False-Positive I in a Young Healthy Woman with Chest Pain J Am Board Fam Pract: first published as on 1 May 2002. Downloaded from

Erika N. Ringdahl, MD, and James J. Stevermer, MD, MSPH

Cardiac troponin I (cTnI) is frequently used to help and 2 months (4.5 ng/mL) after she was released determine whether a patient with chest pain has from the hospital. Her renal function was normal, had cardiac damage. A false-positive finding can and rheumatoid factor was not present. Her tropo- result in patients undergoing unnecessary hospital- nin T level was within normal limits. ization and potentially invasive tests. Clinicians need to be aware of the possibility of spuriously elevated troponin I levels. Discussion Several biochemical markers have been developed Case Report to aid in the diagnosis of acute coronary syndromes. has been used as a marker A 43-year-old healthy woman came to the emer- since 1957,1 but it is not specific for cardiac tissue. gency department complaining of a dull ache in her –MB band (CK-MB) is more spe- left chest, which radiated to her left shoulder. The cific than lactate dehydrogenase, but because pain was minimal but constant and unaffected by CK-MB is also found to some degree in skeletal exertion or rest. She had no associated diaphoresis, muscle, levels can be elevated with muscle disease nausea, or vomiting. The patient did not smoke and or trauma. did not have hypertension, diabetes, or a family The troponin complex is located on the thin history of cardiac disease. Her low-density lipopro- filament of striated and and regu- tein was less than 130 mg/dL 1 year earlier. Find- lates the movement of calcium between and http://www.jabfm.org/ ings of her physical examination, including temper- . The cardiac troponin is a unique form of ature, blood pressure, pulse, and respiratory rate, troponin found only in myocardial tissue. Cardiac were normal. An electrocardiogram was normal troponin has three components, T, C, and I.2 Car- and without evidence of ischemic changes. Labo- diac (cTnT) is also found in diseased or ratory studies disclosed the following values: crea- regenerating , and levels can be el- tine kinase (CK) was 67 U/L (normal 21–215 U/L), evated in patients with muscular dystrophy or poly- with a CK-MB of 0.5 ng/mL (normal 0.0–5.0 ng/ on 28 September 2021 by guest. Protected copyright. myositis.3 mL); and troponin I (AxSYM, Abbott Laboratories, cTnI is specific to cardiac tissue and is released Abbott Park, Ill) was 9.3 ng/mL (normal 0.0–0.4 into serum after myocardial necrosis.4 Enzyme- ng/mL). linked immunosorbent assay for cTnI was first Given this single laboratory abnormality, the available in 1995.5 cTnI has become a marker of patient was admitted to the family practice inpa- choice for detecting myocardial damage because of tient service. The patient was asymptomatic at ad- its specificity to cardiac tissue. Currently, there are mission and remained so throughout the hospital- two major commercial immunoassays that measure ization. A dipyridamole nuclear scan showed no cTnI levels. The Access System (Beckman Coulter, ischemia, a normal ejection fraction, and normal Fullerton, Calif) uses monoclonal mouse antibodies wall motion. A repeat troponin I measurement later as both the capture and the conjugate antibodies. in the day of admission was 9.8 ng/mL. Levels of The AxSYM system uses monoclonal mouse anti- troponin I remained elevated 1 week (3.3 ng/mL) bodies as the capture antibody and goat anti-cTnI as the conjugate antibody.1 Table 1 lists the pri- mary causes of elevated troponin I. Submitted, revised, 7 February 2002. From the Department of Family and Community Medi- Nonischemic heart disease can raise cTnI levels. cine (ENR, JJS), University of Missouri–Columbia. Address Elevated cTnI levels are found in about one half of reprint requests to Erika N. Ringdahl, MD, MA303 Medical Sciences Building, University of Missouri–Columbia, Co- younger patients with pericarditis, especially those lumbia, MO 65212. who have had a recent infection.6 Elevated cTnI

242 JABFP May–June 2002 Vol. 15 No. 3 Table 1. Causes of Elevated Troponin I. Table 2. Positive and Negative Predictive Value of the Troponin I Test at 6 Hours of Chest Pain.

Area Condition J Am Board Fam Pract: first published as on 1 May 2002. Downloaded from Patients Who Will Positive Negative 2,4 Coronary artery syndromes Acute myocardial infarction Predicted Have Positive Test Predictive Predictive Prevalence (%) Value (%) Value (%) Noncoronary heart disease Pericarditis6 1% 5.7 13.6 99.8 Myocardial injury (trauma) 2% 6.7 24.1 99.5 Myocarditis2 4% 7.9 39.4 99.0 Congestive heart failure7 6% 9.4 50.0 98.5 Arrhythmia7 8% 10.8 57.6 98.0 Pulmonary embolism7,8 20% 19.6 79.6 94.5 Renal failure9 50% 41.5 94.0 81.2 Assay interference Heterophile antibody1 80% 63.4 98.4 51.9 Rheumatoid factor10 Specialized fluids (eg albumin, Note: Sensitivity estimated at 6 hours after the onset of chest plasmin)11 pain ϭ 78%, specificity ϭ 95%.19,20 Excess fibrin12

rheumatoid factor. A rheumatoid factor blocking levels have also been reported in patients with con- agent can be used to remove this interference.16 gestive , pulmonary embolism, and In addition to clinical entities that interfere with ventricular arrhythmias.7,8 Levels of cTnI might be the immunoassays, analytical errors contribute to elevated in patients with renal insufficiency, though the number of falsely elevated cTnI results. If the for CK-MB elevated levels are more of a problem.2 specimen is centrifuged before a clot is completely Heterophilic antibodies can cause false-positive formed, the remaining fibrin can nonspecifically 12 cTnI results. The antibodies bind to the capture bind the antibody. Blood from patients who have and the conjugate antibodies, simulating cTnI. Us- a coagulopathy or who are taking anticoagulants http://www.jabfm.org/ 17 ing antibodies from two different species, as in the will be less likely to clot completely. Finally, one AxSYM system, might decrease the impact of the study showed that malalignment of a solution dis- penser produced elevated cTnI levels without reg- heterophilic antibodies.1 The increased use of istering an error by the instrument.18 Overall, the monoclonal mouse antibodies in cancer imaging Abbott AxSYM system was found to be more likely and treatment increases the chances a person will than the Bayer Immuno I Assay system to have develop heterophilic antibodies. Persons with more on 28 September 2021 by guest. Protected copyright. false-positive cTnI results.4 frequent exposure to animal (such as vet- erinarians, farmers, and pet owners) can also de- Implications for Practice velop heterophilic antibodies.13 It has been re- Clinicians can often estimate the chance that a ported that up to 40% of the general population patient has a given disease before testing for a has heterophilic antibodies,14 though it is unlikely particular disease. An a priori estimate is helpful that these elevations are clinically relevant in most when interpreting the test results. Table 2 displays cases. Recently developed enhanced assays can re- the probability of a positive test (false- and true- 15 duce the chance of interference. positives), the positive predictive value (the proba- In a similar fashion, rheumatoid factor can in- bility that a positive result represents true disease), terfere with the immunoassay. Five percent of and the negative predictive value (the probability healthy patients might have circulating rheumatoid that a negative result excludes true disease) as prev- factor, and about 1% of patients who have elevated alence is increased within a range of values. cTnI levels have this elevation purely because of The patient described here had an estimated risk 10 the rheumatoid factor. Because patients with cir- of 4% for serious coronary artery disease, based on culating rheumatoid factor can also have myocar- her history and normal electrocardiogram.21,22 Be- dial ischemia, in some circumstances it could be a cause it takes time for troponin to be released from challenge to determine whether the elevated cTnI injured , the test characteristics of levels are due to infarction or interference from the cTnI improve the longer the patient has had symp-

False-Positive Troponin I 243 toms before the test sample is drawn. The overall blood ELISA for quantification of troponin I in pa- sensitivity and specificity of cTnI, especially after tients with acute chest pain. Clin Chem 1999;45: J Am Board Fam Pract: first published as on 1 May 2002. Downloaded from 24 hours, are more than 90%. The estimated sen- 1789–96. sitivity about 6 hours from onset of symptoms is 6. Bonnefoy E, Godon P, Kirkorian G, Fatemi M, 19,20 Chevalier P, Touboul P. Serum cardiac troponin I 78%, and the specificity is 95%. Even with this and ST-segment elevation in patients with acute patient’s low risk of coronary artery disease, the pericarditis. Eur Heart J 2000;21:832–6. chance that the positive cTnI measurement indi- 7. Heeschen C, Goldmann BU, Moeller RH, Hamm cated myocardial damage was 40% (positive pre- CW. Analytical performance and clinical application dictive value). It is worth noting that (at 6 hours of a new rapid bedside assay for the detection of after pain onset) more than one half of the positive serum cardiac troponin I. Clin Chem 1998;44:1925– results do not represent chest pain until the a priori 30. probability exceeds 6%. 8. Douketis JD, Crowther MA, Stanton EB, Ginsberg JS. Elevated cardiac troponin levels in patients with submassive pulmonary embolism. Arch Intern Med Conclusion 2002;162:79–81. This case shows how a spuriously elevated troponin 9. Mockel M, Schindler R, Knorr L, et al. Prognostic I reading resulted in the hospitalization and addi- value of cardiac troponin T and I elevations in renal disease patients without acute coronary syndromes: a tional testing in this healthy 43-year-old woman. 9-month outcome analysis. Nephrol Dial Transplant When test results do not correspond to the clinical 1999;14:1489–95. picture, the possibility of a false-positive result 10. Krahn J, Parry DM, Leroux M, Dalton J. High must be considered. Testing for clinical entities, percentage of false positive cardiac troponin I results such as rheumatoid factor or heterophile antibod- in patients with rheumatoid factor. Clin Biochem ies, might be appropriate. It is also important to 1999;32:477–80. look for analytical errors by the various instru- 11. Lagneau F, Beyne P, Letteron P, Laperche T, Marty ments. In these cases, clinical pathologists can be J. Fluid therapy directly interferes with immunoassay for cardiac troponin I. Intensive Care Med 1999;25: quite helpful. In the patient described here, the

625–7. http://www.jabfm.org/ cause of the persistently elevated test result was 12. Abbott AxSYM system. Package insert: troponin-I. never determined. List No. 3C29. Abbott Park, Ill: Abbott Laborato- The accuracy of cTnI as a biochemical marker is ries, 1997. helpful in the early evaluation of patients with pos- 13. Fitzmaurice TF, Brown C, Rifai N, Wu AH, Yeo sible coronary syndromes. The likelihood of an KT. False increase of cardiac troponin I with het- incorrect result is somewhat less than with previous erophilic antibodies. Clin Chem 1998;44:2212–4. biochemical markers for cardiac ischemia. Clini- 14. Boscato LM, Stuart MC. Incidence and specificity of on 28 September 2021 by guest. Protected copyright. cians who appreciate the accuracy of this test must interference in two-site immunoassays. Clin Chem still be cognizant of the possibility of inaccurate 1986;32:1491–5. results and evaluate their patients appropriately. 15. Yeo KT, Storm CA, Li Y, et al. Performance of the enhanced Abbott AxSYM cardiac troponin I reagent in patients with heterophilic antibodies. Clin Chim References Acta 2000;292:13–23. 1. Volk AL, Hardy R, Robinson CA, Konrad RJ. False- 16. Dasgupta A, Banerjee SK, Datta P. False-positive positive cardiac troponin I results - Two case re- troponin I in the MEIA due to the presence of ports. LabMed 1999;30:610–2. rheumatoid factors in serum. Elimination of this 2. Coudrey L. The . Arch Intern Med 1998; interference by using a polyclonal antisera against 158:1173–80. rheumatoid factors. Am J Clin Pathol 1999;112: 3. Bodor GS, Survant L, Voss EM, Smith S, Porterfield 753–6. D, Apple FS. Cardiac troponin T composition in 17. Roberts WL, Calcote CB, De BK, Holmstrom V, normal and regenerating human skeletal muscle. Narlock C, Apple FS. Prevention of analytical false- Clin Chem 1997;43:476–84. positive increases of cardiac troponin I on the Stratus 4. Parry DM, Krahn J, Leroux M, Dalton J. False II analyzer. Clin Chem 1997;43:860–1. positive analytical interference of cardiac troponin I 18. Galambos C, Brink DS, Ritter D, Chung HD, Creer assays: an important consideration for method selec- MH. False-positive plasma troponin I with the tion. Clin Biochem 1999;32:667–9. AxSYM analyzer. Clin Chem 2000;46:1014–5. 5. Heeschen C, Goldmann BU, Langenbrink L, Mats- 19. Tucker JF, Collins RA, Anderson AJ, Hauser J, Kalas chuck G, Hamm CW. Evaluation of a rapid whole J, Apple FS. Early diagnostic efficiency of cardiac

244 JABFP May–June 2002 Vol. 15 No. 3 troponin I and troponin T for acute myocardial 21. Pryor DB, Harrell FE Jr, Lee KL, Califf RM, Rosati infarction. Acad Emerg Med 1997;4:13–21. RA. Estimating the likelihood of significant coronary

20. Apple FS, Maturen AJ, Mullins RE, et al. Multi- artery disease. Am J Med 1983;75:771–80. J Am Board Fam Pract: first published as on 1 May 2002. Downloaded from center clinical and analytical evaluation of the 22. Pryor DB, Shaw L, McCants CB, et al. Value of the AxSYM troponin-I immunoassay to assist in the di- history and physical in identifying patients at in- agnosis of myocardial infarction. Clin Chem 1999; creased risk for coronary artery disease. Ann Intern 45:206–12. Med 1993;118:81–90. http://www.jabfm.org/ on 28 September 2021 by guest. Protected copyright.

False-Positive Troponin I 245