| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
Department of Clinical Biochemistry, The Princess Royal Hospital, Haywards Heath, West Sussex, England RH17 4EX, Fax 44-1444-459635, E-mail SJFrost{at}bigfoot.com
To the Editor:
In their recent article, Kaplan and Levinson (1) provide an interesting account of the distinction between weak multispecific heterophile antibodies and the highly avid and specific human anti-animal (e.g., mouse) antibodies that may arise from medical treatment with animal proteins. I would like to draw attention to a recent case study I coauthored (2) that was not cited in the article. This highlights the potential confusion of the terms "heterophile antibodies" and "HAMAs", as discussed by Kaplan and Levinson. However, the case also shows the difficulty in being too pedantic in making the distinction between the two along the lines they suggest.
We reported a case of an interference with a thyroid-stimulating hormone assay, and my initial draft referred to this as attributable to heterophile antibodies. However, one of the referees of the article preferred the term HAMAs (human anti-mouse antibodies), so I amended the manuscript accordingly (being unsure myself what to call the interference). The next time I discussed the case with a clinician, I used the term HAMA, but he looked puzzled. "Had the interference changed?" he asked, so I reverted to calling it heterophile.
The patient has received no medication with animal products and insists that she has had no contact with mice but had handled pet hamsters. According to the authors' proposed terminology, because there was no clear specific immunogen I should have stuck to the term heterophile all along.
The problem is that in our case the antibodies were sufficiently avid to interfere in two competitive immunoassays, for free thyroxine and free triiodothyronine. This interference was sufficiently marked to cause potentially clinically misleading results. According to the authors, this should occur only if the interference is attributable to more avid HAMAs and not heterophile antibodies. I would be interested to hear whether the authors consider our interference attributable to exceptionally strong avidity heterophile antibodies or if, despite the patient's testimony, she has somehow been exposed to a specific mouse immunogen. The patient, incidentally, has Crohn disease, which may have contributed to an unusual presentation.
The authors' suggestion that heterophile antibody interference would, by definition, be too weak to interfere with our non-sandwich assays might have led us astray in the reported case. It would seem pragmatic, once an interfering factor is discovered, to always consider possible interference in other immunoassays. This can usually be established by simple and well-known methods, as described in our case report.
References
1
Department of Pathology, and Laboratory Medicine, University of Louisville, and, Laboratory Service, Veterans Administration Medical Center, 800 Zorn Ave., Louisville, KY 40206-1466
2
LabCorp, 4500 Conaem Dr., Louisville, KY 40213
a Address correspondence to this author at: Veterans Administration Medical Center, 800 Zorn Ave., Louisville, KY 40206-1466. Fax 502-894-6265; e-mail levinson{at}louisville.edu
To the Editor:
The letter by Dr. Frost highlights the difficulty in classifying antibody interferences. We believe that the complexity of the subject requires the attention of an Expert Committee, providing a way for the experiences of many to be brought together to best solve this problem of terminology. In the absence of such a Committee, we made some suggestions as to a reasonable nomenclature that would clarify terminology for most cases (1). We hoped that our report would stimulate the organization of such a Committee by an appropriate agency, such as AACC. We are glad to have an opportunity to respond to the specific case outlined in the letter.
The letter refers to a report by Frost et al. (2) that describes an interesting case in which a two-site thyroid-stimulating hormone (TSH) assay, and free triiodothyronine (T3) and free thyroxine (T4) competitive protein binding (CPB) assays were affected by antibody interference. According to the letter, the authors initially called this interference heterophile, but the reviewer preferred "HAMA" because the assay antibodies were mouse.
Based on the nomenclature that we suggested, heterophile antibodies that were shown to react only with mouse immunoglobulin and received no other species challenge could be called human anti-mouse antibodies (HAMAs), but not specific HAMAs. Thus, the classification suggested by the reviewer would be correct because there was no good history of direct exposure and because the report described adsorption using non-immune mouse immunoglobulin but no other (2). As we pointed out in our Opinion (1), the question of animal handling is a very difficult one and "it may be unclear whether these are heterophile antibodies or not". But, in most cases these antibodies are cross-reacting because of similar epitopes on antibodies of different species. For this reason, these also tend to be weak antibodies similar to heterophile antibodies, and unless they are tested against non-immune globulin of various species, in our opinion, they may be considered heterophile (1).
In the report by Frost et al. (2), if the authors tested the serum with other types of non-immune globulin, especially hamster, they could have defined the specificity of the interference better, but they did not. In his letter, Dr. Frost notes that by the reckoning in our Opinion (1), heterophile antibodies would not be expected to interfere with CPB assays. Actually, we stated that the CPB assay "showed much less interference" (1). From reviewing Table 1 of the article by Frost et al. (2), we conclude that the two-site TSH assay showed the most interference, whereas the CPB free T4 and T3 assays showed less. In fact, for the many repetitive samples assayed, usually the correct clinical condition (hyperthyroid) was reached from the free T4 and free T3 results, although the TSH assay was consistently erroneous (2). Considering that the TSH two-site assay contained blocking agent and most probably the CPB assays did not, the data would be consistent with the effect of heterophile antibodies. We suggest that specific HAMAs would have strongly interfered with all the of the mouse assays.
Another issue is worth attention. Are all CPB assays the same? In the older radioimmunoassays, tritium or radioactive carbon caused the least alteration of the antigen, whereas radioactive iodine can modify the avidity of an antibody toward the altered antigen. Many of the newer assays, such as the solid-phase antigen-linked technique (SPALT) used for free T4 in the report by Frost et al. (2), use an antigen that is substantially modified from the endogenous species. For example, in the SPALT, T4 is bound by a protein conjugate to a solid phase. In such a case, the avidity of the assay antibody for the bound antigen may be substantially reduced, causing the interfering antibody to have more effect than would be expected for a heterophile antibody. This may have also been the case here.
We thank Dr. Frost for bringing this interesting case to our attention. We believe this further verifies the contention that an Expert Committee should ultimately define a nomenclature for bringing better order to this complicated area.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  V. Marks False-Positive Immunoassay Results: A Multicenter Survey of Erroneous Immunoassay Results from Assays of 74 Analytes in 10 Donors from 66 Laboratories in Seven Countries Clin. Chem., November 1, 2002; 48(11): 2008 - 2016. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
