DOI: 10.1161/CIRCULATIONAHA.115.016848
Is the 99th Percentile the Optimal Reference Limit to Diagnose Myocardial Infarction with High Sensitivity Cardiac Troponin Assays in Patients with Chronic Kidney Disease?
Running title: Chaitman; The 99th percentile URL, hs-cTn Assays and CKD
Bernard R. Chaitman, MD
Saint Louis University Schooll of o Medicine, St. Louis, Louuis is,, MO
Addr Ad dres dr esss for es for Correspondence: Corr Co rres rr espo es pond po nden nd ence en ce:: ce Address Bernard R. Chaitman, MD Professor of Medicine, Director of Cardiovascular Research Saint Louis University School of Medicine 1034 S. Brentwood Blvd., Suite 1550 St. Louis, MO 63117 Tel: 314-725-4668 Fax: 314-725-2171 E-mail: [email protected] Journal Subject Codes: [3] Acute coronary syndromes, [4] Acute myocardial infarction
Key words: Editorial, myocardial infarction, diagnosis, coronary artery disease
1 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
The estimated annual incidence of myocardial infarction in the United States is ~525,000 new attacks/year and ~210,000 recurrent attacks1. The large number of events results in many emergency room visits for diagnostic evaluation to determine presence or absence of acute coronary syndrome (ACS) and need for hospital admission. The Third Universal Definition of Myocardial Infarction requires abnormal cardiac biomarkers in the context of acute myocardial ischemia supported by clinical, ECG, or cardiac imaging findings2. Cardiac troponin is recommended as the preferred biomarker to document myocardial necrosis. A rise or fall of c-Tn with at least one value > 99th percentile of a reference control population using an assay with total imprecision at the 99th percentile 99th percentile are recommended. In the the last lasst few f w years, more sensitive as well fe wel elll as hs-cTn assays have haave ave been introduced that allow allo al ow more rapid rap pid diagnosis diaagnos osis os is of of myocardial myo yoccard card rdiaal ne necrosis ecrosis aand ndd de dete detection tect te ctio ct on of ssmaller malller M ma MII ev events ven entss tthan hann wa ha wass previously pr rev evio i usly ppossible io o sibl os blee us bl using sin ng th the he st stan standard a da dard rd cT cTn Tn assay. asssay. M Many any ny ppatients attieentts pr previously rev vio oussly cl clas classified a sifi fied fi ed ass uunstable nstab able angina angi an gina gi na are are nnow ow aable ow blee to be bl be reclassified recl re clas cl assi as sifi si fied fi ed aass an M MII usin us using sin ingg mo more re ssensitive ensi en siti si tive ti vee T Tn n as assa assays. says sa ys. s Ea Earl Early rly detection rl dete de tect te ctio ct ionn io of acute myocardial necrosis may allow introduction of earlier treatment that would reduce morbidity and mortality. However, detection of lower levels of cTn than was previously possible using older assays allows detection of cTn in patients with stable coronary artery disease (CAD), and in patients with non-ischemic conditions associated with low-level cTn elevations in the absence of coronary disease, thus, increasing the false positive rate for MI diagnosis. This decreased specificity results in more complex triage in the emergency room and cardiac consultations due to elevated troponin levels.
2 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
Chronic Kidney Disease Patients with chronic kidney disease (CKD) represent a large group of patients where the diagnosis of ACS can be challenging, particularly the diagnosis of a non-ST elevation myocardial infarction (NSTEMI). It is estimated that > 20 million United States (US) adults have CKD. ACS evaluation in a patient with CKD is complex; initial c-Tn levels using a standard cTn assay often exceed the 99th percentile, elevated troponin levels may be the result of chronic structural heart disease (e.g. heart failure) rather than acute myocardial ischemia, and ECG repolarization abnormalities are common3. Sensitive, and high-sensitivity c-Tn assays (hs-cTn is available in Europe and some parts of the world) further increase the number of patients with initial nitial abnormal troponin values, and raise the question as to whether or not the 99 9th ppercentile erce er cent ce ntil nt ilee is il the he appropriate CDV to use for CKD patients presenting with ACS. To address add d reess this question, Twerenbold et al. al te ttested sted 3 sensitive and nd 4 high-sensitivity cTnaassays ssaays in the Ad Adva Advantageous v nttag va ageo eous eo us P Predictors redi re dict di ctor ct o s of Ac Acu Acute ut C ute Coronary orrona nary na ry S Syndrome yndr yn d om dr me Ev Eva Evaluation aluaati tion on ((APACE) APAC AP ACE) AC E) multicenter prospective observational registry, MII in ppatients with multtic mu i enter pr prospe pectiv pe ve obse erv vat a io ona nal re regi g sttry ry,, examining exam min nin ng the th he frequency f eque fr uenc ue n y ooff M atienntss wit at th CKD, CKD CK D, ((eGFR eGFR eG FR < 660 0 ml ml/m ml/min/1.73m /min /m in/1 in /1.73m 73m2) an andd without wiith with thou outt CKD CKD4. A strength str tren engt en gthh of A gt APACE PACE PA CE iiss th that at aall ll cc-Tn -Tn Tn assays were measured in a core lab, patients were enrolled within 12 hours of symptom onset, the protocol specified that blood samples be taken at presentation, and 1, 2, 3 and 6 hours until the diagnosis of MI was ruled in/out, and the end-point of myocardial infarction was adjudicated by a clinical event committee. The committee used the Universal Definition of Myocardial Infarction, serial levels of hs-cTnT using the Roche assay, all available medical records including whether c-Tn levels were elevated on prior admissions, ECG, and angiographic findings when available, to classify events. Myocardial infarction was considered if at least one hs-cTnT value exceeded the 99th percentile. A significant absolute change in hs-cTnT was
3 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
defined as a rise or fall of at least 10 ng/l within 6 hours or a change of 6 ng/l within 3 hours. Of the 2,813 APACE patients, 447 (16%) had CKD. Adjudicated MI was the final diagnosis in 36% vs. 18% of the CKD vs. non-CKD patients. Receiver operator characteristics (ROC) analysis revealed diagnostic accuracy for MI diagnosis was less in patients with vs. without CKD. Optimal ROC-derived cTn cut-off levels in patients with renal dysfunction were significantly higher compared to patients with normal renal function (factor 1.9-3.4). For sensitive c-Tn assays, the optimal cut-point determined by ROC analysis in the CKD cohort approximated the manufacturer’s 99th percentile whereas, the optimal cut-point was significantly higher than the 99th percentile for the 4 hs-cTn assays. Long-term survival was worse in patients with elevated Tn levels using the more sensitive c-Tn assays, similar to data publi published liish shed ed uusing sing si ng lless ess sensitive ensitive assays3. CKD patients that ruled out for adjudicated MI had higher baseline levels of cTn than tha hann non-CKD nonno n-CK nCKD CK D patients with all 7 cTn assays s; th the frequency of eleva vaate ted baseline values above assays; elevated the he 99th percent percentile til ile ra ranged ang nged ed ffrom rom ro m 12 12-7 12-71% -71% -7 1% ddepending ep pen ending g oon n th thee as ass assay say ty say ttype pe uused. s d. L se Level evel ev el ooff cT cTnn va valu values lues lu es we were eree inverse inversely seely ccorrelated orrelaateed to C or CKD KD sseverity. eveeriity. ity Thee findings Th find fi ndin nd ings in gs from fro rom m the the APACE APAC AP ACE AC E registry regi re gist gi stry st ry using usi sing si ng the the hshs cTn cTn assays assa as says sa yss indicate ind ndiica cate te that tha hatt patients pati pa tien ti ents en ts with CKD presenting with an ACS are more likely to have elevated cTn levels on initial presentation than was previously published using older cTn assays. Increased number of patients with elevated cTn will predictably lead to more cardiac consultations from the emergency room, and perhaps, more hospital admissions, raising the question as to whether the hs-cTn assays offer an advantage over the older cTn assays in these patients. Many institutions in Europe have already switched over to hs-cTn assays. Therefore, clinicians need to take into consideration the assay type used at their particular institution, degree of renal impairment, and initial and subsequent c-Tn levels when determining if a patient with CKD and ACS has had an MI. For
4 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
patients with CKD using hs-cTn assays (which are not available in the U.S. at the time of this editorial), a cut-point > 99th percentile should improve diagnostic accuracy. Other approaches such as using the delta (changing pattern of values) rather than absolute change in hs-cTn as used by Twerenbold et al. requires further study in a CKD population5,6. The 99th Percentile Upper Reference Limit The finding that different c-Tn assays are not biologically equivalent is well known due to biochemical differences between assays and the fact that various manufacturers use different reference populations to determine the 99th percentile URL. Several studies have shown the 99th percentile URL as reported by the manufacturer may not replicate values seen in large community-based cohort studies and can result in major differences in determinin determining ng th thee 99th percentile URL7. To further examine the issue of establishing a an appropriate CDV for hs-cTn assa ays iin n a ge gener ral population of patients presen ral nti ting ng with an ACS, Wildi di et al. studied 2300 assays general presenting ppatients atiients in the AP APAC APACE AC CE re regi registry, gist gi stry st ry, 47 ry 4733 (2 (21%)) off who whom om ha hadd an n aadjudicated djjudic icat ated at ed MI MI diagnosis diaagno di agno nosi s s8. Th si The he primary prrim imar a y outcome outc tccomee measure measure was was the t e percentage th p rccenta pe en age of patients patien pa nts ts with withh an adjudicated adj djud udiccated ud ed M MII or or nno o MI M diagnosis diag di agno ag nosi no siss at ppresentation si rese re sent se ntat nt atio at ionn usi io us using sing tthe he aapproved ppro pp rove ro ved ed ma manu manufacturer nufa fact fa ctur ct urer rer C CDV DV fo forr th thee hs hs-cTnI -cTn cTn TnII Ab Abbo Abbot bott an bo andd hshs cTnT Roche assays; frequency of discordant MI results with the two assays were compared. For hs-cTn levels measured at presentation and at all time points, hs-cTnI and hs-cTnT levels were closely correlated (r=0.813 and r=0.790), and had comparable accuracy for early diagnosis of MI. However, among the 473 adjudicated MI patients, 86 (18%) had discordant MI diagnoses using the approved CDV; 14.1% for women and 22.7% for men. Approximately 20% of the time, MI was diagnosed by one assay but not the other. The authors tested other sensitive and hscTn assays and obtained similar results, not confirming the superiority of one assay over the other, but rather illustrating that all assays are not biologically equivalent. Inconsistencies
5 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
between the hs-cTnI and hs-cTnT assay were able to be reduced to ~10% after adjustment of the cTnI CDV to approximate biological equivalence to the cTnT assay. The findings from both APACE papers in Circulation this week expand and enhance multiple prior observations from this group of investigators and point out additional data that should be considered when the 99th percentile is used to diagnose MI. APACE Registry An important consideration when evaluating data from the APACE registry is the cardiologists’ use of the Roche hs-cTnT assay to diagnose MI. The cardiologists used the 99th percentile URL of 14 ng/L as reported by the manufacturer. In population based studies, the 99th percentile URL for the Dallas Heart Study (DHS), the Atherosclerosis Risk in Communities (ARIC) (ARIIC) study, stu tudy dy,, an dy andd the he Cardiovascular Health Study (CHS) was 14, 21, and 28 ng/L, respectively7. Within each coho ort rt,, th thee 99 99th ppercentile ercentile URL value increasedd wi with t age and was highe herr in men. Thus, the he cohort, higher manufacturer’s m annufacturer’ss U URL RL L vvalue alue al ue ddoes oess not oe not alw always lways re lw replicate epliccatte va valu values lues lu es sseen eeen inn ot oth other her “hea her “healthy” ealt ea lthy lt hy”” pa hy pati patient tien ti en nt cohorts. co oho hort r s. In AP APAC APACE, CE, the he card cardiologists diol olog ol ogis og ists is t use uused sed aan n abso absolute olu utee cchange hannge nge de defined efi fine need as as a rrise i e or ffall is all off ccTnT al TnT T at least eas astt 10ng/l 10ng 10 ng/l ng /l within wit ithi it hinn 6 hours hi hour ho urs rs or a change cha hang ngee of 6 ng/l ng ng/ g/ll within wit ithi it hinn 3 hours. hi hour ho urs rs. It It is not not clear clea cl earr if these ea the hese se absolute abs bsol olut ol ute te numerical values optimize performance of the other assays tested, if gender considerations would improve performance, and if delta rather than absolute change would improve diagnostic accuracy5,6,9. These differences might explain some of the inter-assay differences observed in both papers as well as the relatively high rate (36%) of confirmed MI (mainly NSTEMI) in the CKD cohort4,8. The implications of the data from the APACE registry for clinical patient management and clinical trials that use MI as an end-point are substantial. Clearly, some patients presenting with non-ST elevation-ACS may be misclassified simply due to the fact that the manufacturer’s
6 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
reference population chosen to establish the 99th percentile URL is not representative of the population being studied. Further, with hs-cTn assays, certain patient phenotypes (e.g. CKD) may require higher thresholds than the 99th percentile or other algorithms such as delta change to optimize MI diagnostic accuracy. The use of a standardized approach for approval of hs-cTn assays that would allow clinicians or investigators to determine biologic equivalence across all of the multiple manufactured cTn assays would be invaluable. For the time being, clinicians involved in the daily management of ACS patients in a hospital using more sensitive assays should be aware of non ACS conditions associated with elevations in cTn, and able to determine the pretest likelihood of myocardial infarction from all available data to estimate posttest MI ikelihood. A standardized blood sampling protocol should be used to facilitate ea arl rlie ierr di ie diag agno ag nosi no ss likelihood. earlier diagnosis of MI, with samples drawn at initial assessment (time 0) and repeated at least 3 and 6 hours later later, as req equi eq uire ui redd to establish re esttab ablish the diagnosis for patientss wi with (or without) initia iall cTn above the CDV. ia required initial
Conflict C on nflict of Interest In ntere rest Disclosures: Dissclossurres: N None. onne..
References: 1. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, de Ferranti S, Després JP, Fullerton HJ, Howard VJ, Huffman MD, Judd SE, Kissela BM, Lackland DT, Lichtman JH, Lisabeth LD, Liu S, Mackey RH, Matchar DB, McGuire DK, Mohler ER III, Moy CS, Muntner P, Mussolino ME, Nasir K, Neumar RW, Nichol G, Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani SS, Willey JZ, Woo D, Yeh RW, Turner MB, on behalf of the American Heart Association Statistics Committee and Stroke Statistics Committee. Heart Disease and Stroke Statistics—2015 Update. A Report from the American Heart Association. Circulation. 2015;131:e29-e322. 2. Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD; the Writing Group on behalf of the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction. Third universal definition of myocardial infarction. Circulation. 2012;126:2020-2035. 3. Stacy SR, Suarez-Cuervo C, Berger Z, Wilson LM, Yeh HC, Bass EB, Michos ED. Role of troponin in patients with chronic kidney disease and suspected acute coronary syndrome. A
7 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
systematic review. Ann Intern Med. 2014;161:502-512. 4. Twerenbold R, Wildi K, Jaeger C, Gimenez MR, Reiter M, Reichlin T, Walukeiwicz A, Gugala M, Krivoshei L, Marti N, Weidmann ZM, Hillinger P, Puelacher C, Rentsch K, Honegger U, Schumacher C, Zurbriggen F, Freese M, Stelzig C, Campodarve I, Bassetti S, Osswald S, Mueller C. Optimal cut-off levels of more sensitive cardiac troponin assays for the early diagnosis of myocardial infarction in patients with renal dysfunction. Circulation. 2015;131:XX-XXX. 5. Morrow DA, Bonaca MP. Real world application of “Delta” troponin. J Am Coll Card. 2013;62:1239-1241. 6. Jaffe AS, Moeckel M, Giannitis E, Huber K, Mair J, Mueller C, Plebani M, Thygesen K, Lindahl B, In search for the Holy Grail: Suggestions for studies to define delta changes to diagnose or exclude acute myocardial infarction: a position paper from the study group on biomarkers of the Acute Cardiovascular Care Association. Eur Heart J: Acute Cardiovasc Care. 2014;3:313-316. 7. Gore MO, Seliger SL, deFilippi CR, Nambi V, Christenson RH, Hashim IA, Hoogeveen Hoog Ho ogev og evee ev eenn RC, ee RC Ayers CR, Sun W, McGuire DK, Ballantyne CM, de Lemos JA. Age- and sex-depe sex-dependent pend pe nden nd entt up en uppe upper perr pe reference eference limits for the high-sensitivity cardiac troponin T assay. J Am Coll Cardiol. 2014;63:1441–1448. 8.. Wildi Wildi K, Gimenez Gimenez MR, Twerenbold R, Reichlin n T, Jaege Jaeger g r C, H Heinzelmann einzelmann A, Arnold C, Nelles N ellles ll B, Drueyy S, Haaf Haa a f P, Hillinger Hil illi ling li nger ng er P, P, Schaerli Scha haeerli N ha N,, Kr Kreutzinger reu euttzin tzin ingger ger P, T Tanglay angl an glay Y Y,, He Herr Herrmann rrma rr mannn T ma T,, Weidmann Rentsch Osswald Mueller Weid We i mann ZM, Krivoshei Krivos oshei L, L, Freese M, M, Stelzig Steelzzig C, C, Puelacher Puel elac el achherr C,, R ac entscch K, O sswa waald d S,, M uelleer myocardial related limitations current approach C Misdiagnosis C. Mis i diag gno nosis off myo yocard yo dia i l in iinfarction farc fa r tiionn rel elatted tto el o lim mittatio io ons ns ooff the the cu urrrent nt rregulatory e ullat eg atooryy ap pproa oach to clinical values cardiac troponin. o ddefine efin ef inee cl in cli inical al ddecision ecis ec isio ionn val lue uess fo for ca ard rdia iacc tr ia trop pon onin in. Circulation Cirrcul Ci rcul ulaationn 2015;131:XX-XXX. 201 0155;13 5;13 131:XX XX-X XX -X XXX X. 9. Reichlin T, Irfan A, Twerenbold R, Reiter M, Hochholzer W, Burkhalter H, Bassetti S, Steuer S, Winkler K, Peter F, Meissner J, Haaf P, Potocki M, Drexler B, Osswald S, Mueller C. Utility of absolute and relative changes in cardiac troponin concentrations in the early diagnosis of acute myocardial infarction. Circulation. 2011;124:136-145.
8 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
Is the 99th Percentile the Optimal Reference Limit to Diagnose Myocardial Infarction with High Sensitivity Cardiac Troponin Assays in Patients with Chronic Kidney Disease? Bernard R. Chaitman Circulation. published online May 6, 2015; Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2015 American Heart Association, Inc. All rights reserved. Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circ.ahajournals.org/content/early/2015/05/06/CIRCULATIONAHA.115.016848
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Circulation is online at: http://circ.ahajournals.org//subscriptions/
Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
Is the 99th Percentile the Optimal Reference Limit to Diagnose Myocardial Infarction with High Sensitivity Cardiac Troponin Assays in Patients with Chronic Kidney Disease?
Running title: Chaitman; The 99th percentile URL, hs-cTn Assays and CKD
Bernard R. Chaitman, MD
Saint Louis University Schooll of o Medicine, St. Louis, Louuis is,, MO
Addr Ad dres dr esss for es for Correspondence: Corr Co rres rr espo es pond po nden nd ence en ce:: ce Address Bernard R. Chaitman, MD Professor of Medicine, Director of Cardiovascular Research Saint Louis University School of Medicine 1034 S. Brentwood Blvd., Suite 1550 St. Louis, MO 63117 Tel: 314-725-4668 Fax: 314-725-2171 E-mail: [email protected] Journal Subject Codes: [3] Acute coronary syndromes, [4] Acute myocardial infarction
Key words: Editorial, myocardial infarction, diagnosis, coronary artery disease
1 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
The estimated annual incidence of myocardial infarction in the United States is ~525,000 new attacks/year and ~210,000 recurrent attacks1. The large number of events results in many emergency room visits for diagnostic evaluation to determine presence or absence of acute coronary syndrome (ACS) and need for hospital admission. The Third Universal Definition of Myocardial Infarction requires abnormal cardiac biomarkers in the context of acute myocardial ischemia supported by clinical, ECG, or cardiac imaging findings2. Cardiac troponin is recommended as the preferred biomarker to document myocardial necrosis. A rise or fall of c-Tn with at least one value > 99th percentile of a reference control population using an assay with total imprecision at the 99th percentile 99th percentile are recommended. In the the last lasst few f w years, more sensitive as well fe wel elll as hs-cTn assays have haave ave been introduced that allow allo al ow more rapid rap pid diagnosis diaagnos osis os is of of myocardial myo yoccard card rdiaal ne necrosis ecrosis aand ndd de dete detection tect te ctio ct on of ssmaller malller M ma MII ev events ven entss tthan hann wa ha wass previously pr rev evio i usly ppossible io o sibl os blee us bl using sin ng th the he st stan standard a da dard rd cT cTn Tn assay. asssay. M Many any ny ppatients attieentts pr previously rev vio oussly cl clas classified a sifi fied fi ed ass uunstable nstab able angina angi an gina gi na are are nnow ow aable ow blee to be bl be reclassified recl re clas cl assi as sifi si fied fi ed aass an M MII usin us using sin ingg mo more re ssensitive ensi en siti si tive ti vee T Tn n as assa assays. says sa ys. s Ea Earl Early rly detection rl dete de tect te ctio ct ionn io of acute myocardial necrosis may allow introduction of earlier treatment that would reduce morbidity and mortality. However, detection of lower levels of cTn than was previously possible using older assays allows detection of cTn in patients with stable coronary artery disease (CAD), and in patients with non-ischemic conditions associated with low-level cTn elevations in the absence of coronary disease, thus, increasing the false positive rate for MI diagnosis. This decreased specificity results in more complex triage in the emergency room and cardiac consultations due to elevated troponin levels.
2 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
Chronic Kidney Disease Patients with chronic kidney disease (CKD) represent a large group of patients where the diagnosis of ACS can be challenging, particularly the diagnosis of a non-ST elevation myocardial infarction (NSTEMI). It is estimated that > 20 million United States (US) adults have CKD. ACS evaluation in a patient with CKD is complex; initial c-Tn levels using a standard cTn assay often exceed the 99th percentile, elevated troponin levels may be the result of chronic structural heart disease (e.g. heart failure) rather than acute myocardial ischemia, and ECG repolarization abnormalities are common3. Sensitive, and high-sensitivity c-Tn assays (hs-cTn is available in Europe and some parts of the world) further increase the number of patients with initial nitial abnormal troponin values, and raise the question as to whether or not the 99 9th ppercentile erce er cent ce ntil nt ilee is il the he appropriate CDV to use for CKD patients presenting with ACS. To address add d reess this question, Twerenbold et al. al te ttested sted 3 sensitive and nd 4 high-sensitivity cTnaassays ssaays in the Ad Adva Advantageous v nttag va ageo eous eo us P Predictors redi re dict di ctor ct o s of Ac Acu Acute ut C ute Coronary orrona nary na ry S Syndrome yndr yn d om dr me Ev Eva Evaluation aluaati tion on ((APACE) APAC AP ACE) AC E) multicenter prospective observational registry, MII in ppatients with multtic mu i enter pr prospe pectiv pe ve obse erv vat a io ona nal re regi g sttry ry,, examining exam min nin ng the th he frequency f eque fr uenc ue n y ooff M atienntss wit at th CKD, CKD CK D, ((eGFR eGFR eG FR < 660 0 ml ml/m ml/min/1.73m /min /m in/1 in /1.73m 73m2) an andd without wiith with thou outt CKD CKD4. A strength str tren engt en gthh of A gt APACE PACE PA CE iiss th that at aall ll cc-Tn -Tn Tn assays were measured in a core lab, patients were enrolled within 12 hours of symptom onset, the protocol specified that blood samples be taken at presentation, and 1, 2, 3 and 6 hours until the diagnosis of MI was ruled in/out, and the end-point of myocardial infarction was adjudicated by a clinical event committee. The committee used the Universal Definition of Myocardial Infarction, serial levels of hs-cTnT using the Roche assay, all available medical records including whether c-Tn levels were elevated on prior admissions, ECG, and angiographic findings when available, to classify events. Myocardial infarction was considered if at least one hs-cTnT value exceeded the 99th percentile. A significant absolute change in hs-cTnT was
3 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
defined as a rise or fall of at least 10 ng/l within 6 hours or a change of 6 ng/l within 3 hours. Of the 2,813 APACE patients, 447 (16%) had CKD. Adjudicated MI was the final diagnosis in 36% vs. 18% of the CKD vs. non-CKD patients. Receiver operator characteristics (ROC) analysis revealed diagnostic accuracy for MI diagnosis was less in patients with vs. without CKD. Optimal ROC-derived cTn cut-off levels in patients with renal dysfunction were significantly higher compared to patients with normal renal function (factor 1.9-3.4). For sensitive c-Tn assays, the optimal cut-point determined by ROC analysis in the CKD cohort approximated the manufacturer’s 99th percentile whereas, the optimal cut-point was significantly higher than the 99th percentile for the 4 hs-cTn assays. Long-term survival was worse in patients with elevated Tn levels using the more sensitive c-Tn assays, similar to data publi published liish shed ed uusing sing si ng lless ess sensitive ensitive assays3. CKD patients that ruled out for adjudicated MI had higher baseline levels of cTn than tha hann non-CKD nonno n-CK nCKD CK D patients with all 7 cTn assays s; th the frequency of eleva vaate ted baseline values above assays; elevated the he 99th percent percentile til ile ra ranged ang nged ed ffrom rom ro m 12 12-7 12-71% -71% -7 1% ddepending ep pen ending g oon n th thee as ass assay say ty say ttype pe uused. s d. L se Level evel ev el ooff cT cTnn va valu values lues lu es we were eree inverse inversely seely ccorrelated orrelaateed to C or CKD KD sseverity. eveeriity. ity Thee findings Th find fi ndin nd ings in gs from fro rom m the the APACE APAC AP ACE AC E registry regi re gist gi stry st ry using usi sing si ng the the hshs cTn cTn assays assa as says sa yss indicate ind ndiica cate te that tha hatt patients pati pa tien ti ents en ts with CKD presenting with an ACS are more likely to have elevated cTn levels on initial presentation than was previously published using older cTn assays. Increased number of patients with elevated cTn will predictably lead to more cardiac consultations from the emergency room, and perhaps, more hospital admissions, raising the question as to whether the hs-cTn assays offer an advantage over the older cTn assays in these patients. Many institutions in Europe have already switched over to hs-cTn assays. Therefore, clinicians need to take into consideration the assay type used at their particular institution, degree of renal impairment, and initial and subsequent c-Tn levels when determining if a patient with CKD and ACS has had an MI. For
4 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
patients with CKD using hs-cTn assays (which are not available in the U.S. at the time of this editorial), a cut-point > 99th percentile should improve diagnostic accuracy. Other approaches such as using the delta (changing pattern of values) rather than absolute change in hs-cTn as used by Twerenbold et al. requires further study in a CKD population5,6. The 99th Percentile Upper Reference Limit The finding that different c-Tn assays are not biologically equivalent is well known due to biochemical differences between assays and the fact that various manufacturers use different reference populations to determine the 99th percentile URL. Several studies have shown the 99th percentile URL as reported by the manufacturer may not replicate values seen in large community-based cohort studies and can result in major differences in determinin determining ng th thee 99th percentile URL7. To further examine the issue of establishing a an appropriate CDV for hs-cTn assa ays iin n a ge gener ral population of patients presen ral nti ting ng with an ACS, Wildi di et al. studied 2300 assays general presenting ppatients atiients in the AP APAC APACE AC CE re regi registry, gist gi stry st ry, 47 ry 4733 (2 (21%)) off who whom om ha hadd an n aadjudicated djjudic icat ated at ed MI MI diagnosis diaagno di agno nosi s s8. Th si The he primary prrim imar a y outcome outc tccomee measure measure was was the t e percentage th p rccenta pe en age of patients patien pa nts ts with withh an adjudicated adj djud udiccated ud ed M MII or or nno o MI M diagnosis diag di agno ag nosi no siss at ppresentation si rese re sent se ntat nt atio at ionn usi io us using sing tthe he aapproved ppro pp rove ro ved ed ma manu manufacturer nufa fact fa ctur ct urer rer C CDV DV fo forr th thee hs hs-cTnI -cTn cTn TnII Ab Abbo Abbot bott an bo andd hshs cTnT Roche assays; frequency of discordant MI results with the two assays were compared. For hs-cTn levels measured at presentation and at all time points, hs-cTnI and hs-cTnT levels were closely correlated (r=0.813 and r=0.790), and had comparable accuracy for early diagnosis of MI. However, among the 473 adjudicated MI patients, 86 (18%) had discordant MI diagnoses using the approved CDV; 14.1% for women and 22.7% for men. Approximately 20% of the time, MI was diagnosed by one assay but not the other. The authors tested other sensitive and hscTn assays and obtained similar results, not confirming the superiority of one assay over the other, but rather illustrating that all assays are not biologically equivalent. Inconsistencies
5 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
between the hs-cTnI and hs-cTnT assay were able to be reduced to ~10% after adjustment of the cTnI CDV to approximate biological equivalence to the cTnT assay. The findings from both APACE papers in Circulation this week expand and enhance multiple prior observations from this group of investigators and point out additional data that should be considered when the 99th percentile is used to diagnose MI. APACE Registry An important consideration when evaluating data from the APACE registry is the cardiologists’ use of the Roche hs-cTnT assay to diagnose MI. The cardiologists used the 99th percentile URL of 14 ng/L as reported by the manufacturer. In population based studies, the 99th percentile URL for the Dallas Heart Study (DHS), the Atherosclerosis Risk in Communities (ARIC) (ARIIC) study, stu tudy dy,, an dy andd the he Cardiovascular Health Study (CHS) was 14, 21, and 28 ng/L, respectively7. Within each coho ort rt,, th thee 99 99th ppercentile ercentile URL value increasedd wi with t age and was highe herr in men. Thus, the he cohort, higher manufacturer’s m annufacturer’ss U URL RL L vvalue alue al ue ddoes oess not oe not alw always lways re lw replicate epliccatte va valu values lues lu es sseen eeen inn ot oth other her “hea her “healthy” ealt ea lthy lt hy”” pa hy pati patient tien ti en nt cohorts. co oho hort r s. In AP APAC APACE, CE, the he card cardiologists diol olog ol ogis og ists is t use uused sed aan n abso absolute olu utee cchange hannge nge de defined efi fine need as as a rrise i e or ffall is all off ccTnT al TnT T at least eas astt 10ng/l 10ng 10 ng/l ng /l within wit ithi it hinn 6 hours hi hour ho urs rs or a change cha hang ngee of 6 ng/l ng ng/ g/ll within wit ithi it hinn 3 hours. hi hour ho urs rs. It It is not not clear clea cl earr if these ea the hese se absolute abs bsol olut ol ute te numerical values optimize performance of the other assays tested, if gender considerations would improve performance, and if delta rather than absolute change would improve diagnostic accuracy5,6,9. These differences might explain some of the inter-assay differences observed in both papers as well as the relatively high rate (36%) of confirmed MI (mainly NSTEMI) in the CKD cohort4,8. The implications of the data from the APACE registry for clinical patient management and clinical trials that use MI as an end-point are substantial. Clearly, some patients presenting with non-ST elevation-ACS may be misclassified simply due to the fact that the manufacturer’s
6 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
reference population chosen to establish the 99th percentile URL is not representative of the population being studied. Further, with hs-cTn assays, certain patient phenotypes (e.g. CKD) may require higher thresholds than the 99th percentile or other algorithms such as delta change to optimize MI diagnostic accuracy. The use of a standardized approach for approval of hs-cTn assays that would allow clinicians or investigators to determine biologic equivalence across all of the multiple manufactured cTn assays would be invaluable. For the time being, clinicians involved in the daily management of ACS patients in a hospital using more sensitive assays should be aware of non ACS conditions associated with elevations in cTn, and able to determine the pretest likelihood of myocardial infarction from all available data to estimate posttest MI ikelihood. A standardized blood sampling protocol should be used to facilitate ea arl rlie ierr di ie diag agno ag nosi no ss likelihood. earlier diagnosis of MI, with samples drawn at initial assessment (time 0) and repeated at least 3 and 6 hours later later, as req equi eq uire ui redd to establish re esttab ablish the diagnosis for patientss wi with (or without) initia iall cTn above the CDV. ia required initial
Conflict C on nflict of Interest In ntere rest Disclosures: Dissclossurres: N None. onne..
References: 1. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, de Ferranti S, Després JP, Fullerton HJ, Howard VJ, Huffman MD, Judd SE, Kissela BM, Lackland DT, Lichtman JH, Lisabeth LD, Liu S, Mackey RH, Matchar DB, McGuire DK, Mohler ER III, Moy CS, Muntner P, Mussolino ME, Nasir K, Neumar RW, Nichol G, Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani SS, Willey JZ, Woo D, Yeh RW, Turner MB, on behalf of the American Heart Association Statistics Committee and Stroke Statistics Committee. Heart Disease and Stroke Statistics—2015 Update. A Report from the American Heart Association. Circulation. 2015;131:e29-e322. 2. Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD; the Writing Group on behalf of the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction. Third universal definition of myocardial infarction. Circulation. 2012;126:2020-2035. 3. Stacy SR, Suarez-Cuervo C, Berger Z, Wilson LM, Yeh HC, Bass EB, Michos ED. Role of troponin in patients with chronic kidney disease and suspected acute coronary syndrome. A
7 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
DOI: 10.1161/CIRCULATIONAHA.115.016848
systematic review. Ann Intern Med. 2014;161:502-512. 4. Twerenbold R, Wildi K, Jaeger C, Gimenez MR, Reiter M, Reichlin T, Walukeiwicz A, Gugala M, Krivoshei L, Marti N, Weidmann ZM, Hillinger P, Puelacher C, Rentsch K, Honegger U, Schumacher C, Zurbriggen F, Freese M, Stelzig C, Campodarve I, Bassetti S, Osswald S, Mueller C. Optimal cut-off levels of more sensitive cardiac troponin assays for the early diagnosis of myocardial infarction in patients with renal dysfunction. Circulation. 2015;131:XX-XXX. 5. Morrow DA, Bonaca MP. Real world application of “Delta” troponin. J Am Coll Card. 2013;62:1239-1241. 6. Jaffe AS, Moeckel M, Giannitis E, Huber K, Mair J, Mueller C, Plebani M, Thygesen K, Lindahl B, In search for the Holy Grail: Suggestions for studies to define delta changes to diagnose or exclude acute myocardial infarction: a position paper from the study group on biomarkers of the Acute Cardiovascular Care Association. Eur Heart J: Acute Cardiovasc Care. 2014;3:313-316. 7. Gore MO, Seliger SL, deFilippi CR, Nambi V, Christenson RH, Hashim IA, Hoogeveen Hoog Ho ogev og evee ev eenn RC, ee RC Ayers CR, Sun W, McGuire DK, Ballantyne CM, de Lemos JA. Age- and sex-depe sex-dependent pend pe nden nd entt up en uppe upper perr pe reference eference limits for the high-sensitivity cardiac troponin T assay. J Am Coll Cardiol. 2014;63:1441–1448. 8.. Wildi Wildi K, Gimenez Gimenez MR, Twerenbold R, Reichlin n T, Jaege Jaeger g r C, H Heinzelmann einzelmann A, Arnold C, Nelles N ellles ll B, Drueyy S, Haaf Haa a f P, Hillinger Hil illi ling li nger ng er P, P, Schaerli Scha haeerli N ha N,, Kr Kreutzinger reu euttzin tzin ingger ger P, T Tanglay angl an glay Y Y,, He Herr Herrmann rrma rr mannn T ma T,, Weidmann Rentsch Osswald Mueller Weid We i mann ZM, Krivoshei Krivos oshei L, L, Freese M, M, Stelzig Steelzzig C, C, Puelacher Puel elac el achherr C,, R ac entscch K, O sswa waald d S,, M uelleer myocardial related limitations current approach C Misdiagnosis C. Mis i diag gno nosis off myo yocard yo dia i l in iinfarction farc fa r tiionn rel elatted tto el o lim mittatio io ons ns ooff the the cu urrrent nt rregulatory e ullat eg atooryy ap pproa oach to clinical values cardiac troponin. o ddefine efin ef inee cl in cli inical al ddecision ecis ec isio ionn val lue uess fo for ca ard rdia iacc tr ia trop pon onin in. Circulation Cirrcul Ci rcul ulaationn 2015;131:XX-XXX. 201 0155;13 5;13 131:XX XX-X XX -X XXX X. 9. Reichlin T, Irfan A, Twerenbold R, Reiter M, Hochholzer W, Burkhalter H, Bassetti S, Steuer S, Winkler K, Peter F, Meissner J, Haaf P, Potocki M, Drexler B, Osswald S, Mueller C. Utility of absolute and relative changes in cardiac troponin concentrations in the early diagnosis of acute myocardial infarction. Circulation. 2011;124:136-145.
8 Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
Is the 99th Percentile the Optimal Reference Limit to Diagnose Myocardial Infarction with High Sensitivity Cardiac Troponin Assays in Patients with Chronic Kidney Disease? Bernard R. Chaitman Circulation. published online May 6, 2015; Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2015 American Heart Association, Inc. All rights reserved. Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circ.ahajournals.org/content/early/2015/05/06/CIRCULATIONAHA.115.016848
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Circulation is online at: http://circ.ahajournals.org//subscriptions/
Downloaded from http://circ.ahajournals.org/ at Charité - Universitaetsmedizin Berlin on May 6, 2015
