| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
Central Laboratory, Norwegian Radium Hospital, Montebello, N-0310 Oslo, Norway
aAuthor for correspondence. Fax 47-22-730725; e-mail johan.bjerner{at}klinmed.uio.no.
Human heterophilic antibodies may bind the animal antibodies used in an immunoassay and thus produce erroneous results. Such interference has been increasingly recognized as a diagnostic problem (1)(2). In this issue of Clinical Chemistry, Adel Ismail (3) discusses the two main issues of heterophilic antibody interference in immunometric assays: how to detect it and how to avoid it. This is an important discussion to which all clinical chemists should feel invited. This field requires not only a profound understanding of heterophilic antibody interference but also hands-on experience with immunoassays.
To detect heterophilic antibody interference in the sample, the first critical step is to suspect it. The starting point is often a clinician contacting the laboratory about a mismatch between the clinical information and laboratory results. Once there is suspicion of such interference, a classic approach has been to make serial dilutions of the sample. As Ismail (3) points out, most samples with heterophilic antibody interference will then display nonlinearity. The author does not explain this nonlinearity but draws attention to the polyclonality of interfering antibodies, with affinities ranging from high to low, and he points out that with human antibodies more than one functional binding site is often involved. Human antibodies with high affinities for animal antibodies are specific for one or more animal species and do not bind human immunoglobulins. High-affinity human anti-human immunoglobulins would instantly form larger complexes, either to be removed from circulation or to cause serious disease. In our experience before the surge in therapeutic use of murine or humanized murine monoclonal antibodies, we found little evidence of interfering immunoglobulins from the immunoglobulin classes that normally exhibit higher affinities, i.e., IgA or IgG (personal observation). Low-affinity human antibodies often bind to both animal and human immunoglobulins, making heterophilic antibodies (human anti-animal) and rheumatoid factors (human anti-human) overlapping and sometimes confusing entities (4). Such antibodies do not form complexes in vivo, although they bind the stacked capture antibodies on the solid phase in immunometric assays. The explanation might be that the stacking of antibodies forms repetitive epitopes that may be targets for more than one of the functional binding sites of the interfering antibodies.
Ismail (3) further describes several aspects of the immunoglobulin structure of importance for the binding of heterophilic antibodies. We suggest further complicating mechanisms. During the first step of an immunometric assay, the assay antibodies are exposed to human immunoglobulins. Not only variable but also constant regions of these human antibodies may bind assay immunoglobulins (5). Immunoglobulins are also bioactive molecules, interacting with nonimmunoglobulins. Human IgM arrayed in immune complexes binds and activates complement C1q (6); in fresh samples, one should thus theoretically expect the interfering antibodies to form complexes with complement on the solid phase, an aspect of heterophilic antibody interference that has never been addressed. Finally, we should regard the first step of an immunoassay as the encounter between a limited number of immobilized animal antibodies in close relation to each other and a huge concentration of human immunoglobulins and nonimmunoglobulins containing a broad spectrum of reactivities. In such settings, steric effects and saturation kinetics and effects are probably important.
In immunometric assays, we must discuss affinities in terms of the ratio of two rates, an association rate and a dissociation rate (7). High-affinity antibodies usually combine a high association rate with a low dissociation rate, making those antibodies excellent for assay use. Low affinities imply a low ratio of the rates. Immunoglobulins have more than one binding site, however, and their binding properties are determined by the combined affinities of all functional binding sites, the avidity. Increasing the number of functional binding sites will not significantly increase the association rate, which is limited mainly by diffusion properties (7), but because multiple bonds are less likely to break, the dissociation rate slows. Theoretically, low-affinity human IgM antibodies interfering in assays thus will display a slow association rate (long time to steady state) combined with a low dissociation rate (explaining why the washing steps fail to clear interfering antibodies). Our experience also tells us that increasing incubation times both increases the number of interference-positive samples and the magnitude of the interference. We thus disagree with Ismail’s suggestion that short incubation times lead to higher apparent interference. However, his statement may yet be correct for individuals with high concentrations of interfering heterophilic antibodies, where the limited number of immobilized capture antibodies may be saturated early during the incubation.
Binding to the capture antibody is only one of two prerequisites for heterophilic antibody interference, the other being the simultaneous binding of tracer antibody. Ismail (3) notes that assay format is important for interference. Studies have shown that binding of human immunoglobulins to animal or human assay antibodies is common but does not always imply interference, i.e., simultaneous binding of the tracer antibody (8). Both we and Ismail(3) suggest that multiple binding sites are involved in the attachment of the interfering immunoglobulins to the stacked capture antibodies. How then may the interfering immunoglobulins bind the single Fc epitope present on the tracer antibody? And why do one-step heterogeneous assays, with only one final washing step to remove unbound tracer antibodies, have a higher frequency of interference than two-step heterogeneous assays, in which unbound serum proteins are removed before the addition of tracer antibody? Are there unknown compounds in serum important for enhancing the binding of tracer antibody?
Homogeneous assays have no washing steps because the analysis principle here allows direct differentiation between bound and free antibody. We would expect such an assay format to be sensitive to interferences, but when we validated a homogeneous assay for carcinoembryonic antigen for heterophilic antibody interference, this was not the case (9). The homogeneous assay format tested had two potential advantages: the lack of a washing step allowed reactions to take place in solution (where multiple binding sites are not involved); and the technology allowed several readings to be taken during the course of the incubation, providing a curve of the assay kinetics. Such information on assay kinetics may be used by the instrument software to identify samples with interference, in which assay kinetics will be either slower or faster than for true antigens.
Other methods are available to detect interference produced by human antibodies against animal immunoglobulins. Boscato and Stuart (10) demonstrated how a "nonsense" assay could be constructed to flag samples showing interference. Such a nonsense assay for analytes with single epitopes may consist of the original capture antibody used as both capture and tracer antibody. With no interference present, readings will be in the blank range. The drawback is that this implies duplicate assays. Shifting to flow cytometry, our group developed an assay for 
-fetoprotein in which the interference testing was performed simultaneously for all samples (11). This method is elegant and easy to automate, but the principle has never been applied to routine commercial assays. The methods for detecting interference by nonsense assays are not completely flawless, however. Under certain conditions, interference may be present but not be detected (9). The advantages outweigh the drawbacks, however, because such a nonsense assay would flag all kinds of interference, immunoglobulin or nonimmunoglobulin, even if its nature is unknown to us.
The second question is perhaps more important: how do we eliminate heterophilic antibody interference? If we eliminate it, we do not need to detect it. The attempts to date can be divided into three groups: (a) removal or inactivation of the interfering immunoglobulins from samples; (b) modification of assay antibodies to make them less prone to react with heterophilic antibodies; and (c) use of buffer additives that reduce interference.
Removal of immunoglobulins from samples has been done by precipitation with polyethylene glycol (PEG) (12). This requires that a known amount of serum be taken from the vial before PEG is added. Aggregates must then be separated from the serum by centrifugation or filtration. Ismail (3) has demonstrated that such precipitation could be put into routine use and suggests automation of this preanalytic step. Our experience, however, is that precipitation efficiency depends on sample immunoglobulin concentration, which in our cancer hospital ranges from 0.5 to 100 g/L. More importantly, coprecipitation of analytes limits the use of the method (3). Finally, the method could significantly delay and add cost to routine assays. PEG precipitation can therefore not be considered as the final solution for all samples. Elimination of interfering immunoglobulins has also been achieved by heat treatment of samples (13), but the use of this harsh method is limited to assays for carcinoembryonic antigen and other extremely robust analytes. Although we do not consider removal of immunoglobulins as the preferred solution, there are possibilities for automation of milder methods. We have seen an automated instrument combining chromatography and immunoassay for the analysis of -fetoprotein with different glycosylations (14). Such an instrument may be modified for the preanalytic removal of immunoglobulins.
Humanization of animal antibodies (15) and modification of assay antibodies by removal of the Fc fragment (16) are well-known methods that often dramatically reduce interference. Encouraged by our experience with F(ab')2 fragments (17), we have developed assays using the smaller single-chain fragments (scFv), which may be even better. Changing from immunoglobulins to nonimmunoglobulin binding substances (18), such as affibodies or aptamers, might also be a solution.
Buffer additives may reduce interference. Native animal immunoglobulins are often effective, and aggregating them potentiates the effect (17). The additives to buffers seem to be hidden corporate knowledge. Sadly, little is published in this field.
We do not believe that the current countermeasures eliminate heterophilic antibody interference completely. Awaiting the final (and radical) solution to come, such interference must be fought on several fronts. We believe that by better understanding of the basic principles of heterophilic antibodies we may substantially reduce interference by combining modified assay antibodies, a proper blocking buffer, and an assay format that is less sensitive to interference. The possibility for interference will still be with us, however, and the use of a simultaneous nonsense assay that detects interference should be included when possible. Finally, we appeal to the producers of commercial immunometric assays: Openness about the measures used in commercial immunoassays to avoid interference would help the clinical chemist in validating assays and predicting possible interference in the clinical setting where the assays affect patient care.
References
-fetoprotein ratio and total -fetoprotein concentration. Anal Chem 1998;70:2110-2114.[Medline]
[Order article via Infotrieve]
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  A A A Ismail Interference from endogenous antibodies in automated immunoassays: what laboratorians need to know J. Clin. Pathol., August 1, 2009; 62(8): 673 - 678. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P M Clark Laboratory services for thyroglobulin and implications for monitoring of differentiated thyroid cancer J. Clin. Pathol., May 1, 2009; 62(5): 402 - 406. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B E Fritz, R J Hauke, and D F Stickle New onset of heterophilic antibody interference in prostate-specific antigen measurement occurring during the period of post-prostatectomy prostate-specific antigen monitoring Ann Clin Biochem, May 1, 2009; 46(3): 253 - 256. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. E.P.B. Ballieux, N. I. Weijl, H. Gelderblom, J. van Pelt, and S. Osanto False-Positive Serum Human Chorionic Gonadotropin (hCG) in a Male Patient with a Malignant Germ Cell Tumor of the Testis: A Case Report and Review of the Literature Oncologist, November 1, 2008; 13(11): 1149 - 1154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. La'ulu and W. L. Roberts Second-Trimester Reference Intervals for Thyroid Tests: The Role of Ethnicity Clin. Chem., September 1, 2007; 53(9): 1658 - 1664. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U.-H. Stenman Improving Immunoassay Performance by Antibody Engineering Clin. Chem., May 1, 2005; 51(5): 801 - 802. [Full Text] [PDF]
|
|
|
|
|
|
D. J. Warren, J. Bjerner, E. Paus, O. P. Bormer, and K. Nustad Use of an In Vivo Biotinylated Single-Chain Antibody as Capture Reagent in an Immunometric Assay to Decrease the Incidence of Interference from Heterophilic Antibodies Clin. Chem., May 1, 2005; 51(5): 830 - 838. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
