of clinically significant alloantibodies.
The advantage of using a methodology with increased sensitivity is better identification of patterns of reactiv- ity specific to an antibody that might not be elucidated by a less sensitive methodology. Studies showed that some clinically significant alloantibodies identified by solid-phase technology displayed a WRNS pattern in gel testing. Few studies reported, however, that gel technology is less sensitive than the tube test method- ology for detecting ABO isohemagglutinins and expected anti-A or anti-B may be missed. The mono- clonal anti-A and -B reagents used in the gel cards are also not known to react with B(A), acquired-B, or poly- agglutinable RBCs. On the other hand, reagents and chemicals present in the reaction environment of vari- ous testing kits may also result in false-positive reactions and trigger extensive testing to rule out the presence of alloantibodies. Similarly, a patient’s autoantibodies can react with reagent RBCs present in the antibody screen and identification panel and conse- quently mask detection of reactivity due to an alloanti- body. Cold agglutinins and warm autoantibodies are known for giving a pan-reactivity pattern and masking possible underlying alloantibodies. Autoantibodies to
reagents usually lead to false-negative reactions due to reagent consumption or less availability for detection of the true target.
Interferences, defined as alterations in the expected reactivity pattern potentially misleading the interpreta- tion and correct identification of significant findings, can occur in various tests performed in the blood bank, but their correct interpretation requires an inte- grated review of the case.
Interferences in ABO/Rh Typing
ABO discrepancies include disagreement between historical and current blood type, discrepancy between forward and reverse reactions, reactivity weaker than expected (41) between the RBC antigens and corre- sponding antibodies, and detection of mf type of reactivity. ABO discrepancies detected in patients to be transfused must be resolved before any blood compo- nent is transfused unless blood is urgently needed, in which case group O RBCs or AB plasma are issued.
Discrepancies occurring in donor samples must be resolved before the blood unit is labeled with a blood type. At the initial read of an ABO discrepancy, it is not known which of the typing reactions—the forward or the reverse—reflects the patient’s true ABO type.
Patient’s age, clinical diagnosis, historical blood type, transfusion history, and the results of other tests are useful hints needed for final interpretation.
Clerical and technical errors must first be ruled out because they are the most common causes for ABO discrepancies. Clerical errors are responsible for more than 95% of fatal transfusion reactions and imply patient or specimen misidentification or blood sample mix-up. Causes of technical errors include failure to follow manufacturer’s instruction, failure to add cells or reagents, incorrect test cell preparation so that the pipette dispenses a lower amount of A or B cells into the well, improper centrifugation or incubation condi- tions, use of contaminated reagents, or defective equip- ment [2,4,10,12]. Because clerical and technical errors are not uncommon, the best course of action is to repeat testing on a better washed RBC sample and use plasma from the original specimen to determine if the discrepancy persists. Repeat blood drawing might be needed, especially if sample mix-up is a concern.
True ABO discrepancies (not due to clerical and tech- nical errors) are generally grouped into four categories:
1. Weak or loss of expected RBC antigen 2. The presence of unexpected RBC antigen-like
reactivity
3. Weak or loss of expected antibody
4. The presence of unexpected antibody reactivity.
Weak or Absent Reactivity of Expected Antigen Weak A or B antigens are seen in subgroups of A and B blood types due to age, such as in newborns and the elderly, or disease process. Certain hematological malignancies (leukemia and Hodgkin’s lymphoma) are associated with aberrant transferase formation leading to fewer antigens being formed on the red cell mem- brane, whereas solid organ cancers are associated with an excess of soluble substance similar to A or B antigen and can neutralize the typing reagents and result in a false-negative reaction. Chimerism, the presence of a dual population of cells, is routinely seen post ABO- mismatched stem cell transplant and, rarely, due to vascular anastomosis in fraternal twins. In chimerism, ABO/Rh testing is expected to be discrepant within approximately 1 month post-transplant. Transfusion of non-ABO-specific blood, such as massive transfusion of O Rh-negative RBC units in trauma patients, as well as fetal maternal hemorrhage, also creates a chimeric state. Therefore, the resolution of an ABO discrepancy starts with checking the patient’s age and clinical condition.
The Presence of Unexpected Red Blood Cell Antigen-Like Reactivity
Detection of unexpected antigen may be due to misidentification of rouleaux for agglutination, poly- agglutinable RBCs, or interference from other sub- stances causing RBC agglutination, such as Wharton’s jelly. Autoagglutination due to cold autoantibodies often causes unexpected reactivity (antigen-like in the ABO/Rh typing but also antibody-like in the antibody screen). Antibodies not specific to blood group anti- gens but reacting with chemicals or drugs present in the reaction microenvironment may also cause this type of discrepancy. Stem cell transplantation, B(A) phenomenon, and fetal maternal hemorrhage are other known causes of unexpected antigen reactivity.
Mixed-field agglutination is usually noted in chi- meras, indicating a mixed RBC population (such as in post-transfusion, massive transfusion of another blood group, and bone marrow transplant).
Polyagglutination is a phenomenon suspected when patient’s RBCs cross reacts with anti-A reagent (mf).
Polyagglutinable RBCs also react with most normal adult sera but not with autologous serum. This phe- nomenon is due to exposure of a cryptantigen on the RBC surface as a result of the action of an enzyme associated with a microorganism (acquired microbial polyagglutination), passive adsorption of microbial structures with antigenic structures similar to A, B, H, T, and Tn antigens or may be associated with inher- ited conditions or acquired due to a somatic stem cell mutation.
Weak or Loss of Expected Antibody
The most frequent cause of this discrepancy is fail- ure to detect anti-B in A or O plasma samples. This can readily be corrected by the traditional tube test, which is usually performed at RT and sometimes at 4C. Weak or absent antibodies are found in new- borns, elderly, immunosuppressed patients, post stem cell transplant, and in severe immunodeficiencies.
The Presence of Unexpected Antibody Reactivity Detection of unexpected reactivity may be due to rouleaux formation, blood subgroups, passively acquired antibodies (via transfusions or administration of Rh immunoglobulins, intravenous immunoglobulins, or other drugs), and passenger lymphocytes syndrome.
Individuals with an ABO subgroup can develop alloan- tibodies that react with reagent RBCs of the reverse typing. The classic example is the A2 subgroup with anti-A1 antibody. Positive reaction can also result from cold agglutinins and true alloantibodies reacting with antigens present on the surface RBCs used for testing (e.g., anti-M, anti-N, anti-P1, and anti-c). A citrate- dependent autoantibody causing errors in blood grouping has also been described[22].
Interferences in the Detection of Antibodies
The antibody identification process is not always straightforward. If clerical and technical errors are ruled out, the reactivity observedin vitromight be due to false positives, real antigen antibody reactions, and nonimmune interactions.
False-positive reactions or pseudoagglutination were described previously. Common causes of interfer- ences in this category include the presence of fibrin clots, rouleaux formation, and agglutination due to albumin and plasma expanders. Real antigen anti- body reactions can result from simple or combined presence of allo- or autoantibodies either directed against specific RBC antigens or not related to them.
Autoantibodies typically react with the patient’s own cells and all reagent RBCs present in the antibody screen and panel in a nonspecific manner. Due to their pan-reactivity, they may mask the detection of alloanti- bodies. Distinguishing between autoantibodies and alloantibodies is essential, but the answer is not always in the positive autocontrol because transfused patients with developing alloantibodies have a positive test. It is rather the reactivity pattern that suggests the type of autoantibody. Autoantibodies reacting only at cold temperatures, including RT (thus in in vitro testing) but not at 37C, are known as cold autoantibodies (CAA) and are generally considered clinically insignifi- cant. They become clinically significant if their 283
INTERFERENCES IN BASIC TRANSFUSION MEDICINE TESTS
reactivity extends beyond 32C into the body tempera- ture, such as in the case of cold agglutinin disease (CAD), or if the patient undergoes hypothermia, such as in cardiac surgery. The typical CAA is (21) or (mf) equally pan-reactive with reagent RBCs present in the antibody screen and extended panel. Autoantibodies reacting at 37C or in the presence of AHG (AHG phase) are described as warm autoantibodies (WAA).
The typical WAA is strong (31 or 41) and equally pan-reactive with all reagent RBCs. The clinical signifi- cance of WAA is related to their propensity to cause hemolysis. Because both CAA and WAA are typically pan-reactive, if present, they may mask or interfere with the detection of alloantibodies. In such cases, fur- ther workup is needed to rule out the presence of underlying alloantibodies against major RBC antigens.
Interferences due to passively transfused antibodies occur in patients with a prior transfusion with plasma products containing alloantibodies (e.g., intravenous immunoglobulins (IVIg) or anti-D) or status post- organ or stem cell transplant (passenger lymphocyte syndrome).
Interferences due to antibodies present in the patient’s plasma not due to RBC antigens have been described for a variety of chemicals, antibiotics, poten- tiators (LISS), or other substances (lactose, lactate, meli- biose, phenol, sucrose, and thrombin) present in the testing environment or commercial antisera [23 25].
Chemicals that may lead to generation of antibodies reacting with RBCs are paraben, thimerosal, sugars, EDTA, inosine, citrate, acriflavine, sodium caprylate, yellow No. 5, and tartrazine. Antibiotics known to cause production of antibodies that are also capable of react- ing with RBCs but not via blood group antigens are penicillin, chloramphenicol, neomycin sulfate, gentami- cin, tetracycline, streptomycin, and vancomycin.
At least three mechanisms have been described for the reactivity of RBCs: (1) Antigen antibody com- plexes may form in the test environment leading to RBC agglutination, (2) antibodies may adsorb onto RBC surface and bind to antigen, and (3) RBC aggluti- nation may be enhanced by the presence of exogenous substances independent of an antigen antibody reac- tion. Chemicals and drugs interfere with testing by either covalent (penicillin) or noncovalent link to RBCs. In the latter case, washing may remove the reac- tivity due to residual plasma or due to antibodies that are not covalently bound. Antibodies against neomy- cin, chloramphenicol, gentamicin, hydrocortisone, sugars, dyes (acriflavine, yellow No.5, and tartrazine), sodium azide, and thimerosal are noncovalently bound, thus the reactivity due to these antibodies is removed after RBC washing, whereas antibodies against penicillin, inosine or EDTA are not removed by washing and additional testing is required.
Occasionally, these antibodies may display blood group antigen specificity. Other immune interactions were described for an RBC age-dependent antibody and anti- bodies to lower oxiranes (ethylene, propylene, and butylene oxides) used in the sterilization of polyvinyl chloride blood donor packs. Sterilization of the outside of the bag with propylene oxide sometimes caused the anticoagulant in the bag to acquire properties that could induce RBCs to acquire a new antigen, termed lower oxirane. If these antibodies are present, they can create problems in pre-transfusion testing and present anoma- lies in ABO, Rh grouping, and antibody detection.