Clinical Chemistry 56:5

702–707 (2010) Clinical Case Study

Persistent Increase of Cardiac Troponin I in Plasma

without Evidence of Cardiac Injury

Downloaded from https://academic.oup.com/clinchem/article/56/5/702/5622432 by Universidad de Navarra. Servicio de Bibliotecas user on 11 November 2024
Lindsay A. Legendre-Bazydlo,1 Doris M. Haverstick,1 Jamie L.W. Kennedy,2 John M. Dent,2 and
David E. Bruns1*

CASE
QUESTIONS TO CONSIDER
A 69-year-old man with diabetes mellitus type II, hy-
pertension, dyslipidemia, and prior ischemic strokes 1. Describe common pathologic reasons for increased
presented to the emergency department with com- plasma concentrations of cardiac troponins in the ab-
plaints of balance difficulties and inability to stand un- sence of an acute coronary syndrome.
assisted of 2 weeks’ duration. The patient’s home 2. What analytical interferences increase measured cTnI
medication regimen included atenolol, lisinopril, am- concentrations?
lodipine, metformin, and glipizide. He is a retired chef
and a former smoker (20 pack-years). He has 2 broth- 3. How would you investigate the abnormal cTnI results in
ers, both of whom had myocardial infarctions in their this patient?
50s. The patient’s physical examination was remark-
able for frequent premature contractions, left lower ex-
tremity weakness, and impaired coordination. His coronary intervention, and the patient remained un-
electrocardiogram revealed sinus rhythm with fre- able to stand unassisted. A neurologic evaluation in-
quent premature ventricular contractions and diffuse cluded magnetic resonance imaging, which found no
nonspecific T-wave abnormalities. evidence of an acute stroke. He was discharged with a
Results of a comprehensive metabolic chemistry diagnosis of orthostatic hypotension.
panel were within the reference intervals except for in- Three months later, the patient presented to the
creases in glucose (158 mg/dL; reference interval, emergency department with several days of balance
74 –99 mg/dL) and creatinine (1.5 mg/dL; reference in- difficulty. The initial cTnI value was increased at 0.10
terval, 0.7–1.3 mg/dL). The hemoglobin A1c value was ␮g/L, prompting admission to the cardiology service.
7.4% (reference interval, ⬍6.0%). Cardiac troponin I Over the subsequent 12 h, cTnI values for 2 additional
(cTnI)3 concentrations were increased at 0.27, 0.22, samples remained stable at 0.10 ␮g/L. Other laboratory
and 0.25 ␮g/L (Abbott Architect assay; 99th percentile, tests included a comprehensive metabolic panel and a
⬍0.03 ␮g/L) over a span of approximately 8 h. The complete blood count, with results within reference in-
patient was admitted to the cardiology service on the tervals except for a hemoglobin concentration of 12.5
basis of these abnormal results. g/dL (reference interval, 14.0 –18.0 g/dL) and a hemat-
A transthoracic echocardiogram revealed a pre- ocrit of 35.5% (reference interval, 40.0%–52.0%). The
served left ventricular systolic function with evidence patient’s symptoms were not consistent with a car-
of impaired diastolic filling. A cardiac catheterization diac etiology, and no further cardiac evaluation was
evaluation revealed an ulcerated plaque in the left an- pursued. A neurologic evaluation again found no
terior descending artery with ⬎70% stenosis, which evidence of acute stroke, and his symptoms were be-
was treated with a bare-metal stent. The presenting lieved to reflect a combination of orthostatic hypo-
symptom of balance difficulty failed to resolve after the tension, pontine gliosis, and cerebellar atrophy. The
unexplained abnormal troponin results prompted
the attending cardiologist to contact the director of
clinical chemistry.
1
Department of Pathology and 2 Division of Cardiovascular Medicine, Depart-
ment of Medicine, University of Virginia School of Medicine and Health Science DISCUSSION
Center, Charlottesville, VA.
* Address correspondence to this author at: Department of Pathology, University
of Virginia Health Science Center, Charlottesville, VA 22908. Fax 434-924-2574;
Cardiac troponins play a central role in diagnosis and
e-mail DEB6J@virginia.edu. risk stratification in acute coronary syndromes (1 ), but
Received October 6, 2009; accepted December 8, 2009. troponins are recognized as markers of cardiac myo-
DOI: 10.1373/clinchem.2009.138164
3
Nonstandard abbreviations: cTnI, cardiac troponin I; HAMA, human antimouse cyte injury, not of the etiology of injury. A wide range of
antibody; CK-MB, creatine kinase isoenzyme MB; cTnT, cardiac troponin T. clinical conditions has been associated with increased

702
 Clinical Case Study

tient’s plasma. We reasoned that such a complex

Table 1. Noncoronary causes of increased would delay the clearance of cTnI from the circula-
cardiac troponin.a tion and thus produce persistently increased tropo-

Downloaded from https://academic.oup.com/clinchem/article/56/5/702/5622432 by Universidad de Navarra. Servicio de Bibliotecas user on 11 November 2024
nin concentrations (3 ). We therefore treated the
Cardiac contusion Pulmonary embolism patient’s plasma either with protein A bound to
Heart failure Renal failure Sepharose to deplete the sample of IgG, along with
Aortic dissection Myocarditis any cTnI complexed with IgG, or with an equal vol-
Aortic valve disease Acute stroke ume of buffer. Protein A decreased the cTnI concen-
Hypertrophic cardiomyopathy Subarachnoid hemorrhage tration from 0.10 ␮g/L to undetectable (⬍0.02
Tachy- or bradyarrhythmias Drug toxicity ␮g/L), whereas buffer produced only the decrease
Heart block Critical illness
expected from dilution, to 0.06 ␮g/L (the between-
day total imprecision of the assay at 0.04 – 0.06 ␮g/L
Apical ballooning syndrome Burns
is ⬍10%). In samples from 5 other patients, the cTnI
Rhabdomyolysis Extreme exertion
concentrations obtained for the protein A–treated
Infiltrative diseases (e.g., sample matched those of the dilutional control. Un-
amyloidosis, sarcoidosis)
fortunately, no plasma from the index patient’s
a
Modified from Thygesen et al. (1 ). initial hospitalization was available for a similar
analysis.

MACROCOMPLEXES OF TROPONIN: CHEMISTRY, PREVALENCE,

troponin values (1 ), most or all of which have been AND CLINICAL FINDINGS
shown to entail cardiac injury. Thus, it is necessary to Naturally occurring immunoglobulin– cardiac tropo-
consider these conditions when investigating unex- nin complexes in plasma were first recognized only re-
pected increases in cardiac troponins. cently. In 1996, Bohner et al. suspected the presence of
In the patient described above, the rise and fall in such a complex in a patient after elective coronary ar-
troponin concentrations that are characteristic of acute tery bypass graft surgery. The patient had undetectable
myocardial infarction was not seen on either admis- cTnI despite increased concentrations of creatine ki-
sion. This finding pointed toward other etiologies for nase isoenzyme MB (CK-MB) and cTnT (4 ). The au-
the increased troponin values (1 ). The clinical find- thors added increasing amounts of cTnI to the sample,
ings and other laboratory information excluded up to 38.5 ␮g/L, and found that cTnI continued to be
most if not all of these other causes of increased tro- undetectable. The authors concluded that an antibody
ponin (Table 1). that binds cTnI was complexing cTnI, rendering it un-
able to bind to the antibodies in the immunoassay
LABORATORY INVESTIGATION OF INCREASED cTnI reagent.
Two important interferences reported to produce In 2002, an immunoglobulin– cTnI complex was
false-positive results in cardiac troponin immunoas- identified as the source of increased cTnI concentra-
says are fibrin clots and heterophile antibodies, such tions (3 ). The patient described in the report had in-
as human antimouse antibodies (HAMAs) or rheu- creasing cTnI concentrations over several months
matoid factor (2 ). No fibrin clots were seen in the without clinical evidence of myocardial infarction.
patient’s sample and increased cTnI values were seen Treatment of a sample with antihuman IgG antiserum
in multiple samples and upon their reanalysis, pro- removed the cTnI. The authors hypothesized that the
viding presumptive evidence that fibrin clots were complex extended the normally brief half-life of the
not involved. This patient’s plasma was evaluated for cTnI and allowed accumulation of cTnI in the blood to
HAMAs by measuring cTnI before and after treatment measurable concentrations.
with a heterophile blocking tube from Scantibodies Recent reports indicate that IgGs that bind cardiac
troponins are common in healthy blood donors. IgG
Laboratory. The plasma cTnI concentration remained
autoantibodies against cTnI, cTnT, or both cTnI and
unchanged, suggesting that HAMAs were not a source
cTnT were found in 12.7% (n ⫽ 750), 9.9% (n ⫽ 467),
of analytical interference in the immunoassay.
and 1.7% (n ⫽ 345) of the donors, respectively (5–7 ).
In a study of patients with cardiomyopathy (8 ), anti-
INVESTIGATION OF POSSIBLE IMMUNOGLOBULIN-cTnI bodies to cTnT (or cTnI) were present in 1.7% (or
COMPLEX 7.7%) of patients with dilated cardiomyopathy (n ⫽
We suspected the presence of a macrocomplex of 272) and in 0.5% (or 9.2%) of 185 patients with isch-
cTnI with an immunoglobulin molecule in the pa- emic cardiomyopathy. The lower prevalence than in

Clinical Chemistry 56:5 (2010) 703

Clinical Case Study

Circulating immunoglobulins complexed with

enzymes and other proteins that are measured in the
diagnostic laboratory have been recognized for

Downloaded from https://academic.oup.com/clinchem/article/56/5/702/5622432 by Universidad de Navarra. Servicio de Bibliotecas user on 11 November 2024
many years, back at least to the recognition of mac-
roamylase (10 ). None of the other macrocomplexes
are likely to have a more serious immediate clinical
impact than an immunoglobulin– cardiac troponin
complex (or macrotroponin). Even small increases
in cardiac troponin in plasma or serum will be iden-
tified as abnormal by the sensitive cardiac troponin
assays currently available (and even more sensitive
assays are forthcoming). Such results carry the po-
tential to lead to unnecessary interventions, as seen
in this case. Results of increased cardiac troponin
concentrations are expected with any cardiac tropo-
nin assay because the increased cardiac troponin
concentration measured does not represent an inter-
Fig. 1. Time course of cardiac troponin in the pres-
ference, but rather an analytically correct result that
ence of antitroponin antibodies in patients studied
is nonetheless misleading.
by Pettersson et al. (9 ) (circles) and in the current
patient (ⴱ). In light of the high prevalence of immunoglobulin–
cardiac troponin complexes and the critical impor-
Indicated are patients with antibodies present at the
time of admission (F) and patients in whom antibodies
tance of cardiac troponin testing, it is necessary to be
appeared by the time of the 1-week sample (E). The aware of the possibility of these antibodies in pa-
shaded area shows the 10th–90th percentile interval of tients whose clinical presentations do not match
the cTnI concentration for the patients of Pettersson et their increased cardiac troponin concentrations. A
al who did not have antibody at either time point. The second blood sample obtained a few hours later
horizontal line at 0.005 ␮g/L represents the analytical should provide clarification. With acute myocardial
detection limit for the assay used by Pettersson et al., injury, cardiac troponin will continue to increase,
and the horizontal line at 0.02 ␮g/L represents the whereas it will be stable with an immunoglobulin–
lowest concentration at which their patients were con- cardiac troponin complex, as it was on both admis-
sidered cTnI positive. The assay used by Pettersson et sions of this patient. Another approach is parallel
al., the second-generation Innotrac Aio!™ troponin I analysis of CK-MB. A rise and fall in the CK-MB
assay, is different from the method used in our labora- concentration suggests an acute myocardial infarc-
tory (Abbott Architect). The concentrations measured tion, but the test is less diagnostically sensitive and
with the 2 assays are likely to be different, but the specific than troponin assays.
slopes of the plotted lines can be compared because the One approach to identify an immunoglobulin–
vertical axis is logarithmic. cardiac troponin complex is to measure cardiac tropo-
nin before and after treatment of the plasma or serum
with protein A, which removes IgG and thus the com-
the healthy blood donors is thought to reflect a dif- plex containing the cardiac troponin and IgG. Hetero-
ference in the types of assays used to identify auto- phile blocking tubes, as were used in this case, provide
antibodies (6 ). a simple approach to exclude the possibility that pro-
Pettersson et al. (9 ) showed that the presence of tein A is removing an interfering HAMA. Other ap-
antibodies to cTnI slowed the apparent clearance of proaches for identifying the presence of macrocom-
cTnI from the circulation (Fig. 1). The plasma cTnI plexes include electrophoresis and immunofixation
concentrations in the patient described here are indi- (10 ).
cated by the asterisks and dashed line added to the fig-
ure of Pettersson et al. (Fig. 1). As can be seen, the rate
of clearance of cTnI from the circulation of this patient CONCLUSION
is like that of the patients of Pettersson et al., who had
anti-cTnI antibodies, and is much slower than that We conclude that this patient’s cTnI most likely circulated
seen in the patients without antibodies to cTnI (shaded as a complex with IgG. Recognition of such complexes is
area). critical to avoid erroneous diagnosis of cardiac injury.

704 Clinical Chemistry 56:5 (2010)

 Clinical Case Study

quirements: (a) significant contributions to the conception and design,

POINTS TO REMEMBER acquisition of data, or analysis and interpretation of data; (b) drafting
or revising the article for intellectual content; and (c) final approval of

Downloaded from https://academic.oup.com/clinchem/article/56/5/702/5622432 by Universidad de Navarra. Servicio de Bibliotecas user on 11 November 2024
• Plasma concentrations of cardiac troponins are increased the published article.
in a large number of conditions that injure cardiac muscle Authors’ Disclosures of Potential Conflicts of Interest: No authors
(Table 1). declared any potential conflicts of interest.

• Autoantibodies to cTnI or cTnT have been identified in Role of Sponsor: The funding organizations played no role in the
roughly 10% of the healthy population and form com- design of study, choice of enrolled patients, review and interpretation
plexes with troponin. of data, or preparation or approval of manuscript.

• Antibodies to cardiac troponin can produce a result of

either a decreased troponin concentration (by blocking References
an epitope recognized by a reagent antibody) or an
1. Thygesen K, Alpert JS, White HD, Jaffe AS, Apple FS, Galvani M, et al.
increased troponin value that appears to reflect the Universal definition of myocardial infarction. Circulation 2007;116:
presence of increased cardiac troponin concentration. 2634 –53.
2. McNeil A. The trouble with troponin. Heart Lung Circ 2007;16(Suppl 3):
• The apparent slow clearance of immunoglobulin– cardiac S13– 6.
troponin complexes from the circulation leads to persistent 3. Plebani M, Mion M, Altinier S, Girotto MA, Baldo G, Zaninotto M.
increases in measured cardiac troponin, especially after False-positive troponin I attributed to a macrocomplex. Clin Chem 2002;
cardiac injury. 48:677–9.
4. Bohner J, von Pape KW, Hannes W, Stegmann T. False-negative immunoas-
• Measurement of cTnT in a patient with an say results for cardiac troponin I probably due to circulating troponin I
immunoglobulin-cTnI complex may or may not be use- autoantibodies. Clin Chem 1996;42:2046.
5. Adamczyk M, Brashear RJ, Mattingly PG. Circulating cardiac troponin-I autoan-
ful, because patients with such complexes may also tibodies in human plasma and serum. Ann N Y Acad Sci 2009;1173:67–74.
have macrocomplexes involving cTnT; measurement of 6. Adamczyk M, Brashear RJ, Mattingly PG. Prevalence of autoantibodies to
CK-MB may thus be useful. cardiac troponin T in healthy blood donors. Clin Chem 2009;55:1592–3.
7. Adamczyk M, Brashear RJ, Mattingly PG. Coprevalence of autoantibodies
• The presence of immunoglobulin– cardiac troponin com- to cardiac troponin I and T in normal blood donors. Clin Chem 2010;56:
plexes can be inferred from exclusion of a HAMA interfer- 676 – 677.
ence, combined with demonstration of cardiac troponin 8. Leuschner F, Li J, Goser S, Reinhardt L, Ottl R, Bride P, et al. Absence of
auto-antibodies against cardiac troponin I predicts improvement of left
removal by protein A, protein G, or antihuman immuno- ventricular function after acute myocardial infarction. Eur Heart J 2008;29:
globulin antiserum. 1949 –55.
9. Pettersson K, Eriksson S, Wittfooth S, Engstrom E, Nieminen M, Sinisalo J.
Autoantibodies to cardiac troponin associate with higher initial concentra-
tions and longer release of troponin I in acute coronary syndrome patients.
Clin Chem 2009;55:938 – 45.
Author Contributions: All authors confirmed they have contributed to 10. Remaley AT, Wilding P. Macroenzymes: biochemical characterization, clinical
the intellectual content of this paper and have met the following 3 re- significance, and laboratory detection. Clin Chem 1989;35:2261–70.

Commentary
Mitchell G. Scott

In the last decade, sensitive and specific immunoassays stratification and treatments. The required sensitivity of
for cardiac troponins have revolutionized the diagnosis these assays, however, has made the occurrence of false-
and management of cardiac disease. The absolute spec- positive cardiac troponin results an infrequent but very
ificity of cardiac troponins for cardiac tissue and the real laboratory problem. Any immunoassay has non-
sensitivity of the immunoassays have redefined risk specific background “noise,” and the signal-to-noise ratio
defines the sensitivity of the assay. In this regard cardiac
troponin assays are particularly challenged because
they must distinguish very small, but clinically signifi-
Department of Pathology and Immunology, Division of Laboratory and Genomic cant, concentrations of cardiac troponin from “noise.”
Medicine, Washington University School of Medicine, St. Louis, MO.
Address correspondence to the author at: Department of Pathology and Immu-
Few immunoassays are asked to distinguish between
nology, Division of Laboratory and Genomic Medicine, Washington University “any antigen” and “none” (drug screens and human
School of Medicine, 660 S. Euclid Ave., Box 8118, St. Louis, MO 63110. E-mail chorionic gonadotropin come to mind). Thus, unlike
mscott@labmed.wustl.edu.
Received February 22, 2010; accepted March 1, 2010. most immunoassays, additional “noise” in a cardiac tro-
DOI: 10.1373/clinchem.2009.142182 ponin assay will be considered clinically important. Sim-

Clinical Chemistry 56:5 (2010) 705