Elevated Troponin Results in a Patient on Immunotherapy for Crohn's Disease

A 72-year-old woman presented with left lower abdominal pain for 1 day, and an episode of chest pain 3 days prior to the presentation. Cardiac troponin (cTn) I was 9.1 ng/mL. Coronary angiogram showed nonobstructive coronary artery disease, with 30%–40% obstruction in mid-right coronary artery; all other coronary arteries were angiographically normal. The patient was discharged when her cTnI decreased to 6.7 ng/mL. Ten days later, the patient was instructed to go to the Emergency Department (ED) by her primary care physician because her cTnI increased to 8.9 ng/mL. After admission, the patient’s cTnI results showed a confusing pattern that fluctuated between positive and negative multiple times. Creatine kinase MB isoenzyme (CK-MB) obtained with an immunoassay on Siemens Vista (Siemens Healthcare Diagnostics) was >300 ng/mL [upper reference limit (URL), 5.5 ng/mL]. Results for AST, creatine kinase, lactate dehydrogenase, C-reactive protein, and BNP were all normal. Her sedimentation rate was elevated at 91 mm/h (URL, 25 mm/h). The clinical team raised a question about the fluctuations of her cTnI results and wanted to know which result was correct. For cTnI testing, we used to use Siemens Vista (Vista) in the main laboratory and Siemens Stratus (Stratus) in the ED laboratory, with <0.1 ng/mL as the 99th percentile cutoff for both. Instrument-to-instrument comparison was performed every 6 months using patient specimens with acceptable results. Upon investigation, we found that all positive cTnI results were obtained with Vista, and negative cTnI results were obtained with Stratus. The chronological order of this patient’s cTnI results were: 8.9 ng/mL (Vista) → <0.04 ng/mL (Stratus) → 5.1 ng/mL (Vista) → 7.4 ng/mL(Vista) → <0.04 ng/mL (Stratus) → 6.1 ng/mL (Vista) → 5.8 ng/mL (Vista). In order to determine if the fluctuation was caused by interference, we performed dilution studies on one of her samples with a cTnI of 8.4 ng/mL and obtained a result of 0.75 ng/mL and 0.11 ng/mL on the Vista when the sample was diluted three and six times, respectively. These results showed that the cTnI values did not change proportionally with dilution, suggesting the presence of an interference because true elevated results will decrease in proportion to dilutions, whereas false elevation due to interference will not. The CK-MB result also showed a nonproportional change from >300 ng/mL to 11.8 ng/mL and 2.5 ng/mL upon five and ten times dilution. In order to determine if the interference was caused by heterophilic antibodies such as human anti-animal (HA) or human anti-mouse antibodies (HAMA), we treated one of her samples with heterophilic antibody blockers HBT (Scantibodies, Santee, CA) and the result changed from 7.4 ng/mL (neat) to 5.49 ng/mL, indicating the interference being HAs because HBT treatment can partially or completely neutralize the effects of HAs produing different results. To prove the validity of HBT treatment, we also treated a control sample with HBT and the cTnI result remained unchanged. In order to eliminate the interference, we treated one of the patient's samples with a cTnI result of 3.73 ng/mL with HBT one, two, three, and four times, Vista’s cTnI result decreased to 2.68 ng/mL, 1.48 ng/mL, 0.59 ng/mL and 0.02 ng/mL, respectively (Table 1). On the contrary, the Stratus cTnI result changed from negative to 0.58 ng/mL and 0.66 ng/mL after the sample was treated with HBT once and twice, respectively. In addition, the CK-MB result by electrophoresis was undetectable, suggesting the Vista’s CK-MB assay was also affected by the interference. Based on the results of these studies, we invalided all the troponin and CK-MB results on this patient. The patient was then discharged home because she was asymptomatic, had stable vital signs, and the unremarkable ECG and echocardiogram, and cardiac MRI showing no evidence of myocardial damage from ischemia, inflammatory state, or infiltrative disease. Troponin results after heterophilic antibody blocker treatment. HBT: heterophilic antibody blocker. 1×, 2×, 3×, and 4×: sample treated with HBT once; twice, three times, and four times. Troponin results after heterophilic antibody blocker treatment. HBT: heterophilic antibody blocker. 1×, 2×, 3×, and 4×: sample treated with HBT once; twice, three times, and four times. A review of patient history showed a 30 years history of Crohn's disease. She was initially treated with mesalamine, methotrexate, and prednisone until she subsequently had a relapsing and remitting disease course. About 20 years ago, infliximab 10 mg/kg every 8 weeks was initiated that improved her symptoms tremendously, being able to avoid hospitalization for about 5 years and experienced only occasional flares characterized by loose bowel movements and cramping. Later on, the patient developed an allergic reaction to infliximab infusion with rash, even when she was premedicated with prednisone. Anti-infliximab antibodies repeatedly tested negative. After she developed lung adenocarcinoma about a year ago, she was switched to vedolizumab with budesonide and has had good control of symptoms with occasional flares. The accuracy of a cTn assay is of utmost importance because a false positive troponin result may lead to unnecessary procedures or interventions, and a false negative result may lead to missed diagnosis or delayed treatments. Immunoassays utilizing one or more mouse monoclonal antibodies (mAbs) specific for different epitopes of cTn are currently used to measure cTn in patient samples. HAMAs or HAs can react with assay antibodies by either bridging the capture and the detection antibodies to produce falsely increased results or blocking one or both antibodies’ binding sites to produce falsely decreased results (1,). In order to minimize HAMA’s interference, manufacturers generally add heterophilic antibody blockers to their reagents of immunoassays, which greatly decrease, but may not always be sufficient to eliminate the interference of HAMAs or HAs as multiple cases of HAMA’s or HAs’ interference with cTn tests being reported (2–7). Although the prevalence of HAMA is unknown, several small studies estimated <1% to 80% depending on the assay and the number of patients involved in each study (1,). Patients receiving immunotherapies or imaging procedures utilizing mAbs are prone to develop HAMAs or antibodies to the particular mAbs used in the procedure or therapy. Our patient had been treated with infliximab for more than 20 years. Infliximab is a mAb to tumor necrosis factor alpha (TNFα), which can neutralize the biological activity of TNFα. Infliximab is a chimeric IgG1κ consisted of human constant and murine variable regions. The patient is currently on Vedolizumab, which is a humanized monoclonal antibody that specifically binds to the α4β7 integrin and blocks the interaction of α4β7 integrin with mucosal addressin cell adhesion molecule-1, thereby preventing the adhesion of lymphocytes to the vascular endothelium and reducing intestinal inflammation. It is less likely for vedolizumab to cause the production of HAMA because it is a humanized mAb. The partially humanized infliximab, on the other hand, is the most likely cause for HAMA production in our patient. Janssen, the manufacturer of Remicade (infliximab), reported the presence of anti-infliximab antibodies in about 52% patients receiving Remicade. Anti-infliximab alone may not cause interference with troponin test because one single-center study examined troponin results in 18 patients receiving infliximab therapies and found no incidence of false positive troponin values (8) probably none of the patients included in that study developed HAMA. Our patient did not develop anti-infliximab antibodies, instead, she developed HAMA, which interfered with the Vista’s troponin and CK-MB tests. Further, the concentration of the HAMA present in her sample was so high that it exhibited HAMA excess effect on immunoassay, which will be discussed here. At first glance, it appeared that the HAMA present in our patient’s samples interfered with the cTnI assay on the Vista, but not on the Stratus even though both the Stratus and Vista use the same pair of mAbs in their cTnI assays with the capture antibody specific for the amino acid (aa) 27-32 epitope and the detection antibody specific for the aa 41-56 epitope of cTnI (9). It is, therefore, reasonable to assume that the different effects of the HAMA on the Vista and Stratus cTnI assays are not caused by different interactions between HAMA and the reagent antibodies, instead caused by the differences in the two methods. The cTnI assay on the Vista uses a homogeneous luminescent oxygen channeling immunoassay (LOCI) that mixes 20 µL of sample with 153 µL of reagents (the sensi-beads and the chemi-beads) with a final volume of 173 µL. Therefore, the final concentration of HAMA in the LOCI reaction is 0.116c (20 µLxc/173 µL), where c denotes the original concentration of HAMA. When HAMA cross-react with reagent antibodies bringing the sensi-bead and the chemi-bead into a closed proximity, a false cTnI result will be produced. The Stratus cTnI assay, on the other hand, uses a solid phase radial partition immunoassay (RIPA) method that utilizes a piece of glass fiber to retain the capture antibodies as well as the immune-complexes. When 50 µL of dendrimer linked mAbs is added to the center of a glass fiber paper, the glass fiber will retain dendrimer-mAbs and water solution will flow radiantly. Then, 90 µL of a sample is added to the central area. cTnI or HAMA will bind to the antibody and the rest will flow away on the paper. After 50 µL of the detection antibodies is added to the center of the paper, immune complexes between antibody-cTnI-antibody or antibody-HAMA-antibody will be formed in the central area of the glass fiber paper. Finally, 90 µL of substrate and wash solution is added to the filter to initiate the enzymatic reaction. Anything that is not bound to the glass fiber will be washed away. Assuming the final reaction volume in the wet area of the glass fiber is 50 µL, the final concentration of cTnI or HAMA would be 1.8c (90µLxc/50 µL). Therefore, the amount of HAMAs readily available to the capture antibodies in the RIPA reaction is at least 16 times more than that in the LOCI reaction (0.116c). The other possibility is that the Stratus cTnI reagents does not have the same amount of heterophilic antibody blocker as that included in the Vista’s assay. In any case, the amount of HAMA in the RIPA reaction is much higher than that in the LOCI reaction that it can saturate all of the reagent antibodies’ binding sites to prevent one HAMA from binding to both the capture and the detection antibodies. This effect of HAMA excess is very similar to that found in Hook effect, where antigen excess prevents the formation of a sandwich between the capture antibody, the antigen, and the detection antibody. Figure 1 illustrates the effect of HAMA excess on sandwich immunoassay. In Fig. 1b, when the amount of HAMA is in excess, all of the reagent antibody binding sites are neutralized by HAMA. Cross-reaction between HAMA and capture and detection antibodies cannot happen. In Fig. 1c, when a significant amount of HAMA are neutralized by HAMA blockers, some antibody binding sites are freed up so that the remaining HAMA can cross-react with the capture and the detection antibodies to generate a false signal. In Fig. 1d, when all of the HAMAs are neutralized by HAMA blockers, cross-reaction between HAMA and reagent antibodies cannot take place. No signal is generated. HAMA excess can be used to explain our observations. As shown in Table 1, the original cTnI result from the Stratus was negative because all of the antibody binding sites were covered by HAMA. After two treatments with HBT, the amount of HAMA was significantly reduced, leaving some antibody binding sites unoccupied. The remaining HAMA can cross-react with the capture and detection antibodies to produce a result of 0.66 ng/mL. Schematics of the effect of HAMA excess. a) HAMA cross-react with both the capture and detection antibodies to generate a false signal in the absence of troponin. b) In HAMA excess, HAMAs can bind and saturate all of the binding sites on reagent antibodies, and no false signal can be generated. This scenario can explain the negative troponin result obtained in neat samples with Stratus. c) Insufficient amount of HAMA blocker can neutralize a fraction of HAMAs, exposing the antibody binding sites to the remaining HAMAs, which can cross-react with both capture and detection antibodies to generate a false signal. This scenario can explain the false positive troponin result of 0.66 ng/mL when the sample was treated with HBT blocker twice. d) Sufficient amount of HAMA blocker can neutralize all of the HAMAs. No signal is produced. To the best of our knowledge, we were the first to report and explain the effect of HAMA excess on immunoassays. HAMA excess can be a serious problem for cTnI or other immunoassay tests because it can cause a false negative result in a patient with a relatively low cTnI result who may have a true heart attack. As a growing number of immunotherapies become available for a wider range of medical conditions, we may encounter more cases of HAMA interference with laboratory tests. Clinicians caring for patients who are on these medications should remain alert to the possibility that HAs or HAMAs present in their samples can cause paradoxical laboratory results or confounding patient presentations. In these unusual situations, communication between clinicians and laboratorians is crucial to the investigation of the problem leading to a resolution. Therapies using monoclonal antibodies can cause the development of human anti-mouse antibodies (HAMA). HAMA interference with immunoassays can cause either false positive or false negative results. Serial dilutions plus treatment with heterophilic antibody blocker can help to identify HAMA interference. In case of HAMA excess, decreased or even negative results can be produced by immunoassays. Nonstandard Abbreviations: cTn, cardiac troponin; URL, upper reference limit; ED, Emergency Department; CK-MB, creatine kinase MB isoenzyme; HA, human anti-animal antibodies; mAbs, monoclonal antibodies; HAMA, human anti-mouse antibodies; TNFα, tumor necrosis factor alpha; LOCI, luminescent oxygen channeling immunoassay; RIPA, radial partition immunoassay. Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Authors’ Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.