Atherosclerosis 241 (2015) 87e91
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Myocarditis or “true” infarction by cardiac magnetic resonance in patients with a clinical diagnosis of myocardial infarction without obstructive coronary disease: A meta-analysis of individual patient data P. Tornvall a, *, E. Gerbaud b, A. Behaghel c, R. Chopard d, O. Collste a, E. Laraudogoitia e, € rensson h, S. Agewall i G. Leurent c, N. Meneveau d, M. Montaudon f, E. Perez-David g, P. So €dersjukhuset, Karolinska Institutet, Sweden Cardiology Unit, Department of Clinical Science and Education So ^pital du Haut L Soins Intensifs Cardiologiques, Plateau de Cardiologie Interventionnelle, CHU de Bordeaux, Ho ev^ eque, 5 Avenue de Magellan, F33604 Pessac, France c CHU de Rennes, Service de Cardiologie et Maladies Vasculaires, INSERM, U1099, Universit e de Rennes 1, LTSI, Rennes, France d Department of Cardiology, EA 3920, University Hospital Jean Minjoz, 25000 Besancon, France e Hospital Galdakao, Vizcaya, Spain f ^pital du Haut L Unit e d'Imagerie Thoracique et Cardiovasculaire, CHU de Bordeaux, Ho ev^ eque, 5 Avenue de Magellan, F33604 Pessac, France g ~ on Madrid, Spain Hospital Gregorio Maran h Cardiology Unit, Department of Medicine, Karolinska University Hospital, Karolinska Institutet, Sweden i Department of Cardiology, Oslo University Hospital Ullevål, Institute of Clinical Medicine, University of Oslo, Oslo, Norway a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 October 2014 Received in revised form 29 April 2015 Accepted 30 April 2015 Available online 1 May 2015
Objective: Myocardial Infarction with Non-Obstructed Coronary Arteries (MINOCA) is common, but the causes are to a large extent unknown. Thus, we aimed to study the prevalence of myocarditis and “true” myocardial infarction determined by cardiac magnetic resonance (CMR) imaging in MINOCA patients, and risk markers for these two conditions in this population. Methods: A search was made in the PubMed and Cochrane databases using the search terms “Myocardial infarction”, “Coronary angiography”, “Normal coronary arteries” and “MRI”. All relevant abstracts were read and seven of the studies fulfilled the inclusion criteria; studies describing case series of patients fulfilling the diagnosis of acute myocardial infarction with normal or non-obstructive coronary arteries on coronary angiography that were investigated with CMR imaging. Data from five of these studies are presented. Results: A total of 556 patients from 5 different sites were included. Fifty-one percent were men with a mean age of 52 ± 16 years. Thirty-three per cent of the patients had myocarditis (n ¼ 183), whereas 21% of the patients had infarction on CMR (n ¼ 115). Young age and a high CRP were associated with myocarditis whereas male sex, treated hyperlipidemia, high troponin ratio and low CRP were associated with “true” myocardial infarction. Conclusion and relevance: The results of this meta-analysis of individual data showed that myocarditis and “true” myocardial infarction are common in MINOCA when determined by CMR imaging. This information emphasizes the importance of performing CMR imaging in MINOCA patients and can be used clinically to guide diagnostics and treatment of MINOCA patients. © 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Myocardial infarction Normal coronary angiography Cardiac magnetic resonance imaging Myocarditis Infarction
1. Introduction
* Corresponding author. Department of Clinical Science and Education €dersjukhuset, Karolinska Institutet, So € dersjukhuset, Sjukhusbacken 10, 118 83 So Stockholm, Sweden. E-mail address: [email protected] (P. Tornvall). http://dx.doi.org/10.1016/j.atherosclerosis.2015.04.816 0021-9150/© 2015 Elsevier Ireland Ltd. All rights reserved.
Myocardial Infarction with normal or Non-Obstructive Coronary Arteries (MINOCA) is common with a prevalence of 1e12% of all patients with a clinical diagnosis of myocardial infarction [1]. Recently, Collste et al. [2] suggested a prevalence of 7e8 %. MINOCA is a heterogeneous condition with patients fulfilling the diagnostic
88
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91
criteria of acute myocardial infarction. There are several different sub-groups of MINOCA, including myocarditis, “true” myocardial infarction and Takotsubo stress cardiomyopathy (TSC). Taking into account the heterogeneity of the condition, the mechanisms behind MINOCA are to a large extent unknown [3]. Cardiac magnetic resonance (CMR) imaging has been shown by several investigators to be of value in the investigation of MINOCA patients [2,4e9]. CMR imaging makes it possible to differentiate between myocarditis, “true” myocardial infarction and TSC although the latter is dependent on a short delay from symptom to investigation and clinical findings. The prevalence of myocarditis has varied from 7 to 63 % [2,6] whereas the prevalence of “true” myocardial infarction has varied from 12 to 30 % [4,5]. The studies have been small and heterogeneous regarding patient characteristics, in particular age and sex have varied between studies. The size and heterogeneity of the studies makes it difficult to find out the true prevalence and risk markers of myocarditis and “true” myocardial infarction among MINOCA patients investigated with CMR imaging. Thus, a large sample size is needed to study CMR imaging results in MINOCA. The presents study, including consecutive individual data from five studies [2,6e9], aimed at establishing the prevalence and risk markers for myocarditis and “true” myocardial infarction among MINOCA patients investigated with CMR imaging. Our hypothesis was that it was possible to find clinical markers to predict the occurrence of myocarditis and “true” myocardial infarction. 2. Material and methods A search was made by two of the authors (SA, PT), who independently of each other performed the search in PubMed using the MeSH search terms “Myocardial infarction”, “Coronary angiography” and “MRI” with the addition of the free search term “Normal coronary arteries”. An additional search was made in the Cochrane Library. All relevant abstracts were read and five of the studies described case series with all patients fulfilling the diagnosis of acute myocardial infarction with normal or non-obstructive coronary arteries that were investigated with CMR imaging [2,6e9]. Two more publications were found by reading references from the relevant publications [4,5]. In total, seven relevant studies with similar methodology were found, and the corresponding authors were invited to collaborate within the present study. The final study was able to combine data from five of those studies in the present analysis. Informed consent was obtained from corresponding and senior authors. Original data was transferred between researchers after agreement by e-mail. All included studies were approved by the local ethics committees. 3. Study designs All studies included consecutive cases of the respective time periods of inclusion from the respective catchment area of the involved centers [2,6e9]. The inclusion criteria were similar although there were differences regarding the angiographic
definitions. Exclusion criteria were similar although one study excluded patients with a high coronary heart disease (CHD) Framingham risk score whereas another study included only patients 35e70 years old (Table 1). 4. Risk markers An Excel sheet was prepared and sent to the interested authors of the included studies. The requested background information included age, sex, history of present smoking and diabetes mellitus, treatment for hypertension and dyslipidemia, and the ratio of maximal troponin (I or T) increase to the upper limit of normal and C-reactive protein (CRP) level in plasma. In cases where the CRP level was given as < 5 mg/l, the value was set to 3. Information about smoking, diabetes mellitus and treated dyslipidemia was missing in four patients. Information about treated hypertension was missing in five patients. Information about troponin ratio was missing in five patients whereas data were lacking in 87 (42 with myocarditis, respectively 16 with infarction) patients regarding CRP. 5. CMR imaging The CMR imaging protocols were similar in all studies and included cine images, T2-weighted oedema sequences and delayed enhancement after gadolinium contrast injection. Left ventricular ejection fraction was determined by the respective software of the CMR equipment (data from 11 patients were missing). The evaluation of CMR images varied between the studies. However, the final CMR imaging diagnosis was based mainly on the myocardial distribution of delayed enhancement of gadolinium in order to differentiate between myocarditis (subepicardial) and infarction (subendocardial). TSC was diagnosed clinically by Zaldumbide el al [6]. and Collste and al [2]. and by a combination of CMR imaging and clinical data in the remaining studies [7e9]. Hence, a CMR imaging diagnosis was only possible for myocarditis and infarction. In all studies, the final CMR imaging diagnosis was made by two experts blinded to clinical data. For information regarding detailed protocols see references [2,6e9] and for en overview of the included studies, see Table 1. 6. Statistics Values are presented as percentage, mean ± standard deviation or median (interquartile range). Uni- and multivariable odds ratios (OR) with 95% confidence intervals (CI) of risk markers for myocarditis and infarction were calculated using logistic regression. For this analysis, the material was divided into clinically relevant categories: troponin ratio 10 or >10 and CRP 10 or >10 mg/l. Uni- and multivariable analyses of the OR of cardiac risk factors for myocarditis and infarction determined by CMR imaging were performed. In model 1 age, sex, smoking, diabetes mellitus, hypertension, dyslipidemia, troponin ratio and CRP were included and model 2 included all variables in model 1 with the addition of
Table 1 Characteristics of studies included in the meta-analysis.
Ref. Ref. Ref. Ref. Ref.
[6] [7] [9] [8] [2]
Number of patients
Age, years
Male sex %
CA findings
Restrictions
Time from AMI to MRI
MRI equipment
80 107 87 130 152
48 44 53 54 58
64 63 40 52 36
Non-obstructive Non-obstructive Normal Non-obstructive Normal
No major No major Low risk score* No major Age 35e70
3 days 5 days 10 days 6 days 12 days
Philips® or Siemens® 1.5 T Philips® 3 T GE® 3 T Siemens® 1.5 T GE®, Philips® or Siemens® 1.5 T
* Framingham risk score > 10% in 10 years. No ¼ Number, CA ¼ Coronary angiography, AMI ¼ Acute Myocardial Infarction, MRI ¼ Magnetic Resonance imaging, GE ¼ General Electric, T ¼ Tesla.
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91 Table 2 Patient characteristics of all patients and patients divided into myocarditis and infarction determined by cardiac magnetic resonance (CMR) imaging.
Age (years) Male gender (%) Present smoking (%) Diabetes mellitus (%) Treated hypertension (%) Treated dyslipidemia (%) Ejection fraction (%) Troponin ratio CRP (mg/l)
All patients n ¼ 556
CMR myocarditis n ¼ 183
CMR infarction n ¼ 115
52 (39e65) 51% 36% 6% 30% 24% 59 (51e65) 40 (12e100) 6 (3e34)
37 (29e49) 78% 42% 1% 12% 16% 58 (51e65) 82 (33e225) 30 (7e61)
54 (44e64) 53% 42% 8% 29% 37% 57 (50e65) 85 (29e154) 5 (3e11)
Values are presented either as percent (%) or median (inter-quartile range). Troponin ratio is defined as the maximal troponin (I or T) divided by the upper limit of the reference value. CRP ¼ C-reactive protein.
site of investigation. The analyses presented were performed without imputation of missing values. As a sensitivity analysis, all analyses were repeated after imputation of missing CRP values with CRP 10 or >10 mg/l. All analyses were performed by a biostatistician using SPSS® version 22 for statistical analysis and p < 0.05 was considered significant. 7. Results Mean age of the patients was 52 ± 16 years. The CHD risk factor profile included approximately one third of active smokers and treated hypertension. The prevalence of diabetes mellitus was low (5%) and approximately one fourth of the patients had treatment for dyslipidemia. Thirty-three percent of the patients had myocarditis whereas 21% of the patients had infarction on CMR imaging (Table 2). Patients with a CMR imaging diagnosis of myocarditis were considerably younger and more often men than patients without myocarditis. With the exception for active smoking, that was more common in patients with myocarditis, the CHD risk factor profile was less pronounced when compared with patients without myocarditis. CRP was higher in patients with myocarditis than without (Tables 2 and 3). Patients with a CMR imaging diagnosis of infarction were of similar age and gender when compared with patients without infarction. The CHD risk factor profile was similar in patients with or without infarction with the exception for treated dyslipidemia that was more common in patients with infarction. Patients with infarction had a higher troponin ratio and a lower CRP than patients without infarction (Tables 2 and 4). 8. Multivariable analysis Age and a high CRP were associated with myocarditis when all
89
variables were included in the analysis simultaneously. There were no major changes in the strength of the associations when site of investigation was included in the multivariate analysis (Table 3). Furthermore, there were no major changes in the associations when CRP 10 or >10 mg/l were imputated for missing CRP values (data not shown). Male sex, treated hyperlipidemia, high troponin ratio (>10) and low CRP (10 mg/l) were all associated with infarction when all variables were included in the analysis simultaneously. High troponin ratio and low CRP showed the strongest associations with infarction with no major changes in the strength of the associations when site of investigation was included in the multivariate analysis (Table 4). Furthermore, there were no major changes in the associations when CRP 10 or >10 mg/l were imputated for missing CRP values (data not shown).
9. Discussion As far as we know this is the largest study on CMR imaging in MINOCA. The results showed that one third of the patients had myocarditis whereas one fifth of the patients had “true” myocardial infarction. The hypothesis was that it was possible to find clinical markers for the occurrence of myocarditis and “true” myocardial infarction. Several clinical markers that were different for the two conditions emerged. Young age and high CRP were associated with myocarditis whereas male sex, treated hyperlipidemia, high troponin ratio and low CRP were associated with “true” myocardial infarction. The prevalence of acute myocarditis was 33% in this selected group of patients with MINOCA. There was a large difference between the five included studies regarding the prevalence that varied from 7 % to 63 % depending on the site of investigation [2,6e9]. To the best of our knowledge there are no CMR imaging studies of unselected healthy subjects showing the prevalence of myocarditis. However, the diagnostic accuracy of CMR in relation to myocardial biopsy in suspected myocarditis has recently been shown to be good (79%) according to Lurz el al [10]. One autopsy study of unselected patients showed that the prevalence of myocarditis was below one percent [11]. Young age was a strong marker for myocarditis which is supported by an autopsy study by Gravani and Sternby [11] who reported that myocarditis patients were young. The other marker of myocarditis was an increased CRP, a finding that was expected but not convincingly shown before in this particular group of patients. The prevalence of “true” myocardial infarction was 21% with minor differences between studies (12e30%). This figure is comparably small considering that the patients fulfilled criteria for myocardial infarction and the notion that one population-based study of 75-year old subjects showed that 30% of 394 men and woman had unrecognized myocardial infarction determined by
Table 3 Uni- and multivariable analyses of the predictive values of cardiac risk factors for myocarditis determined by cardiac magnetic resonance (CMR) imaging.
Age, years (n ¼ 556) Male sex (n ¼ 286) Present smoking (n ¼ 197) Diabetes mellitus (n ¼ 31) Treated hypertension (n ¼ 163) Treated dyslipidemia (n ¼ 135) Troponin ratio > 10 (n ¼ 408) CRP > 10 mg/l (n ¼ 205)
Univariable
Multivariable model 1
Multivariable model 2
0.92 5.51 1.50 0.13 0.22 0.45 1.31 6.23
0.93 1.40 0.66 0.35 0.90 0.67 0.85 4.29
0.94 1.30 0.54 0.43 0.86 0.66 0.80 2.98
(0.90e0.93) (3.67e8.26) (1.04e2.16) (0.03e0.56) (0.14e0.37) (0.28e0.72) (0.87e1.99) (4.00e9.70)
(0.92e0.95) (0.81e2.45) (0.39e1.12) (0.07e1.81) (0.45e1.80) (0.36e1.27) (0.47e1.52) (2.56e7.18)
(0.92e0.96) (0.71e2.36) (0.30e0.98) (0.07e2.80) (0.42e1.75) (0.34e1.29) (0.42e1.51) (1.70e5.22)
Values are presented as odds ratio (confidence interval). Troponin ratio is defined as the maximal troponin (I or T) divided by the upper limit of the reference value. CRP ¼ Creactive protein. Model 1 included age, sex, smoking, diabetes mellitus, hypertension, dyslipidemia, troponin ratio and CRP. Model 2 included all variables in model 1 with the addition of site of investigation.
90
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91
Table 4 Uni- and multivariable analyses of the predictive values of cardiac risk factors for infarction determined by cardiac magnetic resonance (CMR) imaging.
Age, years (n ¼ 556) Male sex (n ¼ 286) Present smoking (n ¼ 197) Diabetes mellitus (n ¼ 31) Treated hypertension (n ¼ 163) Treated dyslipidemia (n ¼ 135) Troponin ratio > 10 (n ¼ 408) CRP > 10 mg/l (n ¼ 205)
Univariable
Multivariable model 1
Multivariable model 2
1.01 0.92 1.37 1.64 0.98 2.20 3.20 0.40
1.02 1.92 1.64 0.74 0.70 3.06 5.26 0.29
1.02 1.87 1.40 0.66 0.66 2.72 5.93 0.25
(0.99e1.02) (0.61e1.39) (0.90e2.09) (0.73e3.67) (0.62e1.54) (1.41e3.43) (1.73e5.92) (0.24e0.65)
(0.99e1.04) (1.08e3.38) (0.96e2.81) (0.24e2.28) (0.37e1.30) (1.76e5.33) (2.46e11.25) (0.17e0.51)
(0.99e1.04) (1.05e3.35) (0.80e2.47) (0.21e2.05) (0.35e1.26) (1.52e4.85) (2.69e13.04) (0.14e0.45)
Values are presented as odds ratio (confidence interval). Troponin ratio is defined as the maximal troponin (I or T) divided by the upper limit of the reference value. CRP ¼ Creactive protein. Model 1 included age, sex, smoking, diabetes mellitus, hypertension, dyslipidemia, troponin ratio and CRP. Model 2 included all variables in model 1 with the addition of site of investigation.
CMR imaging [12]. The reason for the low prevalence of “true” myocardial infarction is probably due to the high prevalence of myocarditis and TSC. The prevalence of reported TSC varied from 11 % to 22 % between the studies. TSC is associated with wall motion disturbances and oedema in the acute stage and a recent large CMR imaging study reported that only nine percent of the TSC patients had signs of permanent delayed gadolinium enhancement as a sign of infarction [13]. Unfortunately, it was not possible to study risk markers for TSC determined by CMR imaging since TSC was diagnosed clinically in two of the studies [2,6] and by a combination of CMR imaging and clinical data in three of the studies [7e9]. Our group of patients with “true” myocardial infarction was predominantly male patients with treatment for dyslipidemia with high troponin ratio and low CRP. The most likely reason for the strength of the marker CRP is that it can differentiate myocardial infarction from myocarditis [14]. Possibly, the mechanism behind MINOCA with “true” myocardial infarction is due to thromboembolism despite normal or near normal coronary arteries [3]. 10. Clinical implications The results of the present study emphasize the importance of CMR imaging in the investigation of MINOCA. It should in this context be acknowledged that only approximately 50% of the patients received a definitive diagnosis by CMR and other clinical information is of great value. Myocarditis is common and important to diagnose since the treatment differ from other causes of MINOCA. The new knowledge about risk markers for myocarditis is clinically relevant. For example, in a clinically stable young patient with high CRP fulfilling diagnostic criteria for acute myocardial infarction an alternative first major investigation could be CMR imaging. In the case of a myocarditis diagnosis on CMR imaging, coronary angiography could be omitted from further investigations decreasing the patient total burden of iatrogenic risk. These patients may also avoid to be treated with the standard acute coronary syndromes (ACS) pharmacological treatment. Regarding the risk markers for “true” myocardial infarction, the new knowledge that patients with MINOCA with high troponin ratio and low CRP are likely to show signs of infarction on CMR imaging emphasize the importance to investigate these patients with CMR imaging. Patients with a CMR imaging diagnosis of infarction should probably have similar pharmacological treatment as patients with ACS. Unfortunately, there is a knowledge gap how to treat these patients and thus an urgent need for clinical guidelines in this area. 11. Limitations The present study has several limitations. First, although the five studies are similar regarding inclusion criteria the study group
characteristics differ between sites, in particular regarding age and sex. The reason for this in not clear but is likely due to the threshold to perform coronary angiography in younger patients where myocarditis is a likely diagnosis. Furthermore, the time interval between presentation to hospital and CMR differed slightly between the studies, which might have affected the results. This interval will mainly affect the results of oedema sequences. However, the results were based on delayed enhancement of gadolinium that is less sensitive to the time interval. Furthermore, site of investigation was added as a variable into the multivariable analysis thus indirectly correcting the results for time interval. The results did not change when site of investigation was included. Second, there are differences regarding methods to analyze troponin and CRP. We tried to circumvent the difficulties with different troponin methods by analyzing the ratio of peak to upper level of normal. Despite this correction there are likely differences between the sites. Third, there are major differences between the sites regarding CMR imaging technology and evaluation of the images that might have influenced the results. Fourth, it is not possible to clearly make the distinction towards secondary type 2 myocardial infarction, which is a myocardial injury with necrosis, where a condition other than CAD contributes to an imbalance between myocardial oxygen supply and/or demand [15]. Furthermore, ECG and echocardiography are commonly performed in the clinical practice before performing cardiac MRI. Unfortunately, we have no echocardiography data and the quality of the ECG data is uncertain. However, some studies included in this meta-analysis have already shown the additive value of cardiac MRI to ECG and echocardiography. For example, in the study by Laraudogoitia Zaldumbide et al. [6], CMR provided a formal diagnosis in 61 patients (69.3%) in which the clinical diagnosis was uncertain between at least two possibilities. Anyway, the lack of echocardiography and ECG data is another limitation of the study. Although the analysis may be limited by the inability to include all studies of interest, we believe that differences regarding age, sex and prevalence of myocarditis between our results and the studies not included [4,5], are unlikely to have significantly affected our results. 12. Conclusion The results of this meta-analysis of individual data showed that myocarditis and “true” myocardial infarction are common in MINOCA when determined by CMR imaging. Several risk markers for the two conditions emerged with low age and high CRP being most important for myocarditis and high troponin and low CRP being highly relevant for infarction. This information emphasizes the importance of performing CMR imaging in MINOCA patients and can be used clinically to guide diagnostics and treatment of MINOCA patients.
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91
Conflict of interest None. [9]
Acknowledgment None of the authors had any relevant conflict of interest.
[10]
References [1] I. Kardasz, R. De Caterina, Myocardial infarction with normal coronary arteries:a conundrum with multiple aetiologies and variable prognosis: an update, J. Intern. Med. 261 (4) (2007 Apr) 330e348. €rensson, M. Frick, S. Agewall, M. Daniel, L. Henareh, [2] O. Collste, P. So C. Ekenb€ ack, L. Eurenius, C. Guiron, T. Jernberg, C. Hofman-Bang, K. Malmqvist, E. Nagy, H. Arheden, P. Tornvall, Myocardial infarction with normal coronary arteries is common and associated with normal findings on cardiovascular magnetic resonance imaging: results from the Stockholm myocardial infarction with normal coronaries study, J. Intern Med. 273 (2) (2013 Feb) 189e196. [3] S. Agewall, L. Eurenius, C. Hofman-Bang, K. Malmqvist, M. Frick, T. Jernberg, P. Tornvall, Myocardial infarction with angiographically normal coronary arteries, Atherosclerosis 219 (1) (2011 Nov) 10e14. [4] J.P. Christiansen, C. Edwards, T. Sinclair, G. Armstrong, A. Scott, H. Patel, H. Hart, Detection of myocardial scar by contrast-enhanced cardiac magnetic resonance imaging in patients with troponin-positive chest pain and minimal angiographic coronary artery disease, Am. J. Cardiol. 97 (6) (2006 Mar 15) 768e771. [5] R.G. Assomull, J.C. Lyne, N. Keenan, A. Gulati, N.H. Bunce, S.W. Davies, D.J. Pennell, S.K. Prasad, The role of cardiovascular magnetic resonance in patients presenting with chest pain, raised troponin, and unobstructed coronary arteries, Eur. Heart J. 28 (10) (2007 May) 1242e1249. rez-David, J.A. Larena, S. Velasco del Castillo, [6] E. Laraudogoitia Zaldumbide, E. Pe J.R. Rumoroso Cuevas, J.J. Onaindía, I. Lekuona Goya, M.A. García-Fern andez, The value of cardiac magnetic resonance in patients with acute coronary syndrome and normal coronary arteries, Rev. Esp. Cardiol. 62 (9) (2009 Sep) 976e983. [7] G. Leurent, B. Langella, C. Fougerou, P.A. Lentz, A. Larralde, M. Bedossa, D. Boulmier, H. Le Breton, Diagnostic contributions of cardiac magnetic resonance imaging in patients presenting with elevated troponin, acute chest pain syndrome and unobstructed coronary arteries, Arch. Cardiovasc Dis. 104 (3) (2011 Mar) 161e170. que, [8] E. Gerbaud, E. Harcaut, P. Coste, M. Erickson, M. Lederlin, J.N. Labe
[11]
[12]
[13]
[14]
[15]
91
J.M. Perron, H. Cochet, P. Dos Santos, C. Durrieu-Jaïs, F. Laurent, M. Montaudon, Cardiac magnetic resonance imaging for the diagnosis of patients presenting with chest pain, raised troponin, and unobstructed coronary arteries, Int. J. Cardiovasc Imaging 28 (4) (2012 Apr) 783e794. R. Chopard, J. Jehl, J. Dutheil, V.D. Genon, M.F. Seronde, B. Kastler, F. Schiele, N. Meneveau, Evolution of acute coronary syndrome with normal coronary arteries and normal cardiac magnetic resonance imaging, Arch. Cardiovasc Dis. 104 (10) (2011 Oct) 509e517. P. Lurz, I. Eitel, J. Adam, J. Steiner, M. Grothoff, S. Desch, G. Fuernau, S. de Waha, M. Sareban, C. Luecke, K. Klingel, R. Kandolf, G. Schuler, M. Gutberlet, H. Thiele, Diagnostic performance of CMR imaging compared with EMB in patients with suspected myocarditis, JACC Cardiovasc Imaging 5 (5) (2012 May) 513e524. M.B. Gravanis, N.H. Sternby, Incidence of myocarditis. A 10-year autopsy € , Sweden, Arch. Pathol. Lab. Med. 115 (4) (1991 Apr) study from Malmo 390e392. €m, L. Lind, E.M. Larsson, C.E. Barbier, R. Nylander, R. Themudo, H. Ahlstro T. Bjerner, L. Johansson, Prevalence of unrecognized myocardial infarction detected with magnetic resonance imaging and its relationship to cerebral ischemic lesions in both sexes, J. Am. Coll. Cardiol. 58 (13) (2011 Sep 20) 1372e1377. I. Eitel, F. von Knobelsdorff-Brenkenhoff, P. Bernhardt, I. Carbone, K. Muellerleile, A. Aldrovandi, M. Francone, S. Desch, M. Gutberlet, O. Strohm, G. Schuler, J. Schulz-Menger, H. Thiele, M.G. Friedrich, Clinical characteristics and cardiovascular magnetic resonance findings in stress (takotsubo) cardiomyopathy, JAMA 306 (3) (2011 Jul 20) 277e286. S. Agewall, M. Daniel, L. Eurenius, C. Ekenb€ ack, M. Skeppholm, K. Malmqvist, C. Hofman-Bang, O. Collste, M. Frick, L. Henareh, T. Jernberg, P. Tornvall, Risk factors for myocardial infarction with normal coronary arteries and myocarditis compared with myocardial infarction with coronary artery stenosis, Angiology 63 (7) (2012 Oct) 500e503. K. Thygesen, J.S. Alpert, A.S. Jaffe, M.L. Simoons, B.R. Chaitman, H.D. White, Writing Group on the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction, Thygesen K, Alpert JS, White HD, Jaffe AS, Katus HA, Apple FS, Lindahl B, Morrow DA, Chaitman BA, 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, P, Ravkilde J, Ohman EM, Antman EM, Wallentin LC, Bassand JP, Menasche Armstrong PW, Simoons ML, 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; ESC Committee for Practice Guidelines (CPG). Third universal definition of myocardial infarction, Eur. Heart J. 33 (20) (2012 Oct) 2551e2567.
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Myocarditis or “true” infarction by cardiac magnetic resonance in patients with a clinical diagnosis of myocardial infarction without obstructive coronary disease: A meta-analysis of individual patient data P. Tornvall a, *, E. Gerbaud b, A. Behaghel c, R. Chopard d, O. Collste a, E. Laraudogoitia e, € rensson h, S. Agewall i G. Leurent c, N. Meneveau d, M. Montaudon f, E. Perez-David g, P. So €dersjukhuset, Karolinska Institutet, Sweden Cardiology Unit, Department of Clinical Science and Education So ^pital du Haut L Soins Intensifs Cardiologiques, Plateau de Cardiologie Interventionnelle, CHU de Bordeaux, Ho ev^ eque, 5 Avenue de Magellan, F33604 Pessac, France c CHU de Rennes, Service de Cardiologie et Maladies Vasculaires, INSERM, U1099, Universit e de Rennes 1, LTSI, Rennes, France d Department of Cardiology, EA 3920, University Hospital Jean Minjoz, 25000 Besancon, France e Hospital Galdakao, Vizcaya, Spain f ^pital du Haut L Unit e d'Imagerie Thoracique et Cardiovasculaire, CHU de Bordeaux, Ho ev^ eque, 5 Avenue de Magellan, F33604 Pessac, France g ~ on Madrid, Spain Hospital Gregorio Maran h Cardiology Unit, Department of Medicine, Karolinska University Hospital, Karolinska Institutet, Sweden i Department of Cardiology, Oslo University Hospital Ullevål, Institute of Clinical Medicine, University of Oslo, Oslo, Norway a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 October 2014 Received in revised form 29 April 2015 Accepted 30 April 2015 Available online 1 May 2015
Objective: Myocardial Infarction with Non-Obstructed Coronary Arteries (MINOCA) is common, but the causes are to a large extent unknown. Thus, we aimed to study the prevalence of myocarditis and “true” myocardial infarction determined by cardiac magnetic resonance (CMR) imaging in MINOCA patients, and risk markers for these two conditions in this population. Methods: A search was made in the PubMed and Cochrane databases using the search terms “Myocardial infarction”, “Coronary angiography”, “Normal coronary arteries” and “MRI”. All relevant abstracts were read and seven of the studies fulfilled the inclusion criteria; studies describing case series of patients fulfilling the diagnosis of acute myocardial infarction with normal or non-obstructive coronary arteries on coronary angiography that were investigated with CMR imaging. Data from five of these studies are presented. Results: A total of 556 patients from 5 different sites were included. Fifty-one percent were men with a mean age of 52 ± 16 years. Thirty-three per cent of the patients had myocarditis (n ¼ 183), whereas 21% of the patients had infarction on CMR (n ¼ 115). Young age and a high CRP were associated with myocarditis whereas male sex, treated hyperlipidemia, high troponin ratio and low CRP were associated with “true” myocardial infarction. Conclusion and relevance: The results of this meta-analysis of individual data showed that myocarditis and “true” myocardial infarction are common in MINOCA when determined by CMR imaging. This information emphasizes the importance of performing CMR imaging in MINOCA patients and can be used clinically to guide diagnostics and treatment of MINOCA patients. © 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Myocardial infarction Normal coronary angiography Cardiac magnetic resonance imaging Myocarditis Infarction
1. Introduction
* Corresponding author. Department of Clinical Science and Education €dersjukhuset, Karolinska Institutet, So € dersjukhuset, Sjukhusbacken 10, 118 83 So Stockholm, Sweden. E-mail address: [email protected] (P. Tornvall). http://dx.doi.org/10.1016/j.atherosclerosis.2015.04.816 0021-9150/© 2015 Elsevier Ireland Ltd. All rights reserved.
Myocardial Infarction with normal or Non-Obstructive Coronary Arteries (MINOCA) is common with a prevalence of 1e12% of all patients with a clinical diagnosis of myocardial infarction [1]. Recently, Collste et al. [2] suggested a prevalence of 7e8 %. MINOCA is a heterogeneous condition with patients fulfilling the diagnostic
88
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91
criteria of acute myocardial infarction. There are several different sub-groups of MINOCA, including myocarditis, “true” myocardial infarction and Takotsubo stress cardiomyopathy (TSC). Taking into account the heterogeneity of the condition, the mechanisms behind MINOCA are to a large extent unknown [3]. Cardiac magnetic resonance (CMR) imaging has been shown by several investigators to be of value in the investigation of MINOCA patients [2,4e9]. CMR imaging makes it possible to differentiate between myocarditis, “true” myocardial infarction and TSC although the latter is dependent on a short delay from symptom to investigation and clinical findings. The prevalence of myocarditis has varied from 7 to 63 % [2,6] whereas the prevalence of “true” myocardial infarction has varied from 12 to 30 % [4,5]. The studies have been small and heterogeneous regarding patient characteristics, in particular age and sex have varied between studies. The size and heterogeneity of the studies makes it difficult to find out the true prevalence and risk markers of myocarditis and “true” myocardial infarction among MINOCA patients investigated with CMR imaging. Thus, a large sample size is needed to study CMR imaging results in MINOCA. The presents study, including consecutive individual data from five studies [2,6e9], aimed at establishing the prevalence and risk markers for myocarditis and “true” myocardial infarction among MINOCA patients investigated with CMR imaging. Our hypothesis was that it was possible to find clinical markers to predict the occurrence of myocarditis and “true” myocardial infarction. 2. Material and methods A search was made by two of the authors (SA, PT), who independently of each other performed the search in PubMed using the MeSH search terms “Myocardial infarction”, “Coronary angiography” and “MRI” with the addition of the free search term “Normal coronary arteries”. An additional search was made in the Cochrane Library. All relevant abstracts were read and five of the studies described case series with all patients fulfilling the diagnosis of acute myocardial infarction with normal or non-obstructive coronary arteries that were investigated with CMR imaging [2,6e9]. Two more publications were found by reading references from the relevant publications [4,5]. In total, seven relevant studies with similar methodology were found, and the corresponding authors were invited to collaborate within the present study. The final study was able to combine data from five of those studies in the present analysis. Informed consent was obtained from corresponding and senior authors. Original data was transferred between researchers after agreement by e-mail. All included studies were approved by the local ethics committees. 3. Study designs All studies included consecutive cases of the respective time periods of inclusion from the respective catchment area of the involved centers [2,6e9]. The inclusion criteria were similar although there were differences regarding the angiographic
definitions. Exclusion criteria were similar although one study excluded patients with a high coronary heart disease (CHD) Framingham risk score whereas another study included only patients 35e70 years old (Table 1). 4. Risk markers An Excel sheet was prepared and sent to the interested authors of the included studies. The requested background information included age, sex, history of present smoking and diabetes mellitus, treatment for hypertension and dyslipidemia, and the ratio of maximal troponin (I or T) increase to the upper limit of normal and C-reactive protein (CRP) level in plasma. In cases where the CRP level was given as < 5 mg/l, the value was set to 3. Information about smoking, diabetes mellitus and treated dyslipidemia was missing in four patients. Information about treated hypertension was missing in five patients. Information about troponin ratio was missing in five patients whereas data were lacking in 87 (42 with myocarditis, respectively 16 with infarction) patients regarding CRP. 5. CMR imaging The CMR imaging protocols were similar in all studies and included cine images, T2-weighted oedema sequences and delayed enhancement after gadolinium contrast injection. Left ventricular ejection fraction was determined by the respective software of the CMR equipment (data from 11 patients were missing). The evaluation of CMR images varied between the studies. However, the final CMR imaging diagnosis was based mainly on the myocardial distribution of delayed enhancement of gadolinium in order to differentiate between myocarditis (subepicardial) and infarction (subendocardial). TSC was diagnosed clinically by Zaldumbide el al [6]. and Collste and al [2]. and by a combination of CMR imaging and clinical data in the remaining studies [7e9]. Hence, a CMR imaging diagnosis was only possible for myocarditis and infarction. In all studies, the final CMR imaging diagnosis was made by two experts blinded to clinical data. For information regarding detailed protocols see references [2,6e9] and for en overview of the included studies, see Table 1. 6. Statistics Values are presented as percentage, mean ± standard deviation or median (interquartile range). Uni- and multivariable odds ratios (OR) with 95% confidence intervals (CI) of risk markers for myocarditis and infarction were calculated using logistic regression. For this analysis, the material was divided into clinically relevant categories: troponin ratio 10 or >10 and CRP 10 or >10 mg/l. Uni- and multivariable analyses of the OR of cardiac risk factors for myocarditis and infarction determined by CMR imaging were performed. In model 1 age, sex, smoking, diabetes mellitus, hypertension, dyslipidemia, troponin ratio and CRP were included and model 2 included all variables in model 1 with the addition of
Table 1 Characteristics of studies included in the meta-analysis.
Ref. Ref. Ref. Ref. Ref.
[6] [7] [9] [8] [2]
Number of patients
Age, years
Male sex %
CA findings
Restrictions
Time from AMI to MRI
MRI equipment
80 107 87 130 152
48 44 53 54 58
64 63 40 52 36
Non-obstructive Non-obstructive Normal Non-obstructive Normal
No major No major Low risk score* No major Age 35e70
3 days 5 days 10 days 6 days 12 days
Philips® or Siemens® 1.5 T Philips® 3 T GE® 3 T Siemens® 1.5 T GE®, Philips® or Siemens® 1.5 T
* Framingham risk score > 10% in 10 years. No ¼ Number, CA ¼ Coronary angiography, AMI ¼ Acute Myocardial Infarction, MRI ¼ Magnetic Resonance imaging, GE ¼ General Electric, T ¼ Tesla.
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91 Table 2 Patient characteristics of all patients and patients divided into myocarditis and infarction determined by cardiac magnetic resonance (CMR) imaging.
Age (years) Male gender (%) Present smoking (%) Diabetes mellitus (%) Treated hypertension (%) Treated dyslipidemia (%) Ejection fraction (%) Troponin ratio CRP (mg/l)
All patients n ¼ 556
CMR myocarditis n ¼ 183
CMR infarction n ¼ 115
52 (39e65) 51% 36% 6% 30% 24% 59 (51e65) 40 (12e100) 6 (3e34)
37 (29e49) 78% 42% 1% 12% 16% 58 (51e65) 82 (33e225) 30 (7e61)
54 (44e64) 53% 42% 8% 29% 37% 57 (50e65) 85 (29e154) 5 (3e11)
Values are presented either as percent (%) or median (inter-quartile range). Troponin ratio is defined as the maximal troponin (I or T) divided by the upper limit of the reference value. CRP ¼ C-reactive protein.
site of investigation. The analyses presented were performed without imputation of missing values. As a sensitivity analysis, all analyses were repeated after imputation of missing CRP values with CRP 10 or >10 mg/l. All analyses were performed by a biostatistician using SPSS® version 22 for statistical analysis and p < 0.05 was considered significant. 7. Results Mean age of the patients was 52 ± 16 years. The CHD risk factor profile included approximately one third of active smokers and treated hypertension. The prevalence of diabetes mellitus was low (5%) and approximately one fourth of the patients had treatment for dyslipidemia. Thirty-three percent of the patients had myocarditis whereas 21% of the patients had infarction on CMR imaging (Table 2). Patients with a CMR imaging diagnosis of myocarditis were considerably younger and more often men than patients without myocarditis. With the exception for active smoking, that was more common in patients with myocarditis, the CHD risk factor profile was less pronounced when compared with patients without myocarditis. CRP was higher in patients with myocarditis than without (Tables 2 and 3). Patients with a CMR imaging diagnosis of infarction were of similar age and gender when compared with patients without infarction. The CHD risk factor profile was similar in patients with or without infarction with the exception for treated dyslipidemia that was more common in patients with infarction. Patients with infarction had a higher troponin ratio and a lower CRP than patients without infarction (Tables 2 and 4). 8. Multivariable analysis Age and a high CRP were associated with myocarditis when all
89
variables were included in the analysis simultaneously. There were no major changes in the strength of the associations when site of investigation was included in the multivariate analysis (Table 3). Furthermore, there were no major changes in the associations when CRP 10 or >10 mg/l were imputated for missing CRP values (data not shown). Male sex, treated hyperlipidemia, high troponin ratio (>10) and low CRP (10 mg/l) were all associated with infarction when all variables were included in the analysis simultaneously. High troponin ratio and low CRP showed the strongest associations with infarction with no major changes in the strength of the associations when site of investigation was included in the multivariate analysis (Table 4). Furthermore, there were no major changes in the associations when CRP 10 or >10 mg/l were imputated for missing CRP values (data not shown).
9. Discussion As far as we know this is the largest study on CMR imaging in MINOCA. The results showed that one third of the patients had myocarditis whereas one fifth of the patients had “true” myocardial infarction. The hypothesis was that it was possible to find clinical markers for the occurrence of myocarditis and “true” myocardial infarction. Several clinical markers that were different for the two conditions emerged. Young age and high CRP were associated with myocarditis whereas male sex, treated hyperlipidemia, high troponin ratio and low CRP were associated with “true” myocardial infarction. The prevalence of acute myocarditis was 33% in this selected group of patients with MINOCA. There was a large difference between the five included studies regarding the prevalence that varied from 7 % to 63 % depending on the site of investigation [2,6e9]. To the best of our knowledge there are no CMR imaging studies of unselected healthy subjects showing the prevalence of myocarditis. However, the diagnostic accuracy of CMR in relation to myocardial biopsy in suspected myocarditis has recently been shown to be good (79%) according to Lurz el al [10]. One autopsy study of unselected patients showed that the prevalence of myocarditis was below one percent [11]. Young age was a strong marker for myocarditis which is supported by an autopsy study by Gravani and Sternby [11] who reported that myocarditis patients were young. The other marker of myocarditis was an increased CRP, a finding that was expected but not convincingly shown before in this particular group of patients. The prevalence of “true” myocardial infarction was 21% with minor differences between studies (12e30%). This figure is comparably small considering that the patients fulfilled criteria for myocardial infarction and the notion that one population-based study of 75-year old subjects showed that 30% of 394 men and woman had unrecognized myocardial infarction determined by
Table 3 Uni- and multivariable analyses of the predictive values of cardiac risk factors for myocarditis determined by cardiac magnetic resonance (CMR) imaging.
Age, years (n ¼ 556) Male sex (n ¼ 286) Present smoking (n ¼ 197) Diabetes mellitus (n ¼ 31) Treated hypertension (n ¼ 163) Treated dyslipidemia (n ¼ 135) Troponin ratio > 10 (n ¼ 408) CRP > 10 mg/l (n ¼ 205)
Univariable
Multivariable model 1
Multivariable model 2
0.92 5.51 1.50 0.13 0.22 0.45 1.31 6.23
0.93 1.40 0.66 0.35 0.90 0.67 0.85 4.29
0.94 1.30 0.54 0.43 0.86 0.66 0.80 2.98
(0.90e0.93) (3.67e8.26) (1.04e2.16) (0.03e0.56) (0.14e0.37) (0.28e0.72) (0.87e1.99) (4.00e9.70)
(0.92e0.95) (0.81e2.45) (0.39e1.12) (0.07e1.81) (0.45e1.80) (0.36e1.27) (0.47e1.52) (2.56e7.18)
(0.92e0.96) (0.71e2.36) (0.30e0.98) (0.07e2.80) (0.42e1.75) (0.34e1.29) (0.42e1.51) (1.70e5.22)
Values are presented as odds ratio (confidence interval). Troponin ratio is defined as the maximal troponin (I or T) divided by the upper limit of the reference value. CRP ¼ Creactive protein. Model 1 included age, sex, smoking, diabetes mellitus, hypertension, dyslipidemia, troponin ratio and CRP. Model 2 included all variables in model 1 with the addition of site of investigation.
90
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91
Table 4 Uni- and multivariable analyses of the predictive values of cardiac risk factors for infarction determined by cardiac magnetic resonance (CMR) imaging.
Age, years (n ¼ 556) Male sex (n ¼ 286) Present smoking (n ¼ 197) Diabetes mellitus (n ¼ 31) Treated hypertension (n ¼ 163) Treated dyslipidemia (n ¼ 135) Troponin ratio > 10 (n ¼ 408) CRP > 10 mg/l (n ¼ 205)
Univariable
Multivariable model 1
Multivariable model 2
1.01 0.92 1.37 1.64 0.98 2.20 3.20 0.40
1.02 1.92 1.64 0.74 0.70 3.06 5.26 0.29
1.02 1.87 1.40 0.66 0.66 2.72 5.93 0.25
(0.99e1.02) (0.61e1.39) (0.90e2.09) (0.73e3.67) (0.62e1.54) (1.41e3.43) (1.73e5.92) (0.24e0.65)
(0.99e1.04) (1.08e3.38) (0.96e2.81) (0.24e2.28) (0.37e1.30) (1.76e5.33) (2.46e11.25) (0.17e0.51)
(0.99e1.04) (1.05e3.35) (0.80e2.47) (0.21e2.05) (0.35e1.26) (1.52e4.85) (2.69e13.04) (0.14e0.45)
Values are presented as odds ratio (confidence interval). Troponin ratio is defined as the maximal troponin (I or T) divided by the upper limit of the reference value. CRP ¼ Creactive protein. Model 1 included age, sex, smoking, diabetes mellitus, hypertension, dyslipidemia, troponin ratio and CRP. Model 2 included all variables in model 1 with the addition of site of investigation.
CMR imaging [12]. The reason for the low prevalence of “true” myocardial infarction is probably due to the high prevalence of myocarditis and TSC. The prevalence of reported TSC varied from 11 % to 22 % between the studies. TSC is associated with wall motion disturbances and oedema in the acute stage and a recent large CMR imaging study reported that only nine percent of the TSC patients had signs of permanent delayed gadolinium enhancement as a sign of infarction [13]. Unfortunately, it was not possible to study risk markers for TSC determined by CMR imaging since TSC was diagnosed clinically in two of the studies [2,6] and by a combination of CMR imaging and clinical data in three of the studies [7e9]. Our group of patients with “true” myocardial infarction was predominantly male patients with treatment for dyslipidemia with high troponin ratio and low CRP. The most likely reason for the strength of the marker CRP is that it can differentiate myocardial infarction from myocarditis [14]. Possibly, the mechanism behind MINOCA with “true” myocardial infarction is due to thromboembolism despite normal or near normal coronary arteries [3]. 10. Clinical implications The results of the present study emphasize the importance of CMR imaging in the investigation of MINOCA. It should in this context be acknowledged that only approximately 50% of the patients received a definitive diagnosis by CMR and other clinical information is of great value. Myocarditis is common and important to diagnose since the treatment differ from other causes of MINOCA. The new knowledge about risk markers for myocarditis is clinically relevant. For example, in a clinically stable young patient with high CRP fulfilling diagnostic criteria for acute myocardial infarction an alternative first major investigation could be CMR imaging. In the case of a myocarditis diagnosis on CMR imaging, coronary angiography could be omitted from further investigations decreasing the patient total burden of iatrogenic risk. These patients may also avoid to be treated with the standard acute coronary syndromes (ACS) pharmacological treatment. Regarding the risk markers for “true” myocardial infarction, the new knowledge that patients with MINOCA with high troponin ratio and low CRP are likely to show signs of infarction on CMR imaging emphasize the importance to investigate these patients with CMR imaging. Patients with a CMR imaging diagnosis of infarction should probably have similar pharmacological treatment as patients with ACS. Unfortunately, there is a knowledge gap how to treat these patients and thus an urgent need for clinical guidelines in this area. 11. Limitations The present study has several limitations. First, although the five studies are similar regarding inclusion criteria the study group
characteristics differ between sites, in particular regarding age and sex. The reason for this in not clear but is likely due to the threshold to perform coronary angiography in younger patients where myocarditis is a likely diagnosis. Furthermore, the time interval between presentation to hospital and CMR differed slightly between the studies, which might have affected the results. This interval will mainly affect the results of oedema sequences. However, the results were based on delayed enhancement of gadolinium that is less sensitive to the time interval. Furthermore, site of investigation was added as a variable into the multivariable analysis thus indirectly correcting the results for time interval. The results did not change when site of investigation was included. Second, there are differences regarding methods to analyze troponin and CRP. We tried to circumvent the difficulties with different troponin methods by analyzing the ratio of peak to upper level of normal. Despite this correction there are likely differences between the sites. Third, there are major differences between the sites regarding CMR imaging technology and evaluation of the images that might have influenced the results. Fourth, it is not possible to clearly make the distinction towards secondary type 2 myocardial infarction, which is a myocardial injury with necrosis, where a condition other than CAD contributes to an imbalance between myocardial oxygen supply and/or demand [15]. Furthermore, ECG and echocardiography are commonly performed in the clinical practice before performing cardiac MRI. Unfortunately, we have no echocardiography data and the quality of the ECG data is uncertain. However, some studies included in this meta-analysis have already shown the additive value of cardiac MRI to ECG and echocardiography. For example, in the study by Laraudogoitia Zaldumbide et al. [6], CMR provided a formal diagnosis in 61 patients (69.3%) in which the clinical diagnosis was uncertain between at least two possibilities. Anyway, the lack of echocardiography and ECG data is another limitation of the study. Although the analysis may be limited by the inability to include all studies of interest, we believe that differences regarding age, sex and prevalence of myocarditis between our results and the studies not included [4,5], are unlikely to have significantly affected our results. 12. Conclusion The results of this meta-analysis of individual data showed that myocarditis and “true” myocardial infarction are common in MINOCA when determined by CMR imaging. Several risk markers for the two conditions emerged with low age and high CRP being most important for myocarditis and high troponin and low CRP being highly relevant for infarction. This information emphasizes the importance of performing CMR imaging in MINOCA patients and can be used clinically to guide diagnostics and treatment of MINOCA patients.
P. Tornvall et al. / Atherosclerosis 241 (2015) 87e91
Conflict of interest None. [9]
Acknowledgment None of the authors had any relevant conflict of interest.
[10]
References [1] I. Kardasz, R. De Caterina, Myocardial infarction with normal coronary arteries:a conundrum with multiple aetiologies and variable prognosis: an update, J. Intern. Med. 261 (4) (2007 Apr) 330e348. €rensson, M. Frick, S. Agewall, M. Daniel, L. Henareh, [2] O. Collste, P. So C. Ekenb€ ack, L. Eurenius, C. Guiron, T. Jernberg, C. Hofman-Bang, K. Malmqvist, E. Nagy, H. Arheden, P. Tornvall, Myocardial infarction with normal coronary arteries is common and associated with normal findings on cardiovascular magnetic resonance imaging: results from the Stockholm myocardial infarction with normal coronaries study, J. Intern Med. 273 (2) (2013 Feb) 189e196. [3] S. Agewall, L. Eurenius, C. Hofman-Bang, K. Malmqvist, M. Frick, T. Jernberg, P. Tornvall, Myocardial infarction with angiographically normal coronary arteries, Atherosclerosis 219 (1) (2011 Nov) 10e14. [4] J.P. Christiansen, C. Edwards, T. Sinclair, G. Armstrong, A. Scott, H. Patel, H. Hart, Detection of myocardial scar by contrast-enhanced cardiac magnetic resonance imaging in patients with troponin-positive chest pain and minimal angiographic coronary artery disease, Am. J. Cardiol. 97 (6) (2006 Mar 15) 768e771. [5] R.G. Assomull, J.C. Lyne, N. Keenan, A. Gulati, N.H. Bunce, S.W. Davies, D.J. Pennell, S.K. Prasad, The role of cardiovascular magnetic resonance in patients presenting with chest pain, raised troponin, and unobstructed coronary arteries, Eur. Heart J. 28 (10) (2007 May) 1242e1249. rez-David, J.A. Larena, S. Velasco del Castillo, [6] E. Laraudogoitia Zaldumbide, E. Pe J.R. Rumoroso Cuevas, J.J. Onaindía, I. Lekuona Goya, M.A. García-Fern andez, The value of cardiac magnetic resonance in patients with acute coronary syndrome and normal coronary arteries, Rev. Esp. Cardiol. 62 (9) (2009 Sep) 976e983. [7] G. Leurent, B. Langella, C. Fougerou, P.A. Lentz, A. Larralde, M. Bedossa, D. Boulmier, H. Le Breton, Diagnostic contributions of cardiac magnetic resonance imaging in patients presenting with elevated troponin, acute chest pain syndrome and unobstructed coronary arteries, Arch. Cardiovasc Dis. 104 (3) (2011 Mar) 161e170. que, [8] E. Gerbaud, E. Harcaut, P. Coste, M. Erickson, M. Lederlin, J.N. Labe
[11]
[12]
[13]
[14]
[15]
91
J.M. Perron, H. Cochet, P. Dos Santos, C. Durrieu-Jaïs, F. Laurent, M. Montaudon, Cardiac magnetic resonance imaging for the diagnosis of patients presenting with chest pain, raised troponin, and unobstructed coronary arteries, Int. J. Cardiovasc Imaging 28 (4) (2012 Apr) 783e794. R. Chopard, J. Jehl, J. Dutheil, V.D. Genon, M.F. Seronde, B. Kastler, F. Schiele, N. Meneveau, Evolution of acute coronary syndrome with normal coronary arteries and normal cardiac magnetic resonance imaging, Arch. Cardiovasc Dis. 104 (10) (2011 Oct) 509e517. P. Lurz, I. Eitel, J. Adam, J. Steiner, M. Grothoff, S. Desch, G. Fuernau, S. de Waha, M. Sareban, C. Luecke, K. Klingel, R. Kandolf, G. Schuler, M. Gutberlet, H. Thiele, Diagnostic performance of CMR imaging compared with EMB in patients with suspected myocarditis, JACC Cardiovasc Imaging 5 (5) (2012 May) 513e524. M.B. Gravanis, N.H. Sternby, Incidence of myocarditis. A 10-year autopsy € , Sweden, Arch. Pathol. Lab. Med. 115 (4) (1991 Apr) study from Malmo 390e392. €m, L. Lind, E.M. Larsson, C.E. Barbier, R. Nylander, R. Themudo, H. Ahlstro T. Bjerner, L. Johansson, Prevalence of unrecognized myocardial infarction detected with magnetic resonance imaging and its relationship to cerebral ischemic lesions in both sexes, J. Am. Coll. Cardiol. 58 (13) (2011 Sep 20) 1372e1377. I. Eitel, F. von Knobelsdorff-Brenkenhoff, P. Bernhardt, I. Carbone, K. Muellerleile, A. Aldrovandi, M. Francone, S. Desch, M. Gutberlet, O. Strohm, G. Schuler, J. Schulz-Menger, H. Thiele, M.G. Friedrich, Clinical characteristics and cardiovascular magnetic resonance findings in stress (takotsubo) cardiomyopathy, JAMA 306 (3) (2011 Jul 20) 277e286. S. Agewall, M. Daniel, L. Eurenius, C. Ekenb€ ack, M. Skeppholm, K. Malmqvist, C. Hofman-Bang, O. Collste, M. Frick, L. Henareh, T. Jernberg, P. Tornvall, Risk factors for myocardial infarction with normal coronary arteries and myocarditis compared with myocardial infarction with coronary artery stenosis, Angiology 63 (7) (2012 Oct) 500e503. K. Thygesen, J.S. Alpert, A.S. Jaffe, M.L. Simoons, B.R. Chaitman, H.D. White, Writing Group on the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction, Thygesen K, Alpert JS, White HD, Jaffe AS, Katus HA, Apple FS, Lindahl B, Morrow DA, Chaitman BA, 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, P, Ravkilde J, Ohman EM, Antman EM, Wallentin LC, Bassand JP, Menasche Armstrong PW, Simoons ML, 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; ESC Committee for Practice Guidelines (CPG). Third universal definition of myocardial infarction, Eur. Heart J. 33 (20) (2012 Oct) 2551e2567.
