The role of troponin and albumin to assess myocardial dysfunction after cardiac surgery and
in the critically ill
van Beek, Dianne E.C.
DOI:
10.33612/diss.101333600
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
van Beek, D. E. C. (2019). The role of troponin and albumin to assess myocardial dysfunction after cardiac
surgery and in the critically ill. Rijksuniversiteit Groningen. https://doi.org/10.33612/diss.101333600
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
to
assess myocardial dysfunction
after
cardiac surgery
and in
the critically ill
Cover design and layout: © evelienjagtman.com printing: Ridderprint BV - www.ridderprint.nl
Printing of this thesis was financially supported by the University Medical Center Groningen © E.C. van Beek 2019
All rights are reserved. No part of this publication may be reproduced, stored, or transmitted in any form or by any means without permission of the copyright owners.
to
assess myocardial dysfunction
after
cardiac surgery
and in
the critically ill
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen
op gezag van de
rector magnificus prof. dr. C. Wijmenga en volgens besluit van het College voor Promoties.
De openbare verdediging zal plaatsvinden op Maandag 2 december 2019 om 16.15 uur
Door:
Egidia Christina van Beek Geboren op 27 juli 1985 te Eindhoven
Prof. dr. M.M.R.F. Struys Prof. dr. J.C.C. van der Horst Beoordelingscommissie: Prof. dr. J.E. Tulleken Prof. dr. W.F. Buhre Prof. dr. M.L. Bots
1. General introduction 7
PART I Optimizing the use of troponin as a marker for excessive myocardial damage after cardiac surgery
2. Implementation of the third universal definition for myocardial
infarction after coronary artery bypass grafting in Western Europe 19 3. How typical is the typical rise and fall of troponin in (peri-procedural)
myocardial infarction 35
4. Improving the use of troponin to assess the risk for mortality after
cardiac surgery: a retrospective cohort study 57
PART II The role of albumin in myocardial dysfunction in the intensive care unit 5. Albumin, a marker for postoperative myocardial damage 75
6. Predictive value of serum albumin levels at ICU admission in a mixed ICU population on noradrenaline and fluid requirements in the first 24 hours
91
7. Low serum albumin and new-onset atrial fibrillation in the ICU:
a prospective cohort study 107
8. Validation of serum albumin as a prognostic marker in the intensive
care unit: a SICS-I sub-study 117
Appendix
9. Summary and general discussion 131
10. Acknowledgements | Dankwoord 145
Chapter
General
introduction
1
Background
Worldwide approximately one million people require cardiac surgery each year, and this need is only likely to increase further with the aging society in the following years.1 Major adverse cardiac events (MACE) resulting from myocardial damage in general and postoperative myocardial infarction (PMI) in particular, are major health concerns after cardiac surgery. About 10% of all patients getting cardiac surgery suffer from PMI.2,3,4 However, the proportion of patients actually diagnosed with PMI is substantially lower in clinical practice, with a substantial percentage of PMIs remaining unrecognized. Underdiagnoses of PMI is a major healthcare problem, because myocardial damage and subsequent MACE relates to an increased mortality and an increased intensive care and hospital length of stay.4,5,6 When PMI is detected, treatment options are readily available to prevent recurrence and/or ongoing damage of the myocardium. However, false positive diagnoses could lead to both worse prognostication and treatment of patients without PMI, who do not benefit from treatment but are exposed to side-effects, therewith introducing harm.
The first challenge with diagnosing PMI is that, according to the Third Universal Definition, there are five distinct types of MI7:
▪ Type 1: spontaneous MI
▪ Type 2: MI due to ischemic imbalance
▪ Type 3: MI resulting in death with no biomarker available ▪ Type 4a: MI related to percutaneous coronary intervention (PCI) ▪ Type 4b: MI related to stent thrombosis
▪ Type 5: MI related to coronary artery bypass grafting (CABG)
The cornerstone for diagnosis of (P)MI is elevated biomarkers. The currently preferred biomarker for detecting myocardial damage is troponin (Tn) subtype T or I, since they are not only highly sensitive but also very specific for myocardial injury.7 Although elevated Tn levels have been shown to be an excellent diagnostic marker for type 1 MI, interpreting elevated Tn levels after cardiac surgery has proven to be more challenging. The reason for this is that elevations of Tn are common after cardiac surgery8 and regarded as inherent to cardiotomy, i.e. elevated levels can be present even in absence of PMI2. On the other hand, elevations of Tn levels have repeatedly been shown to be associated with MACE8;9;10;11;12 and mortality10;13. How to distinguish between acceptable and unacceptable Tn elevation after cardiac surgery remains unclear. Currently, the recommended cut-off level for Tn after coronary artery bypass surgery (CABG) to diagnose excessive myocardial damage is 10 times the 99th percentile.7 However, this cut-off level is expert based and there is not yet solid evidence whether this cut-off level results in a distinction between acceptable and unacceptable Tn elevation.
We hypothesize that: early identification of patients at risk after cardiac surgery allows for pre-emptive measures to reduce morbidity and mortality and to redirect medical resources to those who might benefit and away from those in which it might induce harm. Adding an additional biomarker for myocardial damage and dysfunction will not only allow to better identify the patients at risk but will provide a framework for designing new preventive and therapeutic measures if a causal relationship can be established. We hypothesized that serum albumin (SA) is a marker and potential causal factor for myocardial damage and dysfunction. A low SA has been associated with an increase in morbidity and mortality in cardiac patients (table 1).
Table 1. The association of low SA with morbidity and mortality in different cardiac populations.
Population Low SA
Patients with stable coronary artery disease ▪ ↑ MACE14;15;16
▪ ↑ Mortality16;17
Patients with acute coronary syndromes ▪ ↑ MACE16
▪ ↑ Heart failure18
▪ ↑ (Cardiac) mortality18;16;19;20
Patients undergoing cardiac surgery ▪ ↑ Blood transfusion21
▪ ↑ Infection21
▪ ↑ Acute kidney injury21;22
▪ ↑ Hospital and ICU stay23
▪ ↑ Mortality21;24;23 MACE: major adverse cardiac events, ICU: intensive care unit
Albumin infusion in patients undergoing cardiac surgery has shown to reduce the risk of several adverse outcomes:
▪ ↓ Positive fluid balance25
▪ ↓ Fluid boluses needed25
▪ ↓ Norepinephrine dosage required25 ▪ ↓ Acute kidney injury26
▪ ↓ Readmission rate27 ▪ ↓ In-hospital mortality27
In patients admitted to the intensive care unit (ICU) a low SA is also associated with mortality.28;29 SA supplementation studies in the ICU patients have mostly been focused on septic patients, in which a meta-analysis showed no clear effect on mortality30. However, in a study focusing on the entire ICU population, the patients with a low SA did benefit from albumin infusion (as it was associated with an improved Sequential Organ Failure Assessment score and a less positive fluid balance).31
1
In addition to these associations, there are also signs that SA potentially has a direct effect on the heart. For instance, when comparing 20% albumin administration intravenously to crystalloid administration in healthy volunteers both the cardiac output and stroke volume increased more (while the afterload decreased).32 In patients after cardiac surgery 5% albumin infusion compared to saline infusion significantly increased the cardiac index33, and compared to Ringer’s lactate it prolonged the obtained hemodynamic stability after infusion34. In endotoxemic rats albumin infusion has been shown to improve ventricular contractility and the myocardial oxygenation.35
Even more so, in patients undergoing a percutaneous coronary intervention (PCI) a lower SA was associated with a prolonged QTc interval (regardless whether the PCI was elective or emergent).36 A (consistently) prolonged QTc interval is associated with the development of atrial fibrillation (AF)37 and sudden cardiac death36.
These direct effects on the heart of SA combined with the clinical association described above (of a low SA and adverse outcomes and albumin infusion with improved outcome), warrant to test the hypothesis that SA might be an important prognostic and/or causal factor for myocardial dysfunction in the critically ill patients and those after cardiac surgery.
Objective of this thesis
The main objective of this thesis is to improve the identification of the patients most at risk after cardiac surgery and in the ICU. To achieve this objective, we first wanted to optimize the use of Tn to assess myocardial damage in patients undergoing cardiac surgery. Second, we wanted to assess the role of SA, as merely a prognostic or potentially a causal factor for myocardial damage and dysfunction in the intensive care in patients undergoing both cardiac and non-cardiac surgery and in medical ICU patients. Finally, our objective was to replicate some of our observations on SA in a distinct larger cohort.
Outline of this thesis
The first part of this thesis focuses on improving the use of Tn as a marker for excessive myocardial damage after cardiac surgery. To do this, we first studied how different cardiac surgery centers in Western Europe currently monitor (excessive) myocardial damage and what their attitude towards the currently recommended criteria is (chapter 2). We subsequently conducted a systematic review to study whether the kinetics of Tn in a PMI (MI type 5) are different from the kinetics of Tn in MI type 1 and MI type 4 (chapter 3).
Understanding the unique characteristics of the kinetics of Tn in type 5 MI, provides the opportunity to form hypothesis about new and improved cut off points for the diagnosis of type 5 MI. In chapter 4, we evaluated the prognostic value of four different Tn measurements to determine which has the highest prognostic value for mortality after cardiac surgery. The methods of Tn analysis that were tested were selected according to the results of our systematic review (chapter 3) and the current available literature. In the second part of this thesis we focused on the association of SA and various outcomes after cardiac surgery and in critically ill patients (table 2). As a first step we evaluated the possibility of an etiological association between low levels of SA and Tn release in patients after cardiac surgery (chapter 5). Subsequently, we evaluated the association between SA and myocardial dysfunction in non-cardiac surgical and medical ICU patients in a prospective cohort (chapter 6). Myocardial dysfunction was defined by the need for vasoactive agents, the need for fluids and elevated arterial lactate blood levels. In chapter 7, we focused on SA and the association with another symptom of myocardial dysfunction in the ICU, namely new-onset atrial fibrillation (NOAF). Finally, we validated several of the strongest associations we found in an independent prospective cohort (chapter 8). This includes the associations between SA and myocardial damage (chapter 5), and SA and symptoms of myocardial dysfunction (chapter 6).
The main results of these studies will be provided in a summary and subsequently a reflection on these results will be given in the general discussion (chapter 9).
Table 2. the role of SA and myocardial damage and different symptoms of myocardial dysfunction. Chapter Study population Focus Outcomes 5 Cardiac surgical ICU Retrospective cohort Myocardial damage ▪ Tn
6 Non-cardiac surgical and medical ICU
Prospective cohort Myocardial dysfunction ▪ Vasoactive agents ▪ Fluids
▪ Arterial lactate level ▪ Mortality
7 Non-cardiac surgical and medical ICU
Prospective cohort Myocardial dysfunction ▪ NOAF ▪ Mortality
8 Cardiac and non-cardiac surgical and medical ICU
Prospective cohort Myocardial damage and myocardial dysfunction
▪ Tn
▪ Arterial lactate level ▪ Mortality
1
References
1. Katlic MR, ed. Cardiothoracic Surgery in the Elderly.; 2011. ISBN: 978-1-4419-0892-6 2. Ramsay J, Shernan S, Fitch J, et al. Increased
creatine kinase MB level predicts postoperative mortality after cardiac surgery independent of new Q waves. J Thorac Cardiovasc Surg. 2005;129(2):300-306.
3. Chen JC, Kaul P, Levy JH, et al. Myocardial infarction following coronary artery bypass graft surgery increases healthcare resource utilization. Crit Care Med. 2007;35(5):1296-1301.
4. Croal BL, Hillis GS, Gibson PH, et al. Relationship between postoperative cardiac troponin I levels and outcome of cardiac surgery. Circulation. 2006;114(14):1468-1475.
5. Ramsay J, Shernan S, Fitch J, et al. Increased creatine kinase MB level predicts postoperative mortality after cardiac surgery independent of new Q waves. J Thorac Cardiovasc Surg. 2005;129(2):300-306.
6. Chen JC, Kaul P, Levy JH, et al. Myocardial infarction following coronary artery bypass graft surgery increases healthcare resource utilization. Crit Care Med. 2007;35(5):1296-1301.
7. Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of myocardial infarction. Eur Heart J. 2012;33(20):2551-2567.
8. Nesher N, Alghamdi AA, Singh SK, et al. Troponin after cardiac surgery: a predictor or a phenomenon? Ann Thorac Surg. 2008;85(4):1348-1354. 9. Stearns JD, Dávila-Román VG, Barzilai B, et
al. Prognostic value of troponin I levels for predicting adverse cardiovascular outcomes in postmenopausal women undergoing cardiac surgery. Anesth Analg. 2009;108(3):719-726. 10. Eigel P, van Ingen G, Wagenpfeil S. Predictive value
of perioperative cardiac troponin I for adverse outcome in coronary artery bypass surgery. Eur J Cardiothorac Surg. 2001;20(3):544-549. 11. Peivandi AA, Dahm M, Opfermann UT, et al.
Comparison of cardiac troponin I versus T and creatine kinase MB after coronary artery bypass grafting in patients with and without perioperative
myocardial infarction. Herz. 2004;29(7):658-664. 12. Jacquet L, Noirhomme P, El Khoury G, et al. Cardiac troponin I as an early marker of myocardial damage after coronary bypass surgery. Eur J Cardiothorac Surg. 1998;13(4):378-384. 13. Paparella D, Cappabianca G, Visicchio G, et al.
Cardiac troponin I release after coronary artery bypass grafting operation: effects on operative and midterm survival. Ann Thorac Surg. 2005;80(5):1758-1764.
14. Suzuki S, Hashizume N, Kanzaki Y, Maruyama T, Kozuka A, Yahikozawa K. Prognostic significance of serum albumin in patients with stable coronary artery disease treated by percutaneous coronary intervention. Lazzeri C, ed. PLoS One. 2019;14(7):e0219044.
15. Wada H, Dohi T, Miyauchi K, et al. Long-term clinical impact of serum albumin in coronary artery disease patients with preserved renal function. Nutr Metab Cardiovasc Dis. 2018;28(3):285-290. 16. Wada H, Dohi T, Miyauchi K, et al. Impact of serum albumin levels on long-term outcomes in patients undergoing percutaneous coronary intervention. Heart Vessels. 2017;32(9):1085-1092.
17. Chien S-C, Chen C-Y, Leu H-B, et al. Association of low serum albumin concentration and adverse cardiovascular events in stable coronary heart disease. Int J Cardiol. 2017;241:1-5.
18. González-Pacheco H, Amezcua-Guerra LM, Sandoval J, et al. Prognostic Implications of Serum Albumin Levels in Patients With Acute Coronary Syndromes. Am J Cardiol. 2017;119(7):951-958. 19. Xia M, Zhang C, Gu J, et al. Impact of serum albumin
levels on long-term all-cause, cardiovascular, and cardiac mortality in patients with first-onset acute myocardial infarction. Clin Chim Acta. 2018;477:89-93.
20. Plakht Y, Gilutz H, Shiyovich A. Decreased admission serum albumin level is an independent predictor of long-term mortality in hospital survivors of acute myocardial infarction. Soroka Acute Myocardial Infarction II (SAMI-II) project. Int J Cardiol. 2016;219:20-24.
21. Gassa A, Borghardt JH, Maier J, et al. Effect of preoperative low serum albumin on postoperative complications and early mortality in patients undergoing transcatheter aortic valve replacement. J Thorac Dis. 2018;10(12):6763-6770.
22. Findik O, Aydin U, Baris O, et al. Preoperative Low Serum Albumin Levels Increase the Requirement of Renal Replacement Therapy after Cardiac Surgery. Heart Surg Forum. 2016;19(3):123. 23. Bhamidipati CM, LaPar DJ, Mehta GS, et al.
Albumin is a better predictor of outcomes than body mass index following coronary artery bypass grafting. Surgery. 2011;150(4):626-634. 24. Hebeler KR, Baumgarten H, Squiers JJ, et al.
Albumin Is Predictive of 1-Year Mortality After Transcatheter Aortic Valve Replacement. Ann Thorac Surg. 2018;106(5):1302-1307.
25. Wigmore GJ, Anstey JR, St. John A, et al. 20% Human Albumin Solution Fluid Bolus Administration Therapy in Patients After Cardiac Surgery (the HAS FLAIR Study). J Cardiothorac Vasc Anesth. March 2019.
26. Lee E-H, Kim W-J, Kim J-Y, et al. Effect of Exogenous Albumin on the Incidence of Postoperative Acute Kidney Injury in Patients Undergoing Off-pump Coronary Artery Bypass Surgery with a Preoperative Albumin Level of Less Than 4.0 g/ dl. Anesthesiology. 2016;124(5):1001-1011. 27. Kingeter AJ, Raghunathan K, Munson SH, et al.
Association between albumin administration and survival in cardiac surgery: a retrospective cohort study. Can J Anesth Can d’anesthésie. 2018;65(11):1218-1227.
28. Pan S-W, Kao H-K, Yu W-K, et al. Synergistic impact of low serum albumin on intensive care unit admission and high blood urea nitrogen during intensive care unit stay on post-intensive care unit mortality in critically ill elderly patients requiring mechanical ventilation. Geriatr Gerontol Int. 2013;13(1):107-115.
29. Yin M, Si L, Qin W, et al. Predictive Value of Serum Albumin Level for the Prognosis of Severe Sepsis Without Exogenous Human Albumin Administration: A Prospective Cohort Study. J Intensive Care Med. 2018;33(12):687-694. 30. Jiang L, Jiang S, Zhang M, Zheng Z, Ma Y. Albumin
versus Other Fluids for Fluid Resuscitation in Patients with Sepsis: A Meta-Analysis. Chalmers JD, ed. PLoS One. 2014;9(12):e114666. 31. Dubois M-J, Orellana-Jimenez C, Melot C, et al.
Albumin administration improves organ function in critically ill hypoalbuminemic patients: A prospective, randomized, controlled, pilot study. Crit Care Med. 2006;34(10):2536-2540. 32. Bihari S, Wiersema UF, Perry R, et al. Efficacy and
safety of 20% albumin fluid loading in healthy subjects: a comparison of four resuscitation fluids. J Appl Physiol. 2019;126(6):1646-1660. 33. Ernest D, Belzberg AS, Dodek PM. Distribution
of normal saline and 5% albumin infusions in cardiac surgical patients. Crit Care Med. 2001;29(12):2299-2302.
34. Arya VK, Nagdeve NG, Kumar A, Thingnam SK, Dhaliwal RS. Comparison of Hemodynamic Changes After Acute Normovolemic Hemodilution Using Ringer’s Lactate Versus 5% Albumin in Patients on β-Blockers Undergoing Coronary Artery Bypass Surgery. J Cardiothorac Vasc Anesth. 2006;20(6):812-818.
35. Tokunaga C, Bateman RM, Boyd J, Wang Y, Russell JA, Walley KR. Albumin resuscitation improves ventricular contractility and myocardial tissue oxygenation in rat endotoxemia*. Crit Care Med. 2007;35(5):1341-1347.
36. Wu C-C, Lu Y-C, Yu T-H, et al. Serum albumin level and abnormal corrected QT interval in patients with coronary artery disease and chronic kidney disease. Intern Med J. 2018;48(10):1242-1251. 37. Zhang N, Gong M, Tse G, et al. Prolonged corrected
QT interval in predicting atrial fibrillation: A systematic review and meta-analysis. Pacing Clin Electrophysiol. 2018;41(3):321-327.
Part
I
Optimizing the use of
troponin as a marker for
excessive myocardial damage
after cardiac surgery
Chapter
Dianne van Beek, Bas van Zaane, Marc Buijsrogge, Wilton van Klei.
Journal of the American Heart Association 2015 Jan; 4(1). doi: 10.1161/JAHA.114.001401
Implementation
of the Third Universal
Definition of Myocardial
Infarction After Coronary
Artery Bypass Grafting:
A Survey Study in
Western Europe
Background: Diagnosing a postoperative myocardial infarction in patients undergoing
coronary artery bypass grafting is challenging, as the normally used criteria are more difficult to interpret. The rate of implementation of the consensus-based new diagnostic criteria for postoperative myocardial infarction proposed by the third universal definition of myocardial infarction is unknown. Therefore, the primary objective of this study was to address the implementation of the third universal definition of postoperative myocardial infarction following coronary artery bypass grafting.
Methods and Results: We conducted a web-based survey by sending 4 waves of invitations
via e-mail to cardiothoracic surgeons in 12 Western European countries. Of the 302 participating cardiothoracic specialists, from 182 different centers, 213 (71%) were aware that troponin is the preferred biomarker and 112 (37%) knew that using a cut-off level of >10 times the 99th percentile is recommended. Overall, 90 (30%) participants (strongly) agreed with implementation of this cut-off level in their clinical practice. Troponin was used in clinical practice by 149 (49%) of the participants. In total, 117 (89%) of the 131 participants with a local guideline confirmed ECG changes as a diagnostic criterion in that guideline. ST segmental changes (75, 64%) were used more often for diagnosing postoperative myocardial infarction than Q waves (64, 55%) or new left bundle branch blocks (34, 29%).
Conclusions: Cardiac biomarkers and ECG changes were not used in concordance with the
third universal definition, and only a minority had a positive attitude toward implementation of the proposed cut-off level for troponin in their clinical practice.
2
Introduction
Approximately 1 million patients undergo cardiac surgery each year worldwide, and 7% to 15% of them will suffer from a postoperative myocardial infarction (PMI),1–3 mainly because of early graft failure.4–5 PMI after cardiac surgery is not only associated with an increased length of hospital stay, but also with a reduced short- and long-term survival.1–3 Diagnosing PMI in cardiac surgery patients is difficult, since pain from the sternal wound and the prescribed opioids may mask the typical symptoms (eg, chest pain, shortness of breath). Furthermore, postoperative changes of the ECG are not uncommon due to direct myocardial damage from the surgery, and postoperative pericarditis.6 Cardiac-specific biomarkers, such as creatine-kinase M-band and troponin (Tn), are normally used to identify myocardial damage. However, biomarker levels after cardiac surgery can easily be above the cut-off value due to direct and indirect myocardial injury from the surgical trauma or from reperfusion injury after cardiopulmonary bypass without a true PMI being present.1
The third universal definition of myocardial infarction provides, by arbitrary convention, diagnostic criteria for the different types of infarction. For PMI (ie, type 5 myocardial infarction associated with coronary artery bypass grafting [CABG]), the cornerstone for the diagnosis is the presence of a biomarker value of >10 times the 99th percentile of the upper reference limit.7 The preferred biomarker is Tn, because of its high sensitivity and specificity and because of its typical rise and fall pattern in myocardial infarction.7 The accuracy of the consensus-based cut-off level for this patient group is unclear, as is the ideal cut-off level for patients with an already elevated biomarker level preoperatively. More importantly, it is not clear whether PMI is always clinically relevant and when re-intervention is required. For these reasons, it cannot be expected that this definition is implemented in daily clinical practice without reservation.
The objective of this study was to address the clinical implementation of the third universal definition of PMI according to cardiothoracic surgeons in Western Europe. The focus of this survey was on the implementation of the consensus-based diagnostic criteria for biomarkers. The clinical implementation of other diagnostics mentioned in the universal definition, such as ECG changes, imaging, symptoms, and consultations, was also addressed.
Methods
Considering the nature of the study, the institutional review committee waived the requirement for medical ethical review committee approval and for informed consent (reference number WAG/om/14/005019).
Study Population
For this web-based survey, we invited cardiothoracic surgeons from a total of 12 Western European countries (Austria, Belgium, Denmark, France, Germany, Ireland, Luxembourg, Netherlands, Norway, Sweden, Switzerland, and the United Kingdom). We chose to include the 10 countries with the highest government expenditure on health care per capita, according to the most recent available data from Eurostat (2008–2010). The United Kingdom and Ireland were included as well. Even though no recent information on healthcare expenditure was available on Eurostat for these 2 countries, it is conceivable that their healthcare expenditure is in the top 10.
The 1627 cardiothoracic surgeons from the included countries were recruited for voluntary participation in this nonanonymous questionnaire using standardized e-mails. Cardiac centers and cardiothoracic surgeons were identified, and e-mail addresses were retrieved by conducting a web-based search that was focused on the European cardiothoracic societies, national cardiothoracic societies, national registries for medical specialists, network websites, patient organizations, hospitals, and relevant publications. When the search did not result in identification of an e-mail address of a cardiothoracic surgeon, that person was excluded from participation (n=291; 18%), as it was considered not feasible to contact all cardiac centers throughout Europe to identify the missing e-mail addresses. The questionnaire was distributed to 1336 cardiac surgeons in 4 waves to optimize the response rate. No incentives were given for participation.
Questionnaire
The questionnaire consisted of 16 questions (Data S1) and was validated by a focus group of colleagues from the University Medical Center Utrecht. Members of the group include a professor in intensive care medicine, a cardiac-anesthesiologist, and an epidemiologist. The questionnaire was evaluated by this focus group on 13 points (ie, 5 general points and 8 quality-related points) including the formulation of the questions, the construction of the survey, the layout, the user-friendliness, and time required to complete the survey. The quality of the content was addressed by validating the face validity (ie, a global evaluation on whether the survey measured what it should measure), and the content validity (ie, in-depth evaluation of whether the survey provided adequate coverage of the topic). The reliability of the survey was addressed on stability (whether repeated measurements on the
2
same individual would yield similar results), equivalence (whether a measure interpreted by different investigators would yield similar results), and homogeneity (whether changing the operational definitions used would change the results).
Knowledge and Attitude
Voluntary implementation has to be supported by adequate knowledge. For successful implementation of the third universal definition for PMI regarding biomarkers, participants had to know about the proposed cut-off level and the preferred biomarker. Complete knowledge for implementation was defined in this study as the percentage of participants answering affirmatively to both “Were you aware that according to the third universal definition of myocardial infarction troponin is the preferred biomarker for the diagnosis of perioperative myocardial infarction?” and to “Were you aware that the third universal definition of myocardial infarction defines the cut-off level for troponin at >10 times the 99th percentile?”
Another prerequisite for successful voluntary implementation is a positive attitude toward the use of the medical guideline. In this study this was addressed by asking participants whether they agreed with the statement: “A cut-off level of troponin >10 times 99th percentile for diagnosis of perioperative myocardial infarction in patients undergoing CABG should be implemented in your local guideline.” Participants could answer on a 5-point Likert scale (strongly disagree, disagree, neutral, agree, and strongly agree). The answers “agree” or “strongly agree” were arbitrarily considered a positive attitude toward implementation.
Clinical Practice
The questionnaire addressed which participants had a local guideline concerning PMI and which diagnostic tools and criteria were mentioned in that local guideline. All participants were asked to rank the following diagnostic tools on importance for diagnosing PMI: biomarkers, consultation of other specialist, ECG changes, imaging, and symptoms. Analysis
The results were analyzed using frequencies and proportions and presented graphically. Missing data due to unanswered questions can provide relevant information in implementation research and were therefore not excluded from the analysis. This is, because item nonresponse can be due to high sensitivity of the topic, not knowing the answer, or editing of the participant.8
The ranking scores for the importance of the different diagnostics were calculated per diagnostic as the sum of all the rankings from all the participants. When a diagnostic was ranked most important by a participant, 5 points were awarded, when it was ranked second 4 points were given, and so on.
Results
From the 1336 cardiac surgeons who were invited, there were 302 (23%) participants and 16 (1%) refusals. The time between the first and the last participation was 61 days. Participants represented 182 unique cardiac centers from all 12 countries included (table 1). The participation rate per country was 29%, with a SD of 16%. The extremes in the participation rate were seen in countries where only a limited number of participants were invited. In total, 288 of the 302 participants were traced back to the invitation list. The 14 other respondents participated either anonymously or via an invitation from a colleague. The remaining 1048 cardiac surgeons were arbitrarily considered nonresponders. About half of the responders and the nonresponders were from an academic hospital (48% and 45%, respectively). Most participants were from centers performing 100 to 500 or 500 to 1000 CABG per year, and the majority indicated a PMI incidence of 0% to 3% in their hospital.
Knowledge, Attitude, and Clinical Practice
Of all 302 participants, 109 (36%) knew both the consensus-based cut-off level and that Tn is the preferred biomarker (table 2). The majority of the 302 participants (213, 71%) knew that Tn is the preferred biomarker, and the proposed cut-off level was known by 112 participants (37%). Forty-nine participants (16%) had complete knowledge and also a positive attitude toward implementation. Overall, 90 (30%) participants agreed or strongly agreed with implementation of the cut-off level of >10 times the 99th percentile in their clinical practice. For 45 participants (15%) the data regarding knowledge and attitude were missing. In ranking the diagnostic criteria on importance, the most common order was biomarkers first, ECG changes second, and imaging third (37, 12%). Participants who had this order of ranking had less often a negative attitude toward implementation compared to participants with a different ranking order (figure 1). The presence of a local guideline concerning PMI in CABG was reported by 131 (43%) participants.
Biomarkers
From the 131 participants with a local guideline, 120 (92%) mentioned the use of biomarkers for diagnosing PMI. Overall, 184 (61%) of all 302 participants determined biomarkers at least twice in all patients regardless of any symptoms, while 31 (10%) only determine biomarkers on indication, whereas 6 (2%) never determined biomarkers (table 3). The majority of 302 participants used a combination of more than 1 biomarker (174, 58%). The combination of creatine-kinase M-band and Tn was the most popular 1 (79, 45%). The majority did not use Tn but creatine-kinase M-band (149, 49% versus 202, 67%). Ten participants (3%) stated the use of liver function enzymes (eg, aspartate
2
aminotransferases, alanine transaminase) in the comment space as other biomarker. Other biomarkers that were mentioned included myoglobin (4, 1%), total creatine kinase (2, 1%), and lactate dehydrogenase (1, <1%). Biomarkers were ranked as the most important diagnostic criterion by 122 (40%) of the 302 participants (table 4).
Table 1. Characteristics of the 302 participants.
N (%) Complete participation 257 (85) Anonymous 14 (5) Profession Cardiothoracic surgeon 278 (92) Other 23 (8) Missing data 1 (<1)
CABG per year in hospital of participant
100 to 500 108 (36) 500 to 1000 122 (40) Other 38 (13) Missing data 34 (11) Country Austria 9 (3) Belgium 24 (8) Denmark 14 (5) Germany 99 (33) France 16 (5) Ireland 1 (<1) Luxembourg 2 (1) Netherlands 39 (13) Norway 9 (3) Sweden 13 (4) Switzerland 24 (8) United Kingdom 50 (17)
CABG: coronary artery bypass grafting.
Electrocardiography
A total of 117 (89%) of the 131 participants with a local guideline confirmed the presence of ECG criteria for PMI in that local guideline. ST segmental changes were used most often (75, 64%), followed by Q waves (64, 55%), new left bundle branch blocks (34, 29%), T-wave inversions (25, 21%), and new R-wave progression (12, 10%). Other criteria that were mentioned included ventricular arrhythmias (2, 2%), T-wave decrease (1, 1%), extrasystoles (1, 1%), and all ECG changes (1, 1%).
Table 2. Knowledge regarding biomarkers and attitude toward implementation. The proportion of participants
who either agreed or strongly agreed with implementation compared with the proportion that does not for 4 different groups. Group 1. Tn: participants who only knew that Tn is the preferred biomarker; Group 2. Cut-off: participants who only knew the cut-off level; Group 3. Both: participants who knew both that Tn is the preferred biomarker and the cut-off level; and Group 4. Neither: participants who knew neither that Tn is the preferred biomarker nor the cut-off level.
Knows Attitude Total, N (%)
(Strongly) Agrees With Implementation, N (%)
Does Not Agree With Implementation, N (%)
Tn 29 (10) 75 (25) 104 (34)
Cut-off level 1 (<1) 2 (1) 3 (1)
Both Tn and cut-off level 49 (16) 60 (20) 109 (36)
Neither 11 (4) 30 (10) 41 (14)
Total 90 (30) 167 (55) 257 (85)*
Tn: troponin. *Data missing from 45 participants (15%).
Figure 1. Ranking and attitude toward implementation. Responses in percentages to the question: “Do you
agree with the following statement? A cut-off level of troponin >10x 99th percentile for diagnosis of perioperative myocardial infarction in patients undergoing CABG should be implemented in your local guideline.” The answers “strongly disagree” and “disagree” were considered a negative attitude and “agree” or “strongly agree” were considered a positive attitude. Comparison was made of the attitude toward implementation of participants who ranked biomarkers 1, ECG changes 2, and imaging 3 on importance as a diagnostic criterion to the participants who ranked the diagnostic criteria in any other order. ECG indicates electrocardiogram.
0% 10% 20% 30% 40% 50% 60%
Negative
Neutral
Positive
2
Table 3. Biomarker use for diagnosis PMI. The question how often biomarkers are determined standardly in all
postoperative patients was a select 1 question (counts to 100%). The question about which biomarkers are used was a multiple-select question (counts to >100%).
How often are biomarkers determined? N (%)
Never 6 (2%) Only on indication 31 (10%) 1 time 0 2 times 0 >3 times 63 (21%) Until decreasing 121 (40%) Missing data 81 (27%)
Which biomarkers are used? N (%)
CK-MB 202 (67%)
Tn 149 (49%)
HS-Tn 105 (35%)
Other 17 (6%)
Missing data 81 (27%)
CK-MB: creatine-kinase MB, HS-Tn: high-sensitive troponin; PMI: postoperative myocardial infarction, Tn: troponin
Table 4. Ranking scores diagnostic criteria. Ranking score per diagnostic criterion and the proportion of
participants who ranked a certain category as most important (ranked #1) and as least important (ranked #5).
Ranking Score Ranked #1, N (%) Ranked #5, N (%) Missing Data, N (%) Biomarkers 1030 122 (40) 8 (3) 57 (19) ECG 963 77 (25) 6 (2) 55 (18) Imaging 681 23 (8) 32 (11) 66 (22) Symptoms 562 26 (9) 48 (16) 81 (27) Consultations 352 1 (<1) 112 (37) 95 (31) ECG: electrocardiogram Other
Of the 131 participants with a local guideline, 85 (65%) included imaging. Transthoracic echocardiography (53, 62%), transesophageal echocardiography (43, 51%), and angiography (34, 40%) were used more commonly than magnetic resonance imaging scans (4, 5%) and computed tomography scans (3, 4%). Over 1 third (52, 40%) of the 131 local guidelines were reported to include symptoms for PMI, such as ischemic symptoms, hemodynamic instability, low cardiac output, and hypotension. Table 5 provides the definitions of these symptoms as defined by the participants in the comment space in the survey.
Consultation of another specialist was the criterion that was ranked most often as least important by the 302 participants (112, 37%) and it had the most missing data (95, 31%). Consultations of specialists were part of the diagnostic criteria in the local guidelines of 46 (35%) of the 131 participants with guidelines. Cardiologists (28, 61%), Cardiothoracic surgeons (24, 52%), and Intensive Care specialists (15, 33%) were the specialists of choice for consultations.
Table 5. Criteria used to define symptoms. Participants who indicated to use either ischemic symptoms,
hemodynamic instability, low cardiac output, and/or hypotension as a diagnostic criterion were asked the follow-up (open text) question on how they defined these criteria. In this table, the most commonly given response in the open text field on how participants defined ischemic symptoms, hemodynamic instability, low cardiac output, and hypotension is given.
Ischemic symptoms (N=36) (Chest) pain
ECG changes
New wall-motion abnormalities Hemodynamic instability (N=32) Catecholamine use
Low blood pressure
Tachycardia/rhythm disturbances Impaired renal function/oliguria Low cardiac output (N=22) Low cardiac index
Low (systolic) blood pressure Low mean arterial pressure (Chest) pain
Hypotension (N=10) Low (systolic) blood pressure Catecholamine use
ECG: electrocardiogram.
Discussion
In this cross-sectional survey, we aimed to address the clinical implementation of the third universal definition of myocardial infarction after CABG according to cardiothoracic surgeons in Western Europe. We found that the majority of the 302 participating surgeons from 182 different cardiac centers used different biomarkers and different ECG criteria than recommended by the third universal definition. In addition, most participants (55%) did not have a positive attitude toward the implementation of the third universal definition in their clinical practice.
2
Comparison With Literature
This is, to the best of our knowledge, the first study addressing the implementation of the third universal definition of myocardial infarction for diagnosing PMI after CABG. We found a self-reported implementation of knowledge and attitude of only 16%. A review evaluating clinical implementation of consensus-based guidelines found a physician self-reported implementation rate between 24% and 85%.9 The lack of implementation that we found might be due to the limited number of participants with sufficient knowledge regarding this topic (36%). Our study showed that the proportion of participants who agreed with implementation of the Tn cut-off level was lower when the participants had insufficient knowledge compared to participants with sufficient knowledge (i.e., knowledge about both Tn and the cut-off level). However, even in participants with sufficient knowledge, only a minority had a positive attitude toward implementation, indicating that there are probably more reasons for the lack of implementation demonstrated in this study. The first reason might be that the applicability of the consensus-based Tn cut-off level is inadequate. This is demonstrated by a recent study showing that patients with Tn levels >10 times the 99th percentile in combination with either specific ECG changes or echocardiographic criteria had an increased 30-day mortality.10 Although patients with an unstable preoperative Tn level were excluded in this study, still 93% had a Tn level >10 times the 99th percentile.10 So, if indeed >90% of the CABG patients meet this criterion, this could mean that the cut-off level, the cornerstone for the diagnosis, is not usable in clinical practice. Especially since the recent development of the high-sensitivity Tn tests, there is a general concern for the specificity of these new highly sensitive tests.11 Therefore, some of the reluctance against implementation could be toward the consensus-based Tn criterion. Second, the diagnostics criteria that were implemented were not used as recommended by the third universal definition. For instance, the majority of the participants used local guidelines that mention the use of ST segmental changes as ECG criterion for the diagnosis of PMI (64%), while the third universal definition recommends using Q waves and/or new left bundle branch blocks. The use of ST segmental changes is remarkable, as ST-segment elevations are seen regularly after CABG in the absence of myocardial infarction and are not associated with the peak Tn level or adverse outcomes.6 A third reason for the lack of (correct) implementation can also be due to indifference toward diagnosing PMI. This can be the case if PMI is considered a standard or irrelevant complication. Indifference could explain the relatively high percentage of missing data (27%) on the questions regarding the type of biomarker used and the frequency of taking lab samples. Since it is possible that participants with indifference are more likely to not know the answers to these questions, not knowing the answer to a question can result in missing data.8 Finally, for a successful implementation program it is crucial that knowledge be increased,
attitudes be changed positively (physicians have to agree with the implementation), behavior be changed, and the patient outcome be affected.12 The effect on outcome of this new guideline specifically and of diagnosing PMI in general is not clear yet, which means that currently there is not a solid foundation for implementation.
Limitations
Our study has limitations. First, potential participants were identified by conducting a web-based search. When contact information was not available on the website of the hospital, e-mail addresses were searched in scientific publications. This could result in overestimation of the implementation of knowledge, as it is possible that researchers are more exposed to publications regarding the third universal definition. Second, despite sufficient efforts the final participation rate was 23%. In a recent web-based survey study aimed at cardiothoracic surgeons, the final response rate was 16%,13 indicating that our response rate is ample. The risk of a low response rate is nonresponse bias, and it is unknown which participation rate is minimally required in web-based surveys to avoid it. Web-based surveys with a low response rate (<35%) were shown to be representative.14 In this study, we reduced the risk of nonresponse bias by selecting multiple participants per cardiac center and by recruiting participants from 182 different cardiac centers. The lower risk of nonresponse bias makes the participation rate of 23% acceptable. Third, directive “yes” or “no” questions were used to assess knowledge, making the survey vulnerable to participants providing sought-after answers. The use of these lead-in questions could have resulted in bias. Moreover, participating in the survey resulted in a learning effect and the survey was not protected against editing previous answers, also allowing for an overestimation of the implementation of knowledge. However, the results demonstrated a large difference between the knowledge regarding the cut-off level and Tn as the preferred biomarker (37% versus 71%); therefore, it is unlikely that participants were editing or giving sought-after answers. Fourth, applicable participants were asked in an open text question how they defined ischemic symptoms, hemodynamic instability, low cardiac output, and/ or hypotension as a diagnostic criterion for PMI. This open text question likely resulted in bias due to under-reporting.
Clinical Implications
The results of this study can be used as a first step toward designing an implementation program. However, for effective implementation of a clinical guideline, either the quality of care or patient outcome needs to be improved.12 Although PMI is associated with an adverse outcome,1–3 it is not clear whether improving diagnostics by using the third universal definition will positively affect patient care or outcome. Therefore, further research, focused on patient outcome, seems to be required first to provide a solid foundation for successful implementation. Such a study will need to investigate not only
2
the added diagnostic value of the third universal definition, but also the clinical significance of using this definition in routine practice. In addition, it would be relevant to study the use of the third universal definition in research, as the results of such a study would allow for the comparison of the implementation in research and clinical practice.
Conclusions
This was a cross-sectional, survey-based study, including 302 cardiothoracic surgeons from 182 European cardiac centers. The implementation of the third universal definition for PMI in Western Europe is limited. In clinical practice, different ECG criteria and different (combinations of) biomarkers are used for the diagnosis of a PMI following CABG. In addition, less than 1 third of the participants agreed with implementation of the cut-off level for cardiac biomarkers as defined in the third universal definition, indicating that there currently does not seem to be consensus for the implementation of the third universal definition regarding biomarker use for diagnosing PMI.
References
1. Ramsay J, Shernan S, Fitch J, Finnegan P, Todaro T, Filloon T, Nussmeier NA. Increased creatine kinase MB level predicts postoperative mortality after cardiac surgery independent of new Q waves. J Thorac Cardiovasc Surg. 2005; 129:300-306.
2. Chen JC, Kaul P, Levy JH, Haverich A, Menasché P, Smith PK, Carrier M, Verrier ED, Van de Werf F, Burge R, Finnegan P, Mark DB, Shernan SK. Myocardial infarction following coronary artery bypass graft surgery increases healthcare resource utilization. Crit Care Med. 2007; 35:1296-1301.
3. Croal CBL, Hillis GS, Gibson PH, Fazal MT, El-Shafei H, Gibson G, Jeffrey RR, Buchan KG, West D, Cuthbertson BH. Relationship between postoperative cardiac troponin I levels and outcome of cardiac surgery. Circulation. 2006; 114:1468-1475.
4. Laflamme M, DeMey N, Bouchard D, Carrier M, Demers P, Pellerin M, Couture P, Perrault LP. Management of early postoperative coronary artery bypass graft failure. Interact Cardiovasc Thorac Surg. 2012; 14:452-456.
5. Thielmann M, Massoudy P, Jaeger BR, Neuhäuser M, Marggraf G, Sack S, Erbel R, Jakob H. Emergency re-revascularization with percutaneous coronary intervention, reoperation, or conservative treatment in patients with acute perioperative graft failure following coronary artery bypass surgery. Eur J Cardiothorac Surg. 2006; 30:117-125.
6. Loeb HS, Gunnar WP, Thomas DD. Is new ST-segment elevation after coronary artery bypass of clinical importance in the absence of perioperative myocardial infarction? J Electrocardiol. 2007; 40:276-281.
7. Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD, Katus HA, Apple FS, Lindahl B, Morrow DA, Chaitman B, Clemmensen PM, Johanson P, Hod H, Underwood R, Bax JJ,
Bonow RO, Pinto F, Gibbons RJ, Fox KA, Atar D, Newby LK, Galvani M, Hamm CW, Uretsky BF, Steg PG, Wijns W, Bassand JP, Menasché P, Ravkilde J, Ohman EM, Antman EM, Wallentin LC, Armstrong PW, Januzzi JL, Nieminen MS, Gheorghiade M, Filippatos G, Luepker RV, Fortmann SP, Rosamond WD, Levy D, Wood D, Smith SC, Hu D, Lopez-Sendon JL, Robertson RM, Weaver D, Tendera M, Bove AA, Parkhomenko AN, Vasilieva EJ, Mendis S. Third universal definition of myocardial infarction. Eur Heart J. 2012; 33:2551-2567.
8. Brick J, Kalton G. Handling missing data in survey research. Stat Methods Med Res. 1996; 5:215-238.
9. Lomas J. Words without action? The production, dissemination, and impact of consensus recommendations. Annu Rev Public Health. 1991; 12:41-65.
10. Wang TK, Stewart RA, Ramanathan T, Kang N, Gamble G, White HD. Diagnosis of MI after CABG with high-sensitivity troponin T and new ECG or echocardiogram changes: relationship with mortality and validation of the Universal Definition of MI. Eur Heart J Acute Cardiovasc Care. 2013; 2:323-333.
11. CADTH Optimal Use Reports. High-Sensitivity Cardiac Troponin for the Rapid Diagnosis of Acute Coronary Syndrome in the Emergency Department: A Clinical and Cost-Effectiveness Evaluation Ottawa Can Agency Drugs Technol Heal 2013 Mar. 2013.
12. Conroy M, Shannon W. Clinical guidelines: their implementation in general practice. Br J Gen Pract. 1995; 45:371-375.
13. D’Amico TA, McKneally MF, Sade RM. Ethics in cardiothoracic surgery: a survey of surgeons’ views. Ann Thorac Surg. 2010; 90:11-13.e4. 14. Bennett L, Nair CS. A recipe for effective
participation rates for web-based surveys. Assess Eval Higher Educ. 2010; 35:357-365.
Chapter
Dianne van Beek, Bas van Zaane, Marjolein Looije, Linda Peelen, Wilton van Klei.
World Journal of Cardiology
2016;8(3):293. doi:10.4330/wjc.v8.i3.293
The typical
rise and fall of troponin
in (peri-procedural)
myocardial infarction,
a systematic review
Background: The typical rise and fall of cardiac troponin (Tn) is crucial for the diagnosis of
myocardial infarction (MI). However, the exact shape of the rise and fall curve is unknown. The aim of this systematic review was to identify the typical shape of the rise and fall curve of Tn following the different types of MI.
Methods: We conducted a systematic search in PubMed and EMBASE including all
studies which focused on the kinetics of Tn in MI type 1, type 4 and type 5. Tn levels were standardized using the 99th percentile, a pooled mean with 95% confidence interval (CI) was calculated from the weighted means for each time point until 72 hours.
Results: A total of 34 of the 2528 studies identified in the systematic search were included.
The maximum peak level of the Tn was seen after 6 hours after successful reperfusion of an acute MI, after 12 hours for type 1 MI and after 72 hours for type 5 MI. In type 1 MI there were additional smaller peaks at 1 hour and at 24 hours. After successful reperfusion of an acute MI there was a second peak at 24 hours. There was not enough data available to analyze the Tn release after MI associated with PCI (type 4).
Conclusions: The typical rise and fall of Tn is different for type 1 MI, successful reperfusion
of an acute MI and type 5 MI, with different timing of the peak levels and different slopes of the fall phase.
3
Introduction
Myocardial infarction (MI) is the collective term for myocardial necrosis in the setting of myocardial ischemia1. There are many different conditions which can result in myocardial ischemia and subsequent MI. Currently, there are five distinct types of MI defined: type 1 spontaneous MI related to atherosclerotic plaque rupture, type 2 MI secondary to an imbalance between oxygen supply and oxygen demand, type 3 MI resulting in death when biomarkers are not available, type 4a MI related to percutaneous coronary intervention (PCI), type 4b MI related to stent thrombosis, and type 5 MI related to coronary artery bypass grafting (CABG)1.
For all different types of MI, excluding type 3, cardiac biomarkers are the cornerstone for diagnosing its occurrence. The preferred cardiac biomarker for the detection of myocardial damage is troponin (Tn)1. Troponin (subtypes I en T) is part of the contractile apparatus of myocardial cells only and is therefore a highly specific biomarker for myocardial damage1. Elevated levels of Tn can be detected within 3-12 hours after the start of ischemia and they reach a peak after 12-48 hours2. However, as Tn is a structural component of myocardial cells, Tn levels will be elevated in patients with chronic heart conditions such as heart failure as well. Therefore, to distinguish between an acute MI and chronic cardiac disease, elevation of Tn alone is not specific enough. There needs to be a significant change in the level of Tn, i.e. a rise and/or a fall. In spontaneous MI a relative difference of more than 20% is considered a significant change1. More specifically, in spontaneous MI any level above the 99th percentile is considered a rise1. The cut off levels according to the third universal definition for a typical rise in PCI associated MI (>5 times 99th percentile) and CABG associated MI (>10 times 99th percentile) are consensus based and not evidence based1.
The typical rise and/or fall of Tn is thus crucial for the diagnosis of MI1. However, the exact shape of the rise and fall curve is largely unknown. Nevertheless, understanding the shape of the rise and fall curve would allow for better timing of Tn blood sampling in clinical practice and would improve diagnostic criteria per type of MI. The aim of this systematic review was to identify the typical shape of the rise and fall curve of Tn following the different types of MI.
Methods
Literature search
Medline (PubMed) and Embase were searched from 1966 through October 2013 for publications. We used synonyms and abbreviations for ‘rising’, ‘falling’, ‘changing’, ‘troponin’ and ‘myocardial infarction’ as keywords (see online supplementary 1 for search strategies). Based on titles and abstracts, all studies evaluating troponin in MI were included. Different types of studies were eligible, for example cross sectional studies of patients with MI, cohort studies including patients with symptoms of cardiac ischemia, randomized controlled trials concerning treatment or diagnosis of MI and case control studies where the cases had MI. We included studies in patients with MI that focused on cardiac troponin, both I and Tn-T, and that reported at least two different Tn-values with at least one sample above the cut off level. Abstracts from conference proceedings, non-human studies, non-English studies, and studies on animals, children, chronic conditions and cardiomyopathy were excluded. First, all titles and abstracts were screened for eligibility. Second, screening was extended to full text for all studies that where either marked as relevant or when the eligibility was unclear from screening titles and abstracts. Eligibility was determined using a standardized form containing the above-mentioned criteria.
The methodological quality of included studies was assessed by two observers (DvB and ML) and in case of doubt by a third observer (BvZ) using an adjusted QUADAS-tool3 (see supplementary 2 for quality criteria). The selected items of the QUADAS-tool enabled us to examine potential sources of bias and variation4. The defined quality domains were; representativeness of the spectrum (i.e. the representativeness of the patients in the study for clinical practice), acceptable reference standard, acceptable delay between tests, partial verification avoided, relevant clinical information, uninterpretable results reported, and withdrawals explained. We did not calculate summary scores estimating the overall-quality of included studies since it has been shown that their interpretation is problematic and may be misleading5.
Data extraction took place using a specifically designed data extraction form. The two observers independently extracted raw data from the included studies to obtain information on Tn levels at different time points. Other elements that were extracted included the year of publication, the type of study, the research question, any subgroups, inclusion and exclusion criteria, the setting (e.g. emergency department, in hospital, post-surgery) and sample size. In addition, the proportion of patients with MI, the mean or median age of patients with MI, the proportion of males with MI, any comorbidities and the diagnostic criteria used for MI were obtained. Finally, test characteristics were extracted
3
such as the type of Tn test, the 99th percentile / upper reference limit / cut off level of the Tn test, limit of detection, number of samples per patient and the sample time points in relation to the event (e.g. admission, surgery).
Data were considered missing if not explicitly mentioned in the text and if impossible to deduct the information directly from other information in the text. Discrepancies between the two observers were resolved by discussion.
Statistical analysis
Studies were divided into four subgroups based on the focus of the articles: studies on type 1 spontaneous MI, studies that focused on successful reperfusion in the setting of an acute MI (where reperfusion was not initiated or its effect not evaluated), studies on MI associated with PCI (type 4a MI), and studies on MI associated with CABG (type 5 MI) . Type 2 MI studies were not included in this systematic review as the etiology behind this type of MI is distinctly different.
In this review we aimed to address the general rise and fall of Tn and not the rise and fall of specific Tn tests. Therefore, all Tn levels that were obtained within 72 hours were included in our analysis. If the timing of the samples was not specified, the study was excluded from analysis. If only one data source was available for a given point in time, we excluded this time-point from our analysis.
For each time point up till 72 hours we conducted the following procedure:
For each study, we first determined the mean and standard deviation (SD) of the Tn values. If available, mean and SD as presented in the article were used. Alternatively, when only a median was available the mean was approximated. For articles with less than 25 patients with MI, we used the formula of Hozo et al. to approximate the mean, for articles with 25 or more patients with MI, the median was used as the best estimate of the mean6. Articles for which the mean could not be approximated were excluded from analysis. When the standard error (SE) was not available from the articles directly, it was calculated from SD, confidence interval (CI), or median absolute deviation (MAD). Articles for which the SE was not available nor could be calculated were excluded from the analysis.
Subsequently, in order to make the Tn levels from different studies comparable, all Tn levels were standardized. Standardization was achieved by dividing the Tn levels by the 99th percentile of that particular Tn test. If the 99th percentile was not available, we used the upper reference limit (URL) or the cut off value for standardization. Studies that did not mention a 99th percentile or an URL or a cut off value for their Tn test were excluded from analysis.
After standardization, results over studies were pooled as follows. Every study was assigned a weight according to the inverse of the variance (𝑆𝑆𝑆𝑆12). The weighted mean per article was calculated by multiplying the mean with the weight. The sum of all weighted means was divided by the sum of all weights to calculate a pooled mean for every timepoint. The SE per timepoint was calculated as follows:𝑆𝑆𝑆𝑆𝑆𝑆 𝑜𝑜𝑜𝑜 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤1 0.5. From the pooled SE the 95% confidence interval (CI) was calculated.
The pooled mean of the standardized Tn levels with the corresponding CI at different time points were analyzed and summarized using a graph.
Results
Search results
Our search resulted in 2528 potentially eligible studies (figure 1). After screening titles and abstracts 2189 studies were excluded. After reviewing and applying the in- and exclusion criteria to the full text of the remaining 339 studies, 34 studies remained for analysis. There were 17 studies on type 1 spontaneous MI, 8 on successful reperfusion, 1 on MI associated with PCI (type 4), and 9 studies on MI associated with CABG (type 5). One study could be included in the analyses for both type 1 MI and reperfusion. The baseline characteristics of the included studies are summarized in table 1.
Quality of the included studies
Table 2 describes the results of the quality assessment. Almost all studies avoided partial verification, worked with relevant clinical information and a representative spectrum of patients with MI. Very few studies reported uninterpretable results or explained withdrawals.
Typical rise and fall of Tn
The pooled mean Tn level in type 1 MI showed an early first peak of 7.0 (CI 6.0-8.0) at 1 hour. This initial peak was followed by a maximum pooled mean Tn level of 84 (CI 82-86) at 12 hours. A third small peak followed at 24 hours (2.7 CI 2.6-2.9) (figure 2). Finally, there was a gradual fall of Tn.
The maximum pooled mean of Tn after successful reperfusion was at 6 hours (1853; CI 1851–1855), another high peak followed at 24 hours (1006 CI 1004-1007) (figure 3). Subsequently, there was a pronounced fall in Tn. The pooled mean Tn in type 5 MI associated with CABG raised the first 24 hours, after which the Tn levels stabilized (figure 4). The maximum pooled mean level of Tn was at 72 hours (2.2 CI 1.8-2.6).
3
Figure 1. Flow chart. MI= type 1 spontaneous myocardial infarction, RP= successful reperfusion during an acute
myocardial infarction, PCI: type 4 myocardial infarction associated with percutaneous coronary intervention, CABG= type 5 myocardial infarction associated with coronary artery bypass surgery. * different data from one study has been included in both the MI and RP analysis
2528 studies
339 studies 52 studies
MI: 17 studies* RP: 8 studies* PCI: 1 study CABG: 9 studies
Title/abstract screening: 2189 studies
excluded
Full text screening: 287 studies excluded
34 studies Excluded from
analysis: 18 studies excluded
Table 1. Baseline characteristics of included studies. First author Year of
publication Number of patients Prevalence MI N (%) Males with MI N (%) Diagnostic criteria MI
Tn test Cut off level Type of cut off level
Time points measured from Type 1: Spontaneous myocardial infarction (MI)
Aldous12 2011 939 200 (21) NA Biomarkers ECG Imaging Symptoms HS-TnT (T) HS-TnI (I) (T) 0.014 μg/L (I) 0.028 μg/L (T): 99th (I): 99th Admission Aldous13 2012 385 82 (21) 59 (72) Biomarkers ECG Imaging Symptoms TnI (I) HS-TnT (T) (T): 0.014 μg/L (I): 0.028 μg/L (T): 99th (I): 99th Admission al-Harbi14 2002 86 51 (59) 46 (90) ECG Symptoms TnI 0.05 ng/mL 99th Admission Apple15 2009 381 52 (13) NA ESC ACC TnI 0.034 μg/L 99th Admission Bahrmann16 2013 306 38 (12) 23 (61) Biomarkers ECG Imaging Symptoms HS-TnT 14 ng/L 99th Admission
Bertinchant17 1996 682 48 (7) 41 (85) WHO TnI 0.1 μg/L cut off Admission
Biener18 2013 459 111 (3) 82 (74) WHO UD HS-TnT 14 ng/mL 99th Admission Bjurman19 2013 1504 1178 (75) 716 (61) Biomarkers ECG Imaging Symptoms HS-TnT 40 ng/L 99th Admission de Winter20 2000 131 131 (100) NA Biomarkers Symptoms TnT 0.1 μg/L URL Symptoms
Falahati21 1999 327 62 (19) NA WHO TnT 0.20 μg/L cut off Symptoms
Haaf22 2012 887 127 (14) 87 (69) Biomarkers ECG Imaging Symptoms HS-TnT (HT) HS-TnI (HI) TnI (I) (HT): 0.014 μg/L (HI): 0.009 μg/L (I:) 0.009 μg/L (HT): 99th (HI): 99th (I:) 99th Admission Lucia23 2001 82 42 (51) 32 (76) Biomarkers ECG Symptoms
TnI 1.5 ng/mL URL Admission
Mohler24 1998 100 21 (21) NA Biomarkers
ECG Symptoms
TnT 0.1 mg/L cut off Admission
3
Table 1. Baseline characteristics of included studies. First author Year of
publication Number of patients Prevalence MI N (%) Males with MI N (%) Diagnostic criteria MI
Tn test Cut off level Type of cut off level
Time points measured from Type 1: Spontaneous myocardial infarction (MI)
Aldous12 2011 939 200 (21) NA Biomarkers ECG Imaging Symptoms HS-TnT (T) HS-TnI (I) (T) 0.014 μg/L (I) 0.028 μg/L (T): 99th (I): 99th Admission Aldous13 2012 385 82 (21) 59 (72) Biomarkers ECG Imaging Symptoms TnI (I) HS-TnT (T) (T): 0.014 μg/L (I): 0.028 μg/L (T): 99th (I): 99th Admission al-Harbi14 2002 86 51 (59) 46 (90) ECG Symptoms TnI 0.05 ng/mL 99th Admission Apple15 2009 381 52 (13) NA ESC ACC TnI 0.034 μg/L 99th Admission Bahrmann16 2013 306 38 (12) 23 (61) Biomarkers ECG Imaging Symptoms HS-TnT 14 ng/L 99th Admission
Bertinchant17 1996 682 48 (7) 41 (85) WHO TnI 0.1 μg/L cut off Admission
Biener18 2013 459 111 (3) 82 (74) WHO UD HS-TnT 14 ng/mL 99th Admission Bjurman19 2013 1504 1178 (75) 716 (61) Biomarkers ECG Imaging Symptoms HS-TnT 40 ng/L 99th Admission de Winter20 2000 131 131 (100) NA Biomarkers Symptoms TnT 0.1 μg/L URL Symptoms
Falahati21 1999 327 62 (19) NA WHO TnT 0.20 μg/L cut off Symptoms
Haaf22 2012 887 127 (14) 87 (69) Biomarkers ECG Imaging Symptoms HS-TnT (HT) HS-TnI (HI) TnI (I) (HT): 0.014 μg/L (HI): 0.009 μg/L (I:) 0.009 μg/L (HT): 99th (HI): 99th (I:) 99th Admission Lucia23 2001 82 42 (51) 32 (76) Biomarkers ECG Symptoms
TnI 1.5 ng/mL URL Admission
Mohler24 1998 100 21 (21) NA Biomarkers
ECG Symptoms
TnT 0.1 mg/L cut off Admission
