Immunopathology / MULTIPLEX ANALYSIS OF HETEROPHIL ANTIBODIES

We hypothesized that heterophil antibodies reactive with animal proteins used in blot preparation caused nonspecific staining (NSS) on HIV Western blot (WB) studies, causing indeterminate results. We analyzed samples showing NSS on HIV WB using a multiplexed immunoassay to simultaneously measure IgG antibodies to animal IgG (bovine, goat, sheep, mouse) and bovine serum albumin. Heterophil antibodies reactive with IgG from several animal species were detected in 23 (49%) of 47 samples showing NSS on HIV WB; 15 positive samples demonstrated antibodies to all 5 antigens. Similar IgG heterophil antibodies were detected in only 2 (8%) of 24 control samples. Of the HIV WB samples with a positive HIV-1 enzymelinked immunosorbent assay (ELISA) result at the time of WB testing (11/47), heterophil antibodies were found in 8 (73%) of 11. Preabsorption with bovine, goat, and sheep IgG removed heterophil antibodies detected by the multiplexed assay and, in some cases, eliminated reactivity in ELISA and WB testing. Heterophil antibodies are associated with indeterminate HIV immunoassay results and are an important cause of false-positive HIV ELISA results. Multiplexed immunoassays provide a powerful tool for screening patients for heterophil antibodies and resolving possible false-positive results. Heterophil antibodies are widely recognized as a cause of interference in immunoassays. Heterophil antibodies are thought to demonstrate multispecificity by having multiple binding sites or by having a single binding site that can recognize a number of antigens with similar structures.1 Heterophil antibodies can create a false-positive result in an immunoassay by binding to the capture and signal antibodies in the assay and mimicking the behavior of the analyte of interest. Heterophil interference has been observed in many immunoassays, including creatine kinase MB,2 human chorionic gonadotropin,3 CA-125,4 thyroid stimulating hormone,5 and others, but has not been well characterized in HIV immunoassays.6 False-positive HIV enzyme-linked immunosorbent assay (ELISA) results can occur with serious medical, social, and legal consequences,7 and false-positive results are thought to account for as many as 4.6% of positive Western blots in a low-risk screening setting.8 The HIV Western blot can yield indeterminate results because of nonviral bands or crossreacting autoantibodies.9,10 Nonspecific staining (NSS) is not addressed specifically in the criteria for interpreting HIV Western blot assays.11 Since NSS can obscure a viral band on the blot, many laboratories will not interpret a blot with NSS as negative but instead report a result of indeterminate with a recommendation for retesting. We hypothesized that the majority of NSS observed on HIV Western blots was due to heterophil antibodies, since animal proteins, including immunoglobulins, are used in the manufacture of the blots. Furthermore, any sample with heterophil antibodies that caused NSS on a Western blot also might show a false-positive result on an ELISA. To detect heterophil antibodies in samples showing NSS that were potentially reactive with many different antigens, Immunopathology / ORIGINAL ARTICLE Am J Clin Pathol 2001;115:764-769 765 © American Society of Clinical Pathologists we used microsphere-based multiplexed immunoassays on the Luminex platform (Luminex, Austin, TX). The Luminex analyzer performs flow cytometric analysis of microspherebased immunoassays with simultaneous measurement of multiple analytes.12 We used a panel of animal serum proteins commonly used in immunoassays, including polyclonal bovine, goat, sheep, and mouse IgG and bovine serum albumin (BSA). In the present study, we screened for a subset of heterophil antibodies likely to cause interference in immunoassays by simultaneously measuring human IgG reactive with these antigens. We report that human serum samples producing NSS on HIV Western blots often contain heterophil IgG antibodies reacting with a variety of serum proteins from other animal phyla. Materials and Methods Patients and Serum Samples During a 5-month period (July 1, 1999-November 30, 1999), 2,518 consecutive samples were submitted to ARUP Laboratories, Salt Lake City, UT, for HIV Western blot testing. Samples were tested according to the manufacturer’s protocol (Sanofi Diagnostics Pasteur, Redmond, WA). According to the manufacturer’s recommendations, “Strips which are obscured by the development of dark blotches or marks should not be interpreted.”13 All blots showing NSS, as determined by the technologist at the time of testing, were given an interpretation of indeterminate for HIV-1 infection in the final laboratory report. Serum from any blot that showed NSS was selected for heterophil antibody analysis. All samples were tested simultaneously with an HIV-1 ELISA assay (Organon Teknika, Durham, NC). Serum was stored at –20°C. When available, patient information was obtained from the submitting physician.