Preanalytical mysteries Description of interference in the measurement of troponin T by a high-sensitivity method Miguel Aliste-Fernández*1, Gemma Sole-Enrech1, Ruth Cano-Corres1, Silvia Teodoro-Marin2, Eugenio Berlanga-Escalera1 1ClinicalLaboratory, Biochemistry Department, Parc Taulí Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Sabadell, Spain 2Pediatric Medicine, Parc Taulí Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Sabadell, Spain *Corresponding author: maliste@tauli.cat Abstract Introduction: Measurement of high-sensitivity troponin T (hs-TnT) has become an essential step in the diagnosis of acute myocardial infarction. This high-sensitivity method allows quantifying the concentration of troponin T in blood of healthy subjects with a lower inaccuracy compared to previous reagent generations. However, the presence of certain compounds in the sample may interfere with the result. We present a patient who had repeatedly high concentrations of hs-TnT in the serum sample that did not agreed with the signs and symptoms. In addition, ultrasensitive tro- ponin I concentration was undetectable. Materials and methods: To investigate the presence of an interfering compound, different analysis were carried out. In order to discard macro complexes in the sample, the serum was precipitated with polyethylene glycol. In addition, the serum was incubated with Scantibodies Heterophilic Blocking Tube, which can block heterophilic antibodies. Finally, a size exclusion chromatography of the sample was performed by the manufacturer. What happened: The interfering substance was allocated into fractions with proteins of 150kDa, corresponding to high molecular weight proteins like immunoglobulin G (IgG). This compound was responsible for the falsely elevated hs-TnT results and it affected only the high-sensitivity methods. Main lesson: The detected interfering compound was probably an IgG. This type of interference must be kept in mind in front of discordant results, even if they are extremely rare. Therefore, interdisciplinary cooperation between clinicians, laboratory and manufacturer is essential. Keywords: high-sensitivity troponin T; interference; case report Received: December 18, 2018 Accepted: March 26, 2019 Introduction The use of the new high-sensitivity troponin (hs- acute event. Since other cardiac and no cardiac pa- Tn) tests has changed the approach to the diagno- thologies can slightly increase the troponin con- sis of the acute myocardial infarction (AMI). These centration the recent measurement improvement methods of high sensitivity are characterized by reduces the specificity for the diagnosis of AMI (2). the improvement of the detection limit, which is Moreover, interfering compounds can alter the defined as the lowest concentration of analyte de- measurement of troponin concentration in the tectable by analytical method (1). As a result, it is sample, obtaining falsely increased or decreased possible to measure the concentration of troponin results. Due to the complexity of the antigen-anti- in blood of healthy subjects with higher sensitivity body interaction, immunoassays are particularly and lower inaccuracy compared to previous rea- susceptible to these kinds of interferences, but op- gent generations, which in turn allows a better timizations in hs-Tn assays designs led to reported definition of the 99th percentile. It is useful in par- outlier rates of < 0.6% (3). ticular to detect earlier troponin changes after the https://doi.org/10.11613/BM.2019.021001 Biochem Med (Zagreb) 2019;29(2):021001 1 Aliste-Fernández M. et al. Interference in the measurement of troponin T Case description obtained during this period was 3.0% for Cobas e601 and 3.5% for Cobas e801. A 12-years-old boy was admitted to the hospital with episodes of chest pain triggered in effort and The first two serum samples from the patient were repose. He was subjected to an electrocardiogram only processed on the Cobas e801 analyser, while and echocardiogram with normal results. The labo­ the third one was processed on both analysers ratory tests showed an elevation of high-sensiti­ and sent to an external laboratory to measure the vity troponin T (hs-TnT) of 51 ng/L (99th percentile concentration of high-sensitivity troponin I (hs-TnI) = 14 ng/L), creatine kinase (CK) within the refer- on the Architect-i System platform (Abbott Diag- ence range and preserved renal function. Exercise nostics, Chicago, USA), using the hs-TnI STAT rea- test was within normality range and thus acute gent. However, it is important to take into account process by ischemia, pericarditis or myocarditis that the interference susceptibility might vary was discarded. among the different measurement systems (e.g. even among various hs-TnI assays) (5). Blood tests were repeated 4 months later and an increase of hs-TnT up to 98 ng/L was found, while Referring to the literature on the study of analyti- CK and other parameters were within normal cal interferences, we followed the recommenda- ranges. A second echocardiography was per- tions described by Ward et al. and other publica- formed and no pathological alterations were de- tions with similar cases, to identify potential inter- tected. fering factors such as fibrin micro clots, haemolysis or human anti-mouse antibodies (3,6). Firstly, we A new blood sample was collected one month la­ ruled out pre-analytical problems and we per- ter and high concentration of hs-TnT (52 ng/L) was formed a serial dilution of the sample using the observed again. Given the absence of compatible “Diluent MultiAssay” (Roche Diagnostics GmbH, symptoms, a falsely positive result caused by inter- Mannheim, Germany). The concentration of hs-TnT ference was suspected. was measured on the Cobas e801 automated plat- form. We also performed a polyethyleneglycol Laboratory analyses (PEG) precipitation (Merck-Schuchardt, Hohenb- runn, Germany) of the serum to rule out the pre­ All patient samples were drawn in gel separator sence of macrocomplexes, by mixing 300 µL of the tubes without anticoagulant (Vacuette Z serum serum sample with the same volume of a solution sep clot activator, Greiner Bio-One, Kremsmünster, of PEG 6000 at 25%. Then, the mixture was centri- Austria). Tubes were left at room temperature for fuged at 1500xg for 20 minutes and supernatant 15 minutes approximately to allow blood coagula- was assayed. tion and then centrifuged at 2030xg for 10 minu­ Since mouse antibodies are also used in the meas- tes. Both analysers for measuring the concentra- urement procedure, other analytes which could tion of hs-TnT available in our laboratory, Cobas potentially be influenced by the same interference e801 and Cobas e601 (Roche Diagnostics Mann­ were measured. In addition, the concentrations of heim, Germany), employ electrochemilumines- immunoglobulins and rheumatoid factor were cence immunoassays (ECLIA) and use a sandwich measured employing an automated nephelo­ method with two mouse monoclonal antibodies metric assay (Immage 800, Beckman Coulter, Brea, (capture and detection), which are specifically di- USA). rected against epitopes 125-131 and 136-147 of hu- man cardiac troponin T (4). In order to confirm the results, another visit was scheduled 3 months later in which all tests were Precicontrol Troponin quality control (ref: repeated and the concentration of various autoanti­ 05095107922, Roche Diagnostics, Mannheim, Ger- bodies by indirect immunofluorescence assay many) was processed during the study period for (Quantalyser 160, Werfen, California, USA) was both analysers. The coefficient of variation (CV) measured (Table 1). In addition, the presence of Biochem Med (Zagreb) 2019;29(2):021001 https://doi.org/10.11613/BM.2019.021001 2 Aliste-Fernández M. et al. Interference in the measurement of troponin T 60 heterophilic antibodies in the sample was as- sessed by incubation for 1 hour with the Scanti- 50 bodies Heterophilic Blocking Tube (HBT) (Scanti- y = 0.82x + 10.12 bodies Laboratory, Santee, USA), which can block 40 R2 = 0.99 these antibodies. Obtained values Finally, the serum sample was sent to the manu- 30 facturer to further investigate the potential inter- 20 ference. The sample was fractionated by size ex- clusion chromatography to detect the molecular 10 size that could cause interference with the hs-TnT test. The principle of this technique includes a diffe­ 0 rent elution speed depending on the molecular 0 10 20 30 40 50 60 weight. Therefore, since the molecular weight of Expected values troponin T is known (~36 kDa), it is possible to infer Figure 1. Linearity test between expected and obtained values of the dilution test (ng/L). Table 1. Measurements performed to elucidate the cause of the falsely elevated results of hs-TnT Possible causes Our results Reference ranges Analyser No significant Serum indices: Interference by endogenous interference up to: Haemolysis (H) = 8 compounds in blood (haemoglobin, H = 100 mg/dL Cobas 8000 Icterus (I) = 1 bilirubin and lipemia) I = 25 mg/dL Lipemia (L) = 4 L = 1500 mg/dL Cross reactivity with skeletal muscle Creatine kinase: 152 U/L 0-195 U/L Cobas 8000 High activity of alkaline phosphatase Alkaline phosphatase: 150 U/L 2-300 U/L Cobas 8000 Kidney failure Serum creatinine: 52 µmol/L 34–65 µmol/L Cobas 8000 Sample before reaction: 52 ng/L Macro complex in the sample Sample after reaction with PEG 6000: 67 ng/L - Cobas e801 (Concentration of hs-TnT) Percent recovery > 100% PTH: 14.4 ng/L 10.0-65.0 ng/L CA 125: 19.8 kU/L < 35.0 kU/L Measurements of other analytes AFP: under detection limit < 6.0 kUI/L which could potentially influence the CA 19.9: 12.6 kU/L < 22.0 kU/L Cobas e801 same interference TPSA: under detection limit < 4.0 µg/L TSH: 4.44 mIU/L 0.40-5.00 mIU/L FT4: 19.8 pmol/L 10.3–23.2 pmol/L Immunoglobulin G: 14.30g/L 7.40–14.00 g/L Concentration of immunoglobulins Immunoglobulin M: 1.52 g/L 0.70-2.00 g/L Immage 800 Immunoglobulin A: 1.35 g/L 0.60-3.56 g/L Rheumatoid factor < 20 IU/mL 70–200 IU/mL Immage 800 ANA, ASMA, AMA and APCA: < 1/40 < 1/40 Quantalyser Autoantibodies titers ANCA: Negative Negative 160 Sample before incubation: 32 ng/L Heterophile antibodies Sample after incubation with HBT: 33 ng/L - Cobas e601 (Concentration of hs-TnT) Absence of heterophile antibodies removable by HBT hs-TnT - high-sensitivity troponin T. HBT - heterophilic blocking tube. PTH - parathyroid hormone. CA - carbohydrate antigen. AFP - alpha-fetoprotein. TPSA - total specific prostate antigen. TSH – thyrotropin. FT4 - free thyroxine. ANA - anti-nuclear antibodies. ASMA - anti-smooth muscle antibodies. AMA - anti-mitochondrial antibodies. APCA - anti-parietal cells antibodies. ANCA - anti-neutrophil cytoplasmic antibodies. https://doi.org/10.11613/BM.2019.021001 Biochem Med (Zagreb) 2019;29(2):021001 3 Aliste-Fernández M. et al. Interference in the measurement of troponin T if the fraction in which troponin T is expected is Size exclusion chromatography results suggest also showing the reactivity with the hs-TnT assay. that the interference can be allocated to fractions with proteins of 150 kDa, corresponding to high molecular weight proteins like immunoglobulin What happened? G (IgG). Quoting the manufacturer report: “the Table 2 shows the evolution of the troponin T con- observed hs-TnT values are falsely elevated in the centration in the patient, as well as other biomark- sample. The interference is caused by a high mo- ers of myocardial injury. lecular weight compound, presumably IgG”. Inde- According to the literature on interference in immu­ pendent studies concluded that such type of in- noassays additional studies were performed, as terference are rare and are covered by disclaimers summarized in Table 1 (3,6). Regarding the dilution in the method sheet (3). The manufacturer also test, expected and observed values are shown in performed troponin T measurements in different Table 3. Moreover, a linearity test between expect- analysers (N = 1) employing ECLIA technology ed and observed values was performed (Figure 1), (Table 4). which showed that existed linearity between them (R2 = 0.99). Table 2. Evolution of the different myocardial damage markers in the patient; single reading Visit Analyte (unit); analyser CK (U/L); hs-TnT (ng/L); hs-TnT STAT (ng/L); hs-TnI STAT (ng/L); Cobas 8000 Cobas e801 Cobas e601 Architect First visit 115 51 - - Second visit 49 98 - - Third visit 152 52 28 < 10 Fourth visit 129 - 32 < 10 CK - creatine kinase. hs-TnT - high-sensitivity troponin T. STAT - short turnaround time. hs-TnI - high-sensitivity troponin I. The 99th percentile for the specific test was: 0 - 195 U/L (CK), ≤ 14 ng/L (hs-TnT and hs-TnT STAT), ≤ 26 ng/L (hs-TnI STAT). Table 3. Results of dilution test (Cobas e801) Sample before dilution Dilution 1:2 Dilution 1:3 Dilution 1:5 Concentration of high-sensitivity troponin T 33 ng/L 24 ng/L 18 ng/L 52 ng/L measured (66 ng/L) (72 ng/L) (88 ng/L) (Dilution result x DF) Concentration of high-sensitivity troponin T 52 ng/L < 26 ng/L < 17 ng/L < 10 ng/L expected DF – dilution factor. Biochem Med (Zagreb) 2019;29(2):021001 https://doi.org/10.11613/BM.2019.021001 4 Aliste-Fernández M. et al. Interference in the measurement of troponin T Table 4. Manufacturer’s study in which troponin T concentration were measured with different analysers with electrochemilumines- cence immunoassay technology hs-TnT STAT TnT STAT hs-TnT Analyser e601 (ng/L) e411 (ng/L) e801 (ng/L) e601 (ng/L) e801 (ng/L) Average of outcomes 34 41 32 < 10 54 hs-TnT STAT - high-sensitivity troponin T short turnaround time. TnT STAT - troponin T short turnaround time. hs-TnT - high-sensitivity troponin T. Discussion We report the case of a 12-years-old boy with high ponin I are both elevated in case of myocardial in- concentration of hs-TnT without evidence of myo- jury, so an undetectable concentration of hs-TnI cardial injury. As a consequence of this discord- meant that there was no real injury and the eleva- ance, a falsely positive result was suspected. tion of hs-TnT was caused by an interfering com- After checking the correct calibration status of the pound. analyser and repeating the hs-TnT measurements, In addition, the presence of macro complexes was possible pre-analytical causes that could alter the investigated but after not observing different re- results were studied. The presence of fibrin clots sults in the sample before and after PEG 6000 ad- or micro particles was ruled out after inspection of dition, such type of interferent was excluded the sample and besides that, the haemolysis, icte­ (10,11). rus and lipemia indices were found to be under After discarding the presence of rheumatoid fac- the interference level (7). Finally, renal failure, tor, an IgM autoantibody with IgG-binding capaci- which may increase troponin concentration, was ty, we investigated another well-known cause of discarded for the patient (8). interference in immunoassays: heterophilic anti- A simple way to detect interference is performing bodies (HA). Its incidence in the population lies be- serial dilutions of the sample with a suitable sol- tween 0.2 - 40% and it includes two specific cir- vent. Normally, in the event that interference is cumstances: the so-called “true” HA, which are present, reported concentrations will not dilute multispecific and low affinity antibodies, and hu- linearly until the substance is eliminated (9). Accor­ man anti-animal antibodies (HAAA), which are ding to the results of the dilution study, slightly high affinity antibodies against specific animal im- higher troponin concentrations were observed munoglobulins (HAMA) (7,12). They are present in when dilutions were made, which could be caused the blood after treatment with monoclonal anti- by the imprecision added to the measurement bodies or by contact with animals, among other due to the dilution factor. Since the concentration causes (8,13). In the case of falsely positive results, of hs-TnT did not decrease as expected and the lin- the HA, in the sample, bridge capture and detec- earity test between expected and obtained values tion antibodies to mimic analyte binding. The ac- showed that existed linearity, we discarded the tivity of this type of antibodies cannot be com- presence of an interferent disappearing with the pletely ignored even if manufacturers add certain dilution. compounds in their reagents to reduce its influ- After performing this dilution test, we could not ence (4). detect the interference, although its presence was It is possible to eliminate the HA by the incubation evident taking into account the observed differ- of the sample with HBT. However, since some HBT ence between the concentration of hs-TnT (52 tubes are not reliable when immunoassays em- ng/L) and hs-TnI (< 10 ng/L). Troponin T and Tro- ploys materials derived from mice, HA elimination https://doi.org/10.11613/BM.2019.021001 Biochem Med (Zagreb) 2019;29(2):021001 5 Aliste-Fernández M. et al. Interference in the measurement of troponin T is not always possible (6). In our case, no differenc- says have consistently shown a higher negative es were observed in the concentration of hs-TnT predictive value compared to the less sensitive after the HBT sample incubation. tests the average clinical performance indicators According to the manufacturer’s additional inves- of the troponin tests need to be taken into ac- tigation, the interference might be eventually due count for a conclusive benefit-risk analysis (14). to an IgG, although a final IgG validation was not proven. This IgG could correspond to an antibody Main lesson that was not removable with a HBT, and thus pos- sibly a HAMA, that might be found in serum of ani- Nowadays, troponin measurements are indispen- mal workers or in patients receiving mouse mono- sable part of AMI definition and related guidelines clonal antibodies for therapy or imaging. In sup- for their interpretation are urgently required in port of that hypothesis, the medical interview re- daily clinical practice. When a high concentration vealed that the patient had several pets a few of troponin is not compatible with the clinic signs years ago, including hamsters. Moreover, the con- and symptoms of the patient, the presence of an centration of IgG in the patient serum sample was interfering factor in the sample should always be elevated (Table 1). suspected. However, it was not possible to identify the cause In the case reported here, falsely elevated results of the discrepancy between the results obtained of hs-TnT in two different analysers might have from the analysers of our laboratory. The only been caused by an IgG antibody that could not be methodological difference lies in the use of tech- removed after incubation with HBT. At low analyte nology STAT (Short turnaround time) in the meas- concentration, this effect was observed only for urement procedure of Cobas e601, which reduces the high-sensitivity methods of troponin T. In ad- the analysis time from 18 minutes to 9, by reduc- dition, the concentration of hs-TnT was lower when ing the number of incubations from two to one. It STAT technology was used, which could be ex- might be possible that an extended incubation plained by the lower incubation time. Even if this time could favour the union of the IgG to the anti- type of interference is disclaimed by the manufac- bodies, and therefore enhancing the interference, turer, the present case described the track to sub- justifying the higher concentration of hs-TnT ob- stantiate such hypothesis. served in the method that uses two incubations Interference by HA is a problem that may affect instead of one, but further validation was not pos- the immunoassays, especially after the monoclo- sible. Nevertheless, such single results should not nal antibody-based therapy use in recent years be over interpreted. (15,16). Therefore, interdisciplinary cooperation be- In addition, a former less sensitive method (Tro- tween clinicians, laboratory and manufacturer is ponin T STAT) provided no evidence of troponin essential for the appropriate interpretation of the elevation in the same patient sample (< 10 ng/L), results. which indicates that the high-sensitivity methods Acknowledgments might be more susceptible to such type of inter- ference. Irrespectively of the used immunoassay We want to express our gratitude to Dr. S. Kunzel- type, either one or two step assays, the frequency mann and Roche Diagnostics® for its valuable con- of such HAMA interference has been reported be- tribution to the study of this interference. tween 0.1 - 3.1% and may involve both Troponin I Potential conflict of interest and Troponin T tests (12). Given that the hs-Tn as- None declared. Biochem Med (Zagreb) 2019;29(2):021001 https://doi.org/10.11613/BM.2019.021001 6 Aliste-Fernández M. et al. Interference in the measurement of troponin T References 1. Jaffe AS, Ordonez-Llanos J. High sensitivity troponin in 9. Mair J, Lindahl B, Müller C, Giannitsis E, Huber K, Möckel M, chest pain and acute coronary syndromes. A step forward? et al. What to do when you question cardiac troponin va- Rev Esp Cardiol. 2010;63:763-9. https://doi.org/10.1016/ lues. Eur Heart J Acute Cardiovasc Care. 2018;7:577-86. S0300-8932(10)70178-0 https://doi.org/10.1177/2048872617708973 2. Ordoñez Llanos J. Troponina T cardíaca ultrasensible. Un 10. Kavsak PA, Roy C, Malinowski P, Mark CT, Scott T, Clark L, et paso adelante en el diagnóstico cardiológico. Roche Di- al. Macrocomplexes and discordant high-sensitivity cardiac agnostics informa 2010;45-51. Available at: https://doc- troponin concentrations. Ann Clin Biochem. 2018; 55:500-4. player.es/25473238-Roche-diagnostics-informa.html. Acce- https://doi.org/10.1177/0004563217734883 ssed November 7th 2018. (in Spanish) 11. Plebani M, Mion M, Altinier S, Girotto MA, Baldo G, Zani- 3. Franeková J, Bláha M, Bělohoubek J, Kotrbatá M, Sečník P notto M. False-Positive Troponin I attributed to a macro- Jr, Kubíček Z, et al. A clinical and laboratory approach used complex. Clin Chem. 2002;48:677-9. to elucídate discordant results of high-sensitivity troponin 12. Lippi G, Aloe R, Meschi T, Borghi L, Cervellin G. Interference T and troponin I. Clin Chim Acta. 2015;446:128-31. https:// from heterophilic antibodies in troponin testing. Case re- doi.org/10.1016/j.cca.2015.03.046 port and systematic review of the literature. Clin Chim Acta. 4. Elecsys Troponin T hs, method sheet - Roche Dialog. Ava- 2013;426:79-84. https://doi.org/10.1016/j.cca.2013.09.004 ilable at: https://pim-eservices.roche.com/eLD_SF/es/es/ 13. Ntelios D, Karamitsos TD, Boufidou A, Thisiadou K, Karyo- Documents/GetDocument?documentId=b567b693-d989- unis H. Inconsistent high sensitivity troponin T and I mea- e711-fbba-00215a9b3428. Accessed October 16th 2018. surements in a patient with rheumatoid arthritis. Hellenic 5. Marinheiro R, Amador P, Parreira L, Rato Q, Caria R. False Po- J Cardiol. 2018 Mar 30 [cited 2019 Jan 3] . [Epub ahead of sitive Troponin I Rendering Two Admissions for “Recurrent print]. https://doi.org/10.1016/j.hjc.2018.03.004. Acute Myopericarditis”. Open Cardiovasc Med J. 2018;12:55- 14. Lipinski MJ, Baker NC, Escárcega RO, Torguson R, Chen F, 8. https://doi.org/10.2174/1874192401812010055 Aldous SJ. Comparison of conventional and high-sensiti- 6. Ward G, Simpson A, Boscato L, Hickman PE. The inve- vity troponin in patients with chest pain: a collaborative stigation of interferences in immunoassay. Clin Bio- meta-analysis. Am Heart J. 2015;169:6-16.e6. https://doi. chem. 2017;50:1306-11. https://doi.org/10.1016/j.clinbio- org/10.1016/j.ahj.2014.10.007 chem.2017.08.015 15. Ntelios D, Mpei E, Gousi E, Potolidis E, Fanourgiakis P, Man- 7. Chueca MP, Güell R, Gelabert JM, Izquierdo F. Interferencias dros C. Are troponin assays occasionally deceiving us? Am en la medición de troponina. Quím Clín. 2007;26:216-23. (in J Emerg Med. 2013;31:997.e1-2. https://doi.org/10.1016/j. Spanish) ajem.2013.01.035 8. Vafaie M, Biener M, Mueller M, Schnabel PA, André F, Ste- 16. Nguyen J, Thachil R, Vyas N, Marino T. Falsely elevated en H, et al. Analytically false or true positive elevations of troponin: rare occurrence or future problem. J Commu- high sensitivity cardiac troponin: a systematic approach. nity Hosp Intern Med Perspect. 2016;6:32952. https://doi. Heart. 2014;100: 508-14. https://doi.org/10.1136/hear- org/10.3402/jchimp.v6.32952 tjnl-2012-303202 https://doi.org/10.11613/BM.2019.021001 Biochem Med (Zagreb) 2019;29(2):021001 7