Prognostic Role of CMR and Conventional Risk Factors in Myocardial Infarction With Nonobstructed Coronary Arteries

JACC: CARDIOVASCULAR IMAGING

VOL.

-, NO. -, 2019

ª 2019 PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

ORIGINAL RESEARCH

Prognostic Role of Cardiac MRI and Conventional Risk Factors in Myocardial Infarction With Nonobstructed Coronary Arteries Amardeep Ghosh Dastidar, MBBS,a,b Anna Baritussio, MBBS, MD, PHD,a,b Estefania De Garate, MBBS,a,b,c Zsofia Drobni, MBBS,a Giovanni Biglino, PHD,a,b Priyanka Singhal, MBCHB,a,b Elena G. Milano, MBBS, MD,a,b Gianni D. Angelini, MD, MCH,a,b,c Stephen Dorman, MBBS,a Julian Strange, MBBS, MD,a Thomas Johnson, BSC (HONS), MBBS, MD,a,b Chiara Bucciarelli-Ducci, MBBS, MD, PHDa,b,c

ABSTRACT OBJECTIVES This study sought to assess the prognostic impact of cardiovascular magnetic resonance (CMR) and conventional risk factors in patients with myocardial infarction with nonobstructed coronaries (MINOCA). BACKGROUND Myocardial infarction with nonobstructed coronary arteries (MINOCA) represents a diagnostic dilemma, and the prognostic markers have not been clarified. METHODS A total of 388 consecutive patients with MINOCA undergoing cardiac magnetic resonance (CMR) assessment were identified retrospectively from a registry database and prospectively followed for a primary clinical endpoint of allcause mortality. A 1.5-T CMR was performed using a comprehensive protocol (cines, T2-weighted, and late gadolinium enhancement sequences). Patients were grouped into 4 categories based on their CMR findings: myocardial infarction (MI) (embolic/spontaneous recanalization), myocarditis, cardiomyopathy, and normal CMR. RESULTS CMR (performed at a median of 37 days from presentation) was able to identify the cause for the troponin rise in 74% of the patients (25% myocarditis, 25% MI, and 25% cardiomyopathy), whereas a normal CMR was identified in 26%. Over a median follow-up of 1,262 days (3.5 years), 5.7% patients died. The cardiomyopathy group had the worst prognosis (mortality 15%; log-rank test: 19.9; p < 0.001), MI had 4% mortality, and 2% in both myocarditis and normal CMR. In a multivariable Cox regression model (including clinical and CMR parameters), CMR diagnosis of cardiomyopathy and ST-segment elevation on presentation electrocardiogram (ECG) remained the only 2 significant predictors of mortality. Using presentation with ECG ST-segment elevation and CMR diagnosis of cardiomyopathy as risk markers, the mortality risk rates were 2%, 11%, and 21% for presence of 0, 1, and 2 factors, respectively (p < 0.0001). CONCLUSIONS In a large cohort of patients with MINOCA, CMR (median 37 days from presentation) identified a final diagnosis in 74% of patients. Cardiomyopathy had the highest mortality, followed by MI. The strongest predictors of mortality were a CMR diagnosis of cardiomyopathy and ST-segment elevation on presentation ECG. (J Am Coll Cardiol Img 2019;-:-–-) © 2019 Published by Elsevier on behalf of the American College of Cardiology Foundation.

From the aBristol Heart Institute, University Hospitals Bristol NHS Foundation Trust, Bristol, United Kingdom; bSchool of Clinical Sciences, Faculty of Health Sciences, University of Bristol, Bristol, United Kingdom; and the cBristol National Institute of Health Research, Biomedical Research Centre, Bristol, United Kingdom. Dr. Bucciarelli-Ducci is supported by the National Institute of Health Research Biomedical Research Centre at University Hospitals Bristol NHS Foundation Trust and the University of Bristol and is a consultant for Circle Cardiovascular Imaging. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the National Institute for Health Research, or the Department of Health and Social Care. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received June 6, 2018; revised manuscript received December 3, 2018, accepted December 14, 2018.

ISSN 1936-878X/$36.00

https://doi.org/10.1016/j.jcmg.2018.12.023

Dastidar et al.

2

JACC: CARDIOVASCULAR IMAGING, VOL.

A

ABBREVIATIONS AND ACRONYMS ACS = acute coronary syndrome

-, NO. -, 2019 - 2019:-–-

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

cute coronary syndrome (ACS) is one

of Cardiology (ESC) guidelines on the management of

of the leading causes of mortality

STEMI and the recent ESC/American College of Car-

and morbidity worldwide (1). Defi-

diology/American

Heart

Association/World

Heart

nite diagnosis of ACS is crucial for ensuring

Federation fourth universal definition of myocardial

appropriate patient management. As per the

infarction has included MINOCA, suggesting a po-

resonance

international guidelines, patients presenting

tential role for cardiovascular magnetic resonance

ECG = electrocardiogram

with

(CMR) in this cohort (3,9).

CMR = cardiac magnetic

ST-segment

elevation

myocardial

infarction (STEMI) or non-ST-segment eleva-

Confirmation or exclusion of myocardial infarction

volume

tion-ACS (NSTE-ACS) proceed to immediate

(MI) by CMR facilitates tailoring of medical therapy, ensuring appropriate long-term secondary preven-

iEDV = indexed end-diastolic

LGE = late gadolinium

or inpatient coronary angiography, respec-

enhancement

tively, with the aim to identify and treat a

tion and modification of risk, and minimizing expo-

LV = left ventricular

culprit coronary artery occlusion or stenosis

sure to antiplatelet therapy and the associated

LVEF = left ventricular ejection

(2,3). In 1% to 14% of these patients, howev-

bleeding risks for those with a noncoronary etiology

fraction

er, no significant coronary obstruction is

for the MINOCA presentation (10). Few previous

MI = myocardial infarction

identified, and thus these cases are classified

studies have shown that CMR can identify the un-

MINOCA = myocardial

as myocardial infarction with nonobstructive

derlying diagnosis, which is made most commonly

infarction with nonobstructed

coronary arteries (MINOCA) (3). MINOCA rep-

based on acute/chronic myocarditis, acute MI with

resents a diagnostic dilemma, with subse-

spontaneous recanalization/embolus, stress cardio-

elevation-acute coronary

quent uncertain clinical management. A

myopathy (Takotsubo), or other cardiomyopathies

syndrome

recent registry study has shown beneficial ef-

(11–17). Only in a small proportion of cases is no ab-

RV = right ventricular

fects

and

normality identified by CMR. The variable proportion

RWMA = regional wall motion

angiotensin-converting enzyme inhibitors/

of normal cases in the different studies may be due to

abnormality

angiotensin receptor blockers on outcome in

the time delay between the acute presentation and

STEMI = ST-segment elevation

patients with MINOCA (4).

CMR (18), suggesting that the time between acute

coronary artery

NSTE-ACS = non-ST-segment

myocardial infarction

of

treatment

with

statins

The underlying pathophysiological mech-

presentation and CMR can play a role in identifying

anisms occurring in MINOCA are complex

the early reversible abnormalities that may resolve

and multifactorial (5). Patients with MINOCA

when patients are scanned later. Recent studies have

are thought to have a better prognosis and conse-

demonstrated the prognosis in the individual diag-

quently do not receive appropriate secondary pre-

nostic categories such as myocarditis and Takotsubo

vention medications (6); however, recent studies

cardiomyopathy (19–21); however, there is no evi-

suggest that all-cause mortality may be as high as

dence in the literature on the prognostic significance

4.7% at 12 months (7,8). The latest European Society

of the CMR findings and conventional risk factors in

STIR = short tau inversion recovery

MINOCA. The only study that looked at midterm mortality was by Mittal et al. (22); however, they

F I G U R E 1 Study Flow Chart

looked at patients presenting via the primary percu-

Emergency/urgent coronary angiography n = 11,757

taneous coronary angiography pathway with nonobstructed coronaries. Sixty percent of the study population had normal troponin and thereby did not

MINOCA referred for CMR n = 410

strictly meet MINOCA criteria. Our study aimed to assess the prognostic impact of CMR and conventional risk factors in patients with MINOCA.

Excluded • Incomplete CMR protocol n = 8 • >50% stenosis n = 14

METHODS STUDY POPULATION. In this longitudinal observa-

Study sample (n = 388)

tional study, consecutive patients presenting with MINOCA (chest pain, elevated troponin, and nonobstructed coronary arteries) per the new ESC STEMI

ST segment elevation n = 74

Non ST segment elevation n = 314

Guideline (3) and undergoing CMR were identified retrospectively and followed prospectively (Figure 1). We excluded patients admitted with suspected heart

CMR ¼ cardiovascular magnetic resonance; MINOCA ¼ myocardial infarction with non-

failure or arrhythmic events at presentation. The

obstructed coronaries.

study was performed at a large cardiothoracic tertiary center in southwest England; data were collected on

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019

Dastidar et al.

- 2019:-–-

3

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

consecutive patients scanned from September 2011 to December 2015. Patients were identified either pre-

T A B L E 1 Demographic Characteristics

Total (n ¼ 388)

MI (n ¼ 97)

Myocarditis (n ¼ 96)

CM (n ¼ 96)

Normal (n ¼ 99)

Global p Value

Age, yrs

56  17

62  12

42  17

64  12

54  16

<0.001

Women

48

56

23

61

54

<0.001

Family history of IHD

3

2

4

1

3

0.57

typical rise or fall of troponin T was used for the

Diabetes

6

8

5

6

4

0.64

diagnosis of acute MI. (9) False-positive troponin

Hypertension

13

16

9

14

14

0.54

possibilities included a single troponin elevation

History of smoking

7

4

5

10

7

0.32

(<5 upper limit of normal) that was not repeated

BMI

26.5

27.3

26.5

25.7

26.4

0.18

during admission or if a single elevated troponin was

Median troponin T, ng/l

497

660

924

419

202

0.03

senting with STEMI or non-STEMI. Nonobstructed coronaries were defined as thrombosis in MI III flow and <50% stenosis in any coronary artery (3). Troponin T level <14 ng/l was considered normal. The

Presentation as STEMI

19

19

19

25

14

0.29

h was excluded from study (2). The study was

Median interval, days (presentation and CMR)

37

37

21

12

47

<0.001

reviewed and approved by the local institutional re-

Median follow-up, days

1,262

1,246

1,336

1,231

1,276

0.09

view board.

All-cause mortality

6 (22)

4 (4)

2 (2)

15 (14)

2 (2)

<0.001

CMR PROTOCOL. CMR was performed at 1.5-T (Mag-

Values are mean  SD, n, or n (%).

netom Avanto, Siemens Healthineers, Erlangen, Ger-

BMI ¼ body mass index; CM ¼ cardiomyopathy; CMR ¼ cardiac magnetic resonance; IHD ¼ ischemic heart disease; MI ¼ myocardial infarction; STEMI ¼ ST-segment elevation myocardial infarction.

followed by a second normal troponin level within 24

many). A comprehensive CMR protocol was carried out, including cine, T2-weighted (myocardial edema), and early and late gadolinium enhancement (LGE)

abnormality (RWMA) (except for dyssynchrony sec-

imaging (23). In particular, cine images were per-

ondary to bundle branch block), no myocardial

formed using a steady-state free-precession sequence

edema, and no myocardial LGE (scarring) (except

acquired in 3 long-axis planes and a stack of short-

nonspecific LGE in left ventricular [LV]/right ven-

axis images covering the left ventricle, and T2-

tricular [RV] insertion points). Myocarditis was diag-

weighted short T1 inversion recovery (T2-STIR)

nosed based on fulfilling 2 of 3 Lake Louise Criteria:

sequence images were acquired in the same short-

T2-STIR sequences detecting myocardial edema;

and long-axis planes as the cine images with standard

early gadolinium sequences detecting hyperemia; or

parameters, as previously described (24,25). Intrave-

epicardial or mid-myocardial LGE, as previously

nous gadolinium-chelate contrast agent (gadobutrol)

described (26). MI was diagnosed based on territorial

was administered at a dose of 0.1 mmol/kg

–1

of body

weight. Images were acquired 2 to 3 min after contrast injection (early gadolinium), whereas LGE images

F I G U R E 2 Age as a Marker of Mortality

were acquired 15 to 20 min after contrast injection Log rank 23.2, p < 0.001

using a standard inversion recovery segmented gradient echo sequence, as previously described

0.3

(24,25).

reported by a consultant with >15 years of CMR experience and with ESC CMR level 3 certification (C.B.D.). Diffuse and focal myocardial edema was analyzed using both the early gadolinium and the T2STIR images. Myocardial edema was considered present when the ratio of signal intensity between the

Cumulative Mortality

CMR ANALYSIS. All CMR studies were analyzed and

0.2

0.1

myocardium and the mean signal intensity of the skeletal muscle was >2 on T2-STIR images, according to the Lake Louise criteria (26). Early gadolinium enhancement ratio used for assessment of hyperemia was evaluated as previously described (26). Patients

0.0 0

500

were grouped into 4 categories based on their CMR characteristics: MI (embolic/spontaneous recanaliza-

1000

1500

2000

<40 yrs

40-59 yrs

60-79 yrs

tion), myocarditis, cardiomyopathy, and normal CMR. The latter corresponded to a structurally normal heart, defined as no regional wall motion

2500

Follow Up in Days

Kaplan-Meier curves showing the risk of mortality according to age group.

>79 yrs

Dastidar et al.

4

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019 - 2019:-–-

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

edema was quantified by calculating the number of

F I G U R E 3 ECG at Presentation

segments involved (20). STUDY PRIMARY ENDPOINT. The endpoint of the

Log rank 7.4, p = 0.007

present study was all-cause mortality. Patient-related

0.15 Cumulative Mortality

data were collected directly from medical records. Follow-up was performed centrally by analyzing the mortality data obtained from the National Health

0.10

Service summary care records. Data were collected and interpreted by 1 of the primary investigators (A.G.D., A.B., E.D.G., or G.B.) in all cases.

0.05

STATISTICAL ANALYSIS. Baseline patient character-

istics, conventional risk markers, and CMR findings 0.00

are described according to the diagnostic groups. Data 0

500

1000

1500

2000

2500

standard deviation or median as appropriate. Cate-

Follow Up in Days

gorical variables are presented as frequencies and

STEMI

NSTE-ACS

for continuous variables are presented as mean 

percentages. Normally distributed continuous variKaplan-Meier curves showing the risk of mortality according to

ables were compared using 1-way analysis of variance

presence or absence of ST-segment elevation on the

with Bonferroni correction for multiple comparisons.

presenting ECG. ECG ¼ electrocardiogram; NSTE-ACS ¼

Continuous variables that were not normally distrib-

non-ST-segment elevation-acute coronary syndrome;

uted were compared by Kruskal-Wallis tests. In uni-

STEMI ¼ ST-segment elevation myocardial infarction.

variate analyses, the association of time variables to mortality was assessed using Kaplan-Meier curves and the log-rank test. Univariable and multivariable

subendocardial or transmural LGE. Takotsubo cardiomyopathy was diagnosed based on the T2-STIR images detecting myocardial edema and regional wall motion in the mid-cavity or apical distribution with no myocardial LGE, all in accordance with the modified Mayo Clinic criteria (27). Dilated cardiomyopathy, hypertrophic cardiomyopathy, and cardiac amyloidosis were detected based on specific tissue characterization characteristics, and all were grouped under cardiomyopathy together with Takotsubo cardiomyopathy (15). The extent of LGE and myocardial

associations of risk covariates with mortality were assessed using Cox proportional hazard regression analyses. Only variables with p < 0.05 in univariable analyses were used in multivariable model. SPSS version 23 (IBM Corp., Armonk, New York), was used for statistical analysis. Probability values were 2sided, and values of p < 0.05 were considered significant. Because of a paucity of mortality data on the MINOCA subgroups, post hoc power calculation was performed based on the study findings.

RESULTS A total of 11,757 patients underwent emergency or

T A B L E 2 CMR Characteristics of the Different Diagnostic Categories

Total (n ¼ 388)

MI (n ¼ 97)

Myocarditis (n ¼ 96)

CM (n ¼ 96)

Normal (n ¼ 99)

Global p Value

urgent coronary angiography during the recruitment period, of which 410 patients with suspected MINOCA

2

81

78

83

90

74

<0.001

underwent CMR assessment. Two percent (n ¼ 8)

iESV, ml/m2

34

30

32

47

26

<0.001

were excluded because of incomplete protocol and

iSV, ml/m2

47

48

50

43

48

<0.001

3% (n ¼ 14) because of coronary artery disease that

LVEF

61

63

62

52

67

<0.001

iLV mass, g/m2

66

63

67

75

59

<0.001

was deemed >50% stenosis, thereby leaving a study

RWMA

39

59

31

59

6

<0.001

LGE

58

100

94

34

6

<0.001

1 (0–2)

1 (1–2)

3 (2–4)

0 (0–1)

0

<0.01

acteristics of the overall population (n ¼ 388) and for

iEDV, ml/m

No. of LGE segments Myocardial edema, % No. of edema segments

sample size of 388. CMR was able to identify a cause for the troponin rise in 74% of cases. Clinical char-

34

52

52

34

0

<0.001

separate patient groups based on the CMR diagnosis

0 (0–2)

1 (0–2)

2 (0–4)

0 (0–5)

0

<0.01

are summarized in Table 1. There was a low prevalence of cardiovascular risk factors (hypertension,

Values are % or median (interquartile range). iEDV ¼ indexed end-diastolic volume; iESV ¼ indexed end-systolic volume; iLV ¼ indexed left ventricular; IQR ¼ interquartile range; iSV ¼ indexed stroke volume; LGE ¼ late gadolinium enhancement; LVEF ¼ left ventricular ejection fraction; MI ¼ myocardial infarction; RWMA ¼ regional wall motion abnormality; other abbreviation as in Table 1.

diabetes, hyperlipidemia, smoking, and family history), with no difference across groups. The median time delay between acute presentation and CMR was 37 days. Thirty-seven of our patients underwent a

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019

Dastidar et al.

- 2019:-–-

5

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

scan within 2 weeks of presentation. In a median follow-up of 1,262 days (3.5 years), the overall all-

T A B L E 3 Characteristics of the CM Subgroups

cause mortality was 5.7%. CONVENTIONAL RISK FACTORS.

TCM (n ¼ 41)

DCM (n ¼ 28)

HCM (n ¼ 17)

Others (n ¼ 10)

Global p Value

Age. The mean

Age, yrs

68  10

63  13

60  17

60  12

0.05

age of the total population was 56  17 years. Kaplan-

Women

98

32

47

33

<0.001

Meier curves were drawn showing the risk of mor-

STEMI

38

18

24

0

0.07

tality according to the different age categories (<40,

iEDV, ml/m2

77

128

67

81

<0.001

40 to 59, 60 to 79, and >79 years). There was a strong

iESV, ml/m2

35

87

20

36

<0.001

iSV, ml/m2

42

41

48

44

0.298

LVEF

56

33

70

59

<0.001

iLV mass, g/m2

61

85

87

88

<0.001

Mortality

6

5

0

3

<0.01

association between increasing age and mortality (log-rank test: 23.2; p < 0.001) (Figure 2). Sex. Women comprised 48% of the total cohort and 98% of the patients diagnosed with Takotsubo cardiomyopathy were women. There was no significant

Values are mean  SD or n.

association between sex and mortality in the overall

DCM ¼ dilated cardiomyopathy; HCM ¼ hypertrophic cardiomyopathy; TMC ¼ Takotsubo cardiomyopathy; other abbreviations as in Tables 1 and 2.

cohort (log-rank test: 2.5; p ¼ 0.10). Electrocardiogram at presentation (STEMI or NSTE-ACS). Nineteen percent of our study population presented

evidence of diffuse myocardial injury in any of the

with ST-segment elevation on the 12-lead electro-

patients.

cardiogram (ECG). Kaplan-Meier curves were drawn

LGE. LGE was present in 58% of the entire cohort. Six

showing the risk of mortality according to the ECG

percent of patients with normal CMR had evidence of

presentation as ST-segment elevation or no ST-

LGE in the LV/RV insertion points only, which

segment elevation. There was a strong association

currently is considered a nonspecific finding.

between ECG presentation as ST-segment elevation and

mortality

(log-rank

test:

7.4;

p

¼ 0.007)

(Figure 3). CMR CHARACTERISTICS AND DIAGNOSIS. The CMR

CARDIOMYOPATHY SUBGROUP ANALYSIS. In the

cardiomyopathy group, 43% had Takotsubo cardiomyopathy, 29% dilated cardiomyopathy, and 18% hypertrophic cardiomyopathy. The remaining had

characteristics are described according to diagnosis

infiltrative cardiomyopathy (amyloidosis; n ¼ 4), hy-

groups (Table 2).

pertensive heart disease (n ¼ 4), and valvular heart

MI was the most prevalent diagnosis (25%, 97 of

disease (n ¼ 2) and were grouped as other. There was

388 patients), followed by myocarditis (25%, 96 of

no statistically significant difference in the mean age

388) and cardiomyopathy 25%, whereas 26% had a

and presentation as STEMI; however, there was a

structurally normal heart and were grouped into the

significant difference in sex among the different car-

normal CMR category.

diomyopathy subgroups. There was also a significant

LV volumes. In the overall cohort, the mean indexed

difference

end-diastolic volume (iEDV) was 81 ml/m 2. On Bon-

different cardiomyopathy subgroups (Table 3).

ferroni correction, the iEDV in the cardiomyopathy was significantly higher than the MI and normal CMR categories. LV ejection fraction. The mean left ventricular ejection fraction (LVEF) for the entire cohort was 61%. On Bonferroni correction, the LVEF in the cardiomyopa-

in

TREATMENT

the

CMR

BEFORE

parameters

CMR. Aspirin,

among other

the anti-

platelets, angiotensin-converting enzyme inhibitor/ angiotensin receptor blocker, beta-blocker, and statin use in the overall cohort was 93%, 61%, 66%, 65%, and 57%, respectively. There was no significant

thy was significantly lower than the MI, myocarditis, and normal CMR categories. In addition, the LVEF in the myocarditis category was significantly lower than

T A B L E 4 Treatment Before CMR

the normal CMR category. RWMA. Overall, 39% of patients had RWMA. The normal CMR category had RWMA secondary to bundle branch block on ECG.

MI (n ¼ 97)

Myocarditis (n ¼ 96)

Aspirin

95

Other antiplatelet

61

ACE inhibitor/ARB

Pre-CMR Treatment

CM (n ¼ 96)

Normal (n ¼ 99)

Global p Value

92

91

92

0.81

59

60

63

0.31

65

66

69

66

0.92

Myocardial edema. The overall prevalence of focal

Beta-blocker

65

64

71

62

0.58

myocardial edema was 34% based on T2 STIR or early

Statin

62

53

59

56

0.66

gadolinium enhancement. Focal myocardial edema was absent in patients with normal CMR. Additional analysis of the normal CMR cohort revealed no

Values are n. ACE ¼ angiotensin-converting enzyme; ARB ¼ angiotensin receptor blocker; other abbreviations as in Table 1.

Dastidar et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019 - 2019:-–-

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

(15% mortality). The mortality rate in the different

F I G U R E 4 Cumulative Mortality According to CMR diagnosis

groups was: cardiomyopathy 15%; MI group 4%; myocarditis group 2%; and normal 2% (p ¼ 0.001).

A

Log rank 19.9, p < 0.001

Kaplan-Meier curves were drawn showing the risk

0.20

of

mortality

according

to

the

different

CMR

diagnoses, showing a strong association between CMR categories and mortality (log-rank test: 19.9; Cumulative Mortality

0.15

p < 0.001) (Figure 4A). An additional Kaplan-Meier curve was drawn for Takotsubo cardiomyopathy, showing a significantly higher mortality in Takotsubo

0.10

group compared with any other diagnosis (log-rank test: 7.3; p ¼ 0.011) (Figure 4B). PREDICTORS OF ALL-CAUSE MORTALITY. Age, pre-

0.05

sentation with ST-segment elevation on ECG, iEDV, log troponin, LVEF, and CMR diagnosis of cardiomyopathy were significant univariable predictors of mortality (p < 0.05 considered significant for uni-

0.00

variate analysis). In a multivariable model, only CMR 0

500

1000

1500

2000

2500

Follow Up in Days Myocarditis

Cardiomyopathy

MI

Normal

diagnosis of cardiomyopathy (hazard ratio: 3.0; 95% confidence interval: 1.08 to 8.37; p ¼ 0.034) and ECG presentation with ST-segment elevation (hazard ratio: 3.10; 95% confidence interval: 1.27 to 7.68; p ¼

B

0.013) remained significant (Table 5).

Log rank 7.3, p = 0.011

MORTALITY RISK MARKERS (ECG PRESENTATION WITH ST-SEGMENT ELEVATION D CMR DIAGNOSIS).

A risk assessment tool for predicting mortality was

0.15

constructed on the basis of 2 variables: CMR diagnosis Cumulative Mortality

6

of cardiomyopathy and ECG presentation with STsegment elevation. The first group was defined as

0.10

242 patients with any other CMR diagnosis except cardiomyopathy and presentation as NSTE-ACS. The second group consisted of 122 patients with only 1 of the factors altered (i.e., either a CMR diagnosis of

0.05

cardiomyopathy segment

or

elevation).

ECG

presentation

Finally,

the

with

third

ST-

group

included 24 patients with CMR diagnosis of cardio-

0.00

myopathy and ECG presentation with ST-segment 0

500

1000

1500

2000

2500

Follow Up in Days Takotsubo Cardiomyopathy

elevation. The mortality rates in these 3 groups were 1.6% (4 of 242), 10.6% (13 of 122), and 20.8% (5 of

Any Other Diagnosis

24), respectively. Kaplan-Meier curves were drawn showing the risk of mortality in these 3 groups,

(A) Kaplan-Meier curves showing the risk of mortality according to CMR

showing a strong association between number of risk

diagnosis (MI vs. myocarditis vs. cardiomyopathy vs. normal). (B) Kaplan-

markers and mortality (log-rank test: 23.6; p < 0.001)

Meier curves showing the risk of mortality according to CMR diagnosis of Takotsubo cardiomyopathy versus any other diagnosis. MI ¼ myocardial infarction; other abbreviation as in Figure 1.

(Figure 5). POST HOC POWER CALCULATION. Using mortality

of 15% in the cardiomyopathy group versus 3% for the difference in the use of medications among the different diagnostic categories (Table 4). PROGNOSIS

IN

DIFFERENT

CMR

CATEGORIES.

rest of the cohort with an alpha of 0.05, our study demonstrated 95% power.

DISCUSSION

In a median follow-up of 1,262 days (3.5 years), the overall all-cause mortality was 5.7%, with worst

This is the largest cohort to date of patients with

prognosis identified in the cardiomyopathy group

MINOCA assessed with CMR, and the first study to our

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019

Dastidar et al.

- 2019:-–-

knowledge to assess the prognostic role of CMR and conventional risk factors in patients with MINOCA.

7

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

T A B L E 5 Univariable and Multivariable Association for Mortality

The main findings are as follows: 1) CMR can identify

Univariate Analysis

a diagnosis in 3 of 4 patients presenting with MINOCA

Multivariate Analysis

95% CI

and suitable for CMR assessment; 2) CMR diagnosis of

95% CI

Significance

HR

Lower

Upper

Significance

HR

Lower

0.008

2.046

1.203

3.48

0.100

1.645

0.908 2.978

Upper

0.669

3.659

cardiomyopathy (imaging) and ECG presentation with

Age group

ST-segment elevation (conventional risk marker)

Sex

0.302

1.564

were the strongest predictors of mortality; and 3) a

BMI

0.484

0.968 0.884

1.06

Log-peak troponin

0.041

0.509

0.267

0.971

0.094

0.565

0.290

1.101

0.01

3.059

1.307

7.157

0.013

3.125

1.271

7.684

combined risk assessment tool using the 2 parameters (cardiomyopathy and ECG presentation with ST-

STEMI iEDV

0.008

1.016

1.004

1.028

0.258

1.01

0.993

1.028

segment elevation) provides further risk stratifica-

LVEF

0.001

0.958

0.936

0.981

0.851

0.996

0.96

1.034

tion in these patients.

RWMA

0.132

1.905

0.823

4.41

DIAGNOSTIC ROLE OF CMR IN MINOCA. It is impor-

LGE

0.161

0.544

0.233

1.274

Edema

0.228

0.542

0.20

1.469

0.227

1.986 0.034

3.013

1.084

8.373

tant to identify a cause for MINOCA to guide ongoing

MI

0.472

0.672

Myocarditis

0.098

0.293 0.068

1.253

noninvasive imaging technique, CMR (done at a me-

CM

0.001

5.628

2.361

13.417

dian of 37 days from presentation) is able to identify a

Normal

0.091

0.285

0.067

1.219

management and provide patient guidance. As a

diagnosis in these patients in up to 3 of 4 cases. The diagnostic role of CMR in this patient group has been

CI ¼ confidence interval; CM ¼ cardiomyopathy; other abbreviations as in Tables 1 and 2.

demonstrated in a small number of previous studies with smaller sample sizes (11–17); however, the diagnostic yield of CMR in the literature was disparate,

univariable predictors of mortality, but not in multi-

ranging from as low as 30% to as high as 90% (11–17).

variable analysis. These findings are particularly

The variation in the diagnostic yield in the literature

useful in justifying the role of CMR tissue character-

can be explained by the use of incomplete CMR pro-

ization in MINOCA.

tocol (edema imaging was not always included) as

The cardiomyopathy group in our study was a

well as the timing of the test, which sometimes occurs

heterogenous group, with Takotsubo cardiomyopathy

several months after the acute event, leading to the

being the most common (43% of the cardiomyopathy

resolution of the reversible cardiac abnormalities

group). Takotsubo cardiomyopathy had the worst

(13,18).

prognosis over a median follow-up of 3.5 years (15%

In our cohort, the most common diagnosis was MI

overall mortality) when compared with any other

(25%, n ¼ 97), closely followed by myocarditis (25%,

CMR diagnosis. The findings are comparable to the

n ¼ 96) and cardiomyopathy (25%); however, the

recent study by Templin et al. (21) in which Takotsubo

literature suggests that myocarditis is much more

cardiomyopathy had a mortality of 5.6% per patient-

common than MI (10–13,15), contrary to our result.

year in long-term follow-up. CMR not only helps di-

The mean age in our study was much higher

agnose, it may also help identify the high risk

compared with the other studies on MINOCA, which

Takotsubo cardiomyopathy cases by delineating the

may have contributed to the higher prevalence of MI

amount of myocardial edema, RV involvement,

(14,15). Another possible explanation could be the use

thrombus, and LV outflow tract obstruction. CMR-

of established criteria (Lake Louise Criteria) for

derived apico-basal myocardial edema gradient may

diagnosing myocarditis (26), in addition to an inclu-

be used as a marker of malignant arrhythmia risk (28).

sion cutoff of <50% coronary artery stenosis (10–15).

CONVENTIONAL RISK FACTORS FOR PROGNOSIS

PROGNOSTIC ROLE OF CMR IN MINOCA. Of all the

ASSESSMENT. Of all the risk factors assessed, ECG

CMR parameters examined (including LV, LV, and LV

presentation with ST-segment elevation was an in-

mass; myocardial edema; LGE; and the overall CMR

dependent predictor of mortality. Age and log-peak

diagnosis), a CMR diagnosis of cardiomyopathy was

troponin were univariable predictors of mortality,

associated with the worst prognosis (15% mortality).

but not in multivariable analysis. Gender was not

On a multivariable analysis involving both traditional

associated with mortality. The finding on peak

markers and CMR characteristics, CMR-derived diag-

troponin is particularly important because it is in

nosis of cardiomyopathy remained significant as an

contrary to the regular practice. High peak troponin

independent predictor of mortality.

level is often used as an arbitrary prognostic marker

Among the other CMR parameters assessed, LVEF

because of its evidence from ACS trials (29). The

and LV indexed end-diastolic volume were significant

importance of the presentation ECG is a novel finding

Dastidar et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019 - 2019:-–-

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

F I G U R E 5 Prognostic Role of CMR Diagnosis and ECG in MINOCA

T2 STIR (Edema)

LGE

Risk stratification 0.30

Log rank 23.6, p < 0.001

0.25

Takotsubo Cardiomyopathy

Cumulative Mortality

8

0.20

0.15

0.10

Myocarditis

0.05

0.00 0

500

1000

1500

2000

2500

37 11 1

0 0 0

Follow Up (Days) A

Myocardial infarction

At Risk Group A - 242 Group B - 122 Group C - 24

239 115 22

B

186 82 16

C 93 31 6

A - Any CMR Diagnosis Except CM & Presentation as NSTE-ACS B - CMR Diagnosis of CM or Presentation with ST-Elevation on ECG C - CMR Diagnosis of CM & Presentation with ST-Elevation on ECG Normal

(Left) In a population of patients presenting with MINOCA, CMR identified 4 main categories: cardiomyopathy (Takotsubo), myocarditis, myocardial infarction, and normal. In the Takotsubo cardiomyopathy case, there is transmural edema of the mid-apical segments but no late gadolinium enhancement (LGE); in the myocarditis case, there is diffuse edema with diffuse patchy epicardial and midwall LGE (black arrow), subsequent endomyocardial biopsy confirmed giant cells myocarditis; in the myocardial infarction case, there is focal transmural edema of the apical anterolateral wall with corresponding focal transmural LGE (black arrow). No abnormalities were noted in the normal case. (Right) Kaplan-Meier curves showing the risk of mortality according to risk group (group A: any CMR diagnosis except cardiomyopathy and presentation as NSTE-ACS; group B: CMR diagnosis of cardiomyopathy or presentation as STEMI; group C: CMR diagnosis of cardiomyopathy and presentation as STEMI). Abbreviations as in Figures 1 and 3.

because it is often overlooked while managing pa-

diagnosis might be useful predictors for risk stratifi-

tients with MINOCA.

cation.

CLINICAL IMPLICATIONS. Our findings strengthen

confirmed in larger multicenter studies.

This

should

be

explored

LIMITATIONS. Several

further

and

the evidence that CMR is a clinically relevant nonin-

STUDY

vasive imaging modality for the assessment of pa-

consideration. This is a single-center study with

tients presenting with MINOCA. In a previous study,

relatively limited sample size. Although our study

we demonstrated that CMR in MINOCA leads to a

was designed to represent a real-world population,

change in diagnosis in 54% and change in manage-

we potentially might have excluded higher-risk pa-

ment in 41% of patients (18). The current study re-

tients with contraindications to CMR (e.g., creati-

inforces the effect of CMR because a diagnosis of

nine clearance <30 ml/min, intracardiac devices).

cardiomyopathy is associated with worst prognosis.

Our study did, however, include a broad range of

Any other diagnosis except cardiomyopathy had a

consecutive patients with MINOCA from a large

relatively low mortality (2% to 4%), thereby putting

catchment area, and only a limited number (2%) of

patients in a good prognostic category.

patients were excluded from CMR. There may have

limitations

merit

Currently, there is no risk stratification algorithm

been a referral bias because the study included pa-

for patients presenting with MINOCA. The results of

tients with MINOCA, referred for a CMR by the

our study suggest that ECG at presentation and CMR

physician providing the care; however, our regular

JACC: CARDIOVASCULAR IMAGING, VOL.

-, NO. -, 2019

Dastidar et al.

- 2019:-–-

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

clinical practice at the Bristol Heart Institute in-

ACKNOWLEDGMENTS The

cludes CMR in patients presenting with MINOCA.

Benny Lawton (superintendent radiographer) and the

The CMR was performed at a median of 37 days

team of cardiovascular magnetic resonance radiogra-

from presentation; this may have affected the diag-

phers in the Bristol Heart Institute for their help in

nostic pickup rate as well as the prevalence of

acquiring the cardiovascular magnetic resonance

myocardial edema because not all patients were

studies. They also thank Alan Davies, clinical data

imaged in the acute phase. Nevertheless, our study

manager at UHBristol NHS Foundation Trust, for

showed a significant prognostic effect irrespective of

providing the mortality data.

authors

thank

Chris

the timing of scan. Also, newer mapping techniques were not performed in this patient cohort. The

ADDRESS

normal CMR group had few patients with bundle

Bucciarelli-Ducci, CMR Unit, Bristol Heart Institute,

FOR

CORRESPONDENCE:

branch block and LGE in the insertion points, which

Upper Maudlin Street, Bristol BS2 8HW, United

may not be classed as entirely normal; however, the

Kingdom. E-mail: [email protected].

Dr.

Chiara

presence of diffuse myocardial injury was excluded on the basis of both the early gadolinium enhancement

and

T2-weighted

images

analysis.

PERSPECTIVES

Last,

myocardial biopsy was not carried out in these pa-

COMPETENCY IN MEDICAL KNOWLEDGE: CMR (median 37

tients;

days from presentation) enabled establishing a diagnosis in 74%

however,

well-established

and

clinically

validated CMR criteria were used for the diagnosis

of patients presenting with MINOCA. Patients presenting with

of individual cases.

ST-segment elevation on ECG and CMR diagnosis of cardiomy-

CONCLUSION

opathy have worse prognoses.

In a large cohort of patients with MINOCA, CMR (median 37 days from presentation) established a diagnosis in almost 3 of 4 cases. Among the conventional risk factors and CMR characteristics, STsegment elevation on presentation ECG and CMR diagnosis of cardiomyopathy were independent predictors of mortality. Combined analysis of CMR diagnosis and ECG at presentation may allow strati-

TRANSLATIONAL OUTLOOK: MINOCA is common and often represents a clinical dilemma. Current ESC/American College of Cardiology guidelines suggest some diagnostic tests, but there is no recommended workup and management of MINOCA. CMR is useful for both diagnosis and prognostication in MINOCA. A large multicenter prospective trial is warranted to confirm these results and thereby enable a tailored treatment strategy in this heterogeneous cohort.

fication of patients with poor outcomes.

REFERENCES 1. Vedanthan R, Seligman B, Fuster V. Global perspective on acute coronary syndrome: a burden on the young and poor. Circ Res 2014;114: 1959–75. 2. Roffi M, Patrono C, Collet J-P, et al. 2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2016;37: 267–315. 3. Ibanez B, James S, Agewall S, et al. 2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with STsegment elevation. Eur Heart J 2018;39:119–77. 4. Lindahl B, Baron T, Erlinge D, et al. Medical therapy for secondary prevention and longterm outcome in patients with myocardial infarction with nonobstructive coronary artery disease clinical perspective. Circulation 2017; 135:1481–9.

6. Maddox TM, Ho PM, Roe M, Dai D, Tsai TT, Rumsfeld JS. Utilization of secondary prevention therapies in patients with nonobstructive coronary artery disease identified during cardiac catheterization: insights from the National Cardiovascular Data Registry Cath-PCI Registry. Circ Cardiovasc Qual Outcomes 2010;3:632–41.

10. Dastidar AG, Rodrigues JCL, Ahmed N, Baritussio A, Bucciarelli-Ducci C. The role of cardiac MRI in patients with troponin-positive chest pain and unobstructed coronary arteries. Curr Cardiovasc Imaging Rep 2015;8:28. 11. Collste O, Sörensson P, Frick M, et al. Myocardial infarction with normal coronary ar-

7. Pasupathy S, Air T, Dreyer RP, Tavella R, Beltrame JF. Systematic review of patients presenting with suspected myocardial infarction and nonobstructive coronary arteries. Circulation 2015; 131:861–70.

teries 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 2013;273:189–96.

8. Gehrie ER, Reynolds HR, Chen AY, et al. Characterization and outcomes of women and men

12. Mahmoudi M, Harden S, Abid N, et al. Troponin-positive chest pain with unobstructed coronary arteries: definitive differential diagnosis using cardiac MRI. Br J Radiol 2012;85:e461–6.

with non-ST-segment elevation myocardial infarction and nonobstructive coronary artery disease: results from the Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes with Early. Am Heart J 2009; 158:688–94.

13. Monney PA, Sekhri N, Burchell T, et al. Acute myocarditis presenting as acute coronary syndrome: role of early cardiac magnetic resonance in its diagnosis. Heart 2011;97:1312–8.

5. Niccoli G, Scalone G, Crea F. Acute myocardial infarction with no obstructive coronary atherosclerosis: mechanisms and management. Eur Heart

9. Thygesen K, Alpert JS, Jaffe AS, et al. Fourth universal definition of myocardial infarction

14. Leurent G, Langella B, Fougerou C, et al. Diagnostic contributions of cardiac magnetic

J 2015;36:475–81.

(2018). J Am Coll Cardiol 2018;72:2231–64.

resonance imaging in patients presenting with

9

10

Dastidar et al.

JACC: CARDIOVASCULAR IMAGING, VOL.

Prognostic Role of CMR and Conventional Risk Factors in MINOCA

elevated troponin, acute chest pain syndrome and unobstructed coronary arteries. Arch Cardiovasc Dis 2011;104:161–70. 15. Assomull RG, Lyne JC, Keenan N, et al. The role of cardiovascular magnetic resonance in patients presenting with chest pain, raised troponin, and unobstructed coronary arteries. Eur Heart J 2007; 28:1242–9. 16. Chopard R, Jehl J, Dutheil J, et al. Evolution of acute coronary syndrome with normal coronary arteries and normal cardiac magnetic resonance imaging. Arch Cardiovasc Dis 2011;104:509–17. 17. Laraudogoitia Zaldumbide E, Pérez-David E, Larena JA, et al. The value of cardiac magnetic resonance in patients with acute coronary syndrome and normal coronary arteries. Rev Española Cardiol 2009;62:976–83. 18. Dastidar AG, Rodrigues JCL, Johnson TW, et al. Myocardial infarction with nonobstructed coronary arteries. JACC Cardiol Img 2017;10:1204–6. 19. Gräni C, Eichhorn C, Bière L, et al. Prognostic value of cardiac magnetic resonance tissue characterization in risk stratifying patients with

-, NO. -, 2019 - 2019:-–-

suspected myocarditis. J Am Coll Cardiol 2017;70: 1964–76.

different techniques and its implications. Eur Heart J Cardiovasc Imaging 2015;16:738–46.

20. Aquaro GD, Perfetti M, Camastra G, et al. Cardiac MR with late gadolinium enhancement in acute myocarditis with preserved systolic function. J Am Coll Cardiol 2017;70:1977–87.

26. Friedrich MG, Sechtem U, Schulz-Menger J, et al. Cardiovascular magnetic resonance in myocarditis: a JACC White Paper. J Am Coll Cardiol 2009;53:1475–87.

21. Templin C, Ghadri JR, Diekmann J, et al. Clin-

27. Kawai S, Kitabatake A, Tomoike H. Guidelines

ical features and outcomes of Takotsubo (stress) cardiomyopathy. N Engl J Med 2015;373:929–38.

for diagnosis of takotsubo (ampulla) cardiomyopathy. Circ J 2007;71:990–2.

22. Mittal TK, Reichmuth L, Ariff B, et al. Imaging diagnoses and outcome in patients presenting for primary angioplasty but no obstructive coronary artery disease. Heart 2016;102:1728–34.

28. Dastidar AG, Frontera A, Palazzuoli A, Bucciarelli-Ducci C. TakoTsubo cardiomyopathy: unravelling the malignant consequences of a benign disease with cardiac magnetic resonance. Heart Fail Rev 2015;20:415–21.

23. Dastidar AG, Rodrigues JC, Baritussio A, Bucciarelli-Ducci C. MRI in the assessment of ischaemic heart disease. Heart 2015;102:239–52. 24. McAlindon EJ, Pufulete M, Harris JM, et al. Measurement of myocardium at risk with cardiovascular MR: comparison of techniques for edema

29. Kontos MC, Shah R, Fritz LM, et al. Implication of different cardiac troponin I levels for clinical outcomes and prognosis of acute chest pain patients. J Am Coll Cardiol 2004;43:958–65.

imaging. Radiology 2015;275:61–70. 25. McAlindon E, Pufulete M, Lawton C, Angelini GD, Bucciarelli-Ducci C. Quantification of infarct size and myocardium at risk: evaluation of

KEY WORDS ACS unobstructed coronaries, CMR, MINOCA, myocardial infarction myocarditis, Takotsubo cardiomyopathy