Myocardial tissue characterization by cardiac magnetic resonance imaging using T1 mapping predicts ventricular arrhythmia in ischemic and non–ischemic cardiomyopathy patients with implantable cardioverter-defibrillators Zhong Chen, MBBS,*† Manav Sohal, MBBS,*† Tobias Voigt, PhD,* Eva Sammut, MBBS,*† Catalina Tobon-Gomez, PhD,* Nick Child, MBBS,*† Tom Jackson, MBBS,*† Anoop Shetty, MD,*† Julian Bostock, PhD,† Michael Cooklin, MD,† Mark O’Neill, MD, PhD,*† Matthew Wright, MD, PhD,† Francis Murgatroyd, MD,‡ Jaswinder Gill, MD,*† Gerry Carr-White, MD, PhD,† Amedeo Chiribiri, MD, PhD,*† Tobias Schaeffter, PhD,* Reza Razavi, MD,*† C. Aldo Rinaldi, MD*† From the *Division of Imaging Science and Biomedical Engineering, Kings College London, London, United Kingdom, †Department of Cardiology, Guy’s and St Thomas’ NHS Trust, London, United Kingdom, and ‡ Department of Cardiology, King’s College Hospital NHS Trust, London, United Kingdom. BACKGROUND Diffuse myocardial fibrosis may provide a substrate for the initiation and maintenance of ventricular arrhythmia. T1 mapping overcomes the limitations of the conventional delayed contrast-enhanced cardiac magnetic resonance (CE-CMR) imaging technique by allowing quantification of diffuse fibrosis. OBJECTIVE The purpose of this study was to assess whether myocardial tissue characterization using T1 mapping would predict ventricular arrhythmia in ischemic and non–ischemic cardiomyopathies. METHODS This was a prospective longitudinal study of consecutive patients receiving implantable cardioverter-defibrillators in a tertiary cardiac center. Participants underwent CMR myocardial tissue characterization using T1 mapping and conventional CE-CMR scar assessment before device implantation. The primary end point was an appropriate implantable cardioverter-defibrillator therapy or documented sustained ventricular arrhythmia. RESULTS One hundred thirty patients (71 ischemic and 59 non– ischemic) were included with a mean follow-up period of 430 ⫾ 185 days (median 425 days; interquartile range 293 days). At follow-up, 23 patients (18%) experienced the primary end point. In multivariableadjusted analyses, the following factors showed a significant association with the primary end point: secondary prevention (hazard ratio [HR] 1.70; 95% confidence interval [95% CI] 1.01–1.91), noncontrast T1_native for every 10-ms increment in value (HR 1.10; CI 1.04–1.16;

Dr Chen has received an investigator-led research grant from Medtronic and a Clinical Research Training Fellowship grant from the British Heart Foundation. Address reprint requests and correspondence: Dr Zhong Chen, Department of Imaging Science and Biomedical Engineering, The Rayne Institute, King’s College London, St. Thomas’ Hospital, 4th Fl Lambeth Wing, London SE1 7EH, United Kingdom. E-mail address: [email protected].

1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.

90-ms difference between the end point–positive and end point– negative groups), and Grayzone_2sd-3sd for every 1% left ventricular increment in value (HR 1.36; CI 1.15–1.61; 4% difference between the end point–positive and end point–negative groups). Other CECMR indices including Scar_2sd, Scar_FWHM, and Grayzone_2sd-FWHM were also significantly, even though less strongly, associated with the primary end point as compared with Grayzone_2sd-3sd. CONCLUSION Quantitative myocardial tissue assessment using T1 mapping is an independent predictor of ventricular arrhythmia in both ischemic and non–ischemic cardiomyopathies. KEYWORDS Ventricular arrhythmia; T1 mapping; Gray zone; Defibrillator ABBREVIATIONS CE-CMR ¼ contrast-enhanced cardiac magnetic resonance; COV ¼ coefficient of variation; ECV ¼ extracellular volume index; FWHM ¼ full-width half-maximum; HR ¼ hazard ratio; ICD ¼ implantable cardioverter-defibrillator; ICM ¼ ischemic cardiomyopathy; LGE ¼ late gadolinium enhancement; LV ¼ left ventricular/ventricle; %LV ¼ percentage of the LV mass; NICM ¼ non–ischemic cardiomyopathy; ROC ¼ receiver-operatingcharacteristic; SI ¼ signal intensity; VT ¼ ventricular tachycardia (Heart Rhythm 2015;12:792–801) I 2015 Heart Rhythm Society. All rights reserved.

Introduction Ventricular arrhythmia is a major cause of sudden cardiac death in patients with impaired left ventricular (LV) function. Implantable cardioverter-defibrillators (ICDs) are effective http://dx.doi.org/10.1016/j.hrthm.2014.12.020

Chen et al

T1 Mapping Predicts Ventricular Arrhythmia

in reducing mortality but are associated with both perioperative complications and postimplant morbidity including inappropriate therapies.1,2 Risk stratification is therefore critical to select patients most likely to benefit from ICD therapy. Although LV ejection fraction plays a central role in patient selection, there is considerable risk heterogeneity among patients with impaired LV ejection fraction.3 Myocardial fibrosis provides a potential substrate for the initiation and maintenance of reentrant ventricular arrhythmia circuits, which propagate around localized regions of scars and along zones of slow conduction containing a mixture of scar and “healthy” myocardial tissue (gray zone). Cardiac magnetic resonance imaging techniques using late contrast enhancement (CE-CMR) are considered the criterion standard for the noninvasive assessment of myocardial regional fibrosis, with a growing body of evidence showing that the presence and magnitude of the region of delayed contrast enhancement predict arrhythmia in ischemic and non–ischemic cardiomyopathies (ICM and NICM).4–7 Diffuse LV fibrosis has also been shown to be arrhythmogenic by affecting electrical propagation between individual myocytes, which might provide evidence of increased risk of ventricular arrhythmia.8,9 The assessment of diffuse fibrosis as seen in patients with NICM is not possible with conventional CE-CMR techniques, since they rely on an arbitrary scale of the relative signal intensity (SI) difference detected between regions of dense scar and regions of user-defined “normal” tissue. Even in patients with ICM, the noninfarct regions of the LV seen as “normal” myocardium on CE-CMR imaging may contain diffuse interstitial fibrosis as a result of adverse remodeling.10,11 Myocardial tissue characterization using T1 mapping has emerged as a new CMR application, which has the potential to overcome the limitations of conventional CECMR techniques and allows characterization of diffuse fibrosis, providing a reliable quantitative assessment of myocardial tissue on a standardized scale. Several studies have shown a good correlation between measured T1 values and its associated derivative extracellular volume index (ECV) with quantitative histological measures of fibrosis.12–14 We therefore hypothesized that T1 mapping techniques, which enable better assessment of diffuse myocardial fibrosis, would help predict the occurrence of ventricular arrhythmia in patients with ICM or NICM undergoing ICD implantation.

Methods Study population After local research ethics committee approval, we prospectively invited all patients undergoing ICD implantation for primary and secondary preventions as per guidelines of our institution between May 2011 and January 2013.15 More than 90% of patients gave written informed consent and were recruited to the study. All study participants underwent CMR assessment and coronary angiography before device implant. Ischemic etiology was defined as the presence of any

793 epicardial coronary artery stenosis with more than 75% or any history of myocardial infarction or coronary revascularization with a scar pattern consistent with myocardial infarction on CMR imaging. Non–ischemic etiology was defined as the absence of the above criteria. CMR assessment was performed at least 6 months after previous myocardial infarction. Patients undergoing ICD implantation for catecholaminergic ventricular tachycardia (VT), Brugada syndrome, or long QT syndromes with normal CMR findings were not included in the final analysis.

CMR imaging protocol CMR imaging was performed using a 1.5-T scanner with a 32-channel coil (Philips Healthcare, Best, The Netherlands). After standardized patient-specific planning, a stack of breath-hold short-axis balanced steady-state free precession cine slices covering the LV was acquired for the quantification of volume, mass, and ejection fraction. Following a Look-Locker scan to find the optimum inversion time, a stack of short-axis slices of an inversion-recovery gradientecho sequence (repetition time/echo time ¼ 3.4/2.0 ms; flip angle ¼ 251; voxel size ¼ 1.8  1.8  8 mm; electrocardiogram triggered to end-diastole with a patient-adapted prepulse delay) was acquired 10–15 minutes after contrast injection (gadobutrol 0.2mmol/kg body weight) for CECMR scar assessment. A single midventricular short-axis slice (in the same geometry as the midventricular CE-CMR slice) using a modified Look-Locker inversion-recovery sequence with 11 phases (3 þ 3 þ 5) was acquired for T1 mapping (repetition time/echo time ¼ 3.3/1.5 ms; flip angle ¼ 501; voxel size ¼ 1.8  1.8  8 mm; with heart rate– adapted trigger delay and adiabatic prepulse to achieve a complete inversion) both before and after the administration of contrast.

Image analysis Ventricular volumes, function, and mass were analyzed using dedicated software (ViewForum, Philips Healthcare) and were indexed to body surface area. CE-CMR images were visually assessed by 2 independent CMR experts blinded to the study outcome (R.R. and A.C.). In case of discrepancy, the data were jointly reviewed by these experts to reach a consensus. The quantitative analysis of late gadolinium enhancement (LGE) was performed using CMR42 (Circle Cardiovascular Imaging Inc., Calgary, Canada). The extent of LGE was quantified using both the 2-standard deviation method (Scar_2sd, defined as the region with SI 2 SD above the remote reference myocardium) and the full-width half-maximum (FWHM) method (Scar_FWHM, defined as the region with SI above the 50% of the maximal SI of the enhanced area). The extent of the gray zone*** was quantified using the Grayzone_2sd-FWHM method (defined as the region with SI between Scar_FWHM and Scar_2sd) and the Grayzone_2sd-3sd method (defined as the region with SI between 2 and 3 SD above the remote reference

794 myocardium). Each of these indices was expressed as a percentage of the LV mass (%LV). T1 relaxation maps were processed using a customized software plug-in, incorporating motion correction with open source software (OsiriX environment, Pixmeo, Geneva, Switzerland). We performed motion correction by applying a multiresolution B-spline deformable registration, as implemented in the Insight Toolkit.16 Because of the potential susceptibility of high T1 values to the effects of heart rate during image acquisition, we performed linear regression analysis of the measured native noncontrast T1 (T1_native) values on heart rate and applied heart rate correction using a function of the mean heart rate of our study population to the noncontrast T1 values, as previously described.17 The T1associated derivative ECV was also calculated using the following formula, as previously described13: ð1hematocritÞ   ð1=postcontrast T1 myocardiumÞð1=precontrast T1 myocardiumÞ  ð1=postcontrast T1 bloodÞð1=precontrast T1 bloodÞ

We chose a region of interest in the septum for the estimation of noncontrast T1 (T1_native), postcontrast T1 (T1_contrast), and ECV, as previously described.18 Care was taken to avoid the endocardium/blood pool interface. If regional enhancement was seen in the septum on the LGE image, a septal region of interest was chosen adjacent to the enhanced region for T1 mapping analysis. The geometry of the midventricular T1 map slice was the same as that of the midventricular LGE images, which allows this comparison.

Follow-up and end point All patients were implanted with an ICD or an ICD combined with cardiac resynchronization therapy device capable of storing electrograms that met the criteria for detection. A standardized program for ventricular arrhythmia detection and therapy was used: ventricular arrhythmias 4170 beats/min were detected (detection count: 16 intervals) and treated with antitachyarrhythmia pacing and then shock therapy if antitachyarrhythmia pacing failed. Ventricular arrhythmias 4210 beats/min were detected (detection count: 24/30 intervals) and were treated with shock therapy as the first line. Standard supraventricular tachycardia detection discriminators were enabled according to the recommendations of device manufacturers (St Jude Medical Inc, St Paul, MN, and Medtronic, Minneapolis, MN; onset: 81% for Medtronic devices and 18% for St Jude Medical devices; stability: 40 ms for both types of devices; morphology: passive). These settings were altered only if clinically indicated. Patients were followed up at 3-month intervals. At each follow-up visit, the device was interrogated for system integrity and any recorded events were reviewed by an experienced device physiologist and a trained electrophysiologist who were blinded to the CMR data. The primary end point for our study was the delivery of appropriate ICD therapy for VT or ventricular fibrillation, or sustained VT of greater than 30-second duration as documented by the device. Multiple therapies may occur in a single triggered

Heart Rhythm, Vol 12, No 4, April 2015 event, and we defined any additional appropriate therapy 5 minutes apart from the preceding therapy to be a separate episode of therapy.

Statistical analysis The baseline characteristics of the participants were studied in detail; continuous variables are reported as mean ⫾ SD and categorical variables as number (percentage) of participants. Between-group comparisons were made using the t test for normally distributed continuous variables; otherwise, the Mann-Whitney U or Wilcoxon test was used (or the χ2 test for categorical variables), with a 2-tailed P value of o.05 indicating statistically significant differences between the 2 groups. The first episode of appropriate ICD therapy or sustained VT was considered as the event of interest for quantifying various associations. Kaplan-Meier survival curves were plotted to study the cumulative event rates between groups of participants, with the log-rank test providing further evidence regarding any significant differences between them. The univariate association between prespecified variables of interest and the primary end point was examined using Cox proportional hazards regression models. For multivariable-adjusted analyses, a forward selection process using variables significantly associated with the outcome of interest in univariate analyses (at P o .05) was used to further determine any independent predictors of the primary end point. All reported associations in this study are hazard ratios (HRs) and their corresponding 95% confidence intervals. To overcome limitations of missing data, only patients who had complete information on CE-CMR–derived indices and T1 mapping–derived indices were included in our multivariate analysis. Moreover, to avoid spurious associations due to the inclusion of correlated variables in a single Cox model, the multivariate analysis studied each of the CE-CMR indices separately. Receiver operating characteristic (ROC) curves were plotted to identify variable cutoff points with 90% sensitivity to discriminate the primary end point. The estimated cutoff values for the chosen scar index and T1 index to predict the primary end point was retrospectively used to reclassify and dichotomize the study subjects into high- and low-risk categories. All data analyses were performed using SPSS statistical software (version 21, IBM, New York, NY).

Ethics The study protocol was approved by the South East London Research Ethics Committee.

Results Study population A total of 138 patients agreed to participate in the study, of whom 3 did not complete the CMR protocol owing to claustrophobia and 2 additional participants were lost to follow-up. Three patients with Brugada syndrome or long QT syndromes were excluded from the analysis, leaving a total of 130 patients for further analysis. Of these, 71 patients

Chen et al Table 1

T1 Mapping Predicts Ventricular Arrhythmia

Baseline patient characteristics

Characteristic No. of patients (%) Age (y) Sex: female ICD CRT-D 11 prevention 21 prevention COPD Diabetes Hypertension CVA AF eGFR o60 mL/min/1.73 m2

ICM

NICM

P

71 (100) 66 ⫾ 10 14 (20) 34 (48) 37 (52) 48 (68) 23 (32) 10 (14) 14 (20) 25 (35) 6 (8) 14 (20) 25 (35)

59 (100) 58 ⫾ 18 12 (20) 25 (42) 34 (57) 44 (75) 15 (25) 10 (17) 7 (12) 20 (34) 2 (3) 19 (32) 15 (25)

NS NS NS NS NS NS NS NS NS NS NS NS

LVEF (%) 26 ⫾ 11 QRS duration (ms) 123 ⫾ 25 Preimplant NSVT (11 prevention) 13 (27)

35 ⫾ 18 124 ⫾ 30 17 (38)

NS NS NS

Aspirin/clopidgrel* β-Blocker Ca antagonist ACEi/ARB Statins* Amiodarone* Digoxin Loop diuretic K-sparing diuretic

65 (92) 70 (99) 2 (3) 69 (97) 65 (92) 12 (17) 7 (10) 27 (38) 33 (45)

28 (47) 53 (90) 3 (5) 57 (97) 37 (63) 3 (5) 6 (10) 30 (51) 25 (42)

o.001 NS NS NS o.001 .036 NS NS NS

Values are presented as mean ⫾ SD or as n (%). ACEi/ARB ¼ angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; AF ¼ atrial fibrillation; COPD ¼ chronic obstructive pulmonary disease; CRT-D ¼ cardiac resynchronization therapy with defibrillator; CVA ¼ cerebral vascular accident; eGFR ¼ estimated glomerular filtration rate; ICD ¼ implantable cardioverter-defibrillator; ICM ¼ ischemic cardiomyopathy; LVEF ¼ left ventricular ejection fraction; NICM ¼ non– ischemic cardiomyopathy; NSVT ¼ nonsustained ventricular tachycardia. * Variable with a statistically significant difference between the 2 groups.

had ICM and 59 had NICM (1 with sarcoidosis, 5 with hypertrophic cardiomyopathy, and 53 with dilated cardiomyopathy). One hundred study participants underwent T1 myocardial tissue characterization, and of this subset, 95

795 paired native and postcontrast T1 mapping images were suitable for analysis and 5 T1 mapping images were not suitable for analysis because of image corruption due to excessive motion artifacts. Table 1 summarizes the baseline characteristics of all 130 participants according to the underlying etiology.

Primary end point During a mean follow-up period of 430 ⫾ 185 days (median 425 days; interquartile range 293 days), 23 patients (18%) experienced appropriate ICD therapy. Three patients experienced more than 1 episode of appropriate ICD therapy. The median time to first ICD therapy was 178 days (interquartile range 420 days). Of the 4 cardiovascular deaths (3 in ICM and 1 in NICM groups; 1 died of primary heart failure and 3 died of heart failure after sepsis), only 1 patient had experienced appropriate ICD therapy after implant and before death. There was a greater proportion of appropriate ICD therapy in the secondary (11 of 38 [29%]) vs primary (12 of 92 [13%)) prevention group (P ¼ .031), as was the cumulative event rate for the primary end point (P ¼ .033) (Figure 1A). The cumulative event rate was similar, however, between the ICM and NICM cohorts (P ¼ .479; Figure 1B).

CE-CMR and T1 myocardial tissue characteristics When patients were dichotomized into end point–positive and end point–negative groups, there were significant differences in the CE-CMR–derived indices, with both scar and gray zone being significantly higher in the end point–positive group. In addition, the noncontrast native T1 value was significantly higher in the end point–positive group (Table 2).

Predictors of ICD therapy Univariate analysis revealed that secondary prevention, CECMR–derived indices of scar and gray zone (n ¼ 130), and

Figure 1 Kaplan-Meier survival curves of appropriate ICD therapy for (A) primary vs secondary prevention and for (B) ICM vs NICM. ICD ¼ implantable cardioverter-defibrillator; ICM ¼ ischemic cardiomyopathy; NICM ¼ non–ischemic cardiomyopathy.

796 Table 2

Heart Rhythm, Vol 12, No 4, April 2015 CMR-derived indices Negative end point

Positive end point

Index ICM LVEF (%) LV mass (g) Scar_2sd (%LV)* Scar_FWHM (%LV)* Grayzone_2sd-FWHM (%LV)* Grayzone_2sd-3sd (%LV)* T1_native (ms)*† T1_contrast (ms)† ECV†

28 ⫾ 10 146 ⫾ 35 23 ⫾ 8 14 ⫾ 6 9⫾4 5⫾2 1031 ⫾ 70 418 ⫾ 56

NICM 32 ⫾ 15 151 ⫾ 42 18 ⫾ 11 11 ⫾ 7 8⫾5 5⫾3 1033 ⫾ 65 421 ⫾ 72

ICM

36 ⫾ 19

31 ⫾ 13

155 ⫾ 49

162 ⫾ 33

13 ⫾ 12

33 ⫾ 9

7⫾8

19 ⫾ 7

6⫾6

14 ⫾ 7

5⫾4

8⫾2

1035 ⫾ 59

1088 ⫾ 63

425 ⫾ 87

421 ⫾ 34

0.29 ⫾ 0.06 0.29 ⫾ 0.06 0.29 ⫾ 0.06

NICM 30 ⫾ 12 155 ⫾ 34 30 ⫾ 11 18 ⫾ 8 12 ⫾ 7 9⫾3 1123 ⫾ 80 415 ⫾ 43

30 ⫾ 12 144 ⫾ 36 26 ⫾ 13 15 ⫾ 9 11 ⫾ 7 9⫾5 1158 ⫾ 93 408 ⫾ 56

0.31 ⫾ 0.05 0.31 ⫾ 0.04 0.32 ⫾ 0.07

P .893 .588 o.001 o.001 .006 o.001 .029 .592 .262

T1_native o 1000 ms was considered to be the normal reference in our institution. A 1.5-T MRI scanner with a 32-channel coil (Philips Healthcare) was used. There were no significant differences (P 4 .05) in age, sex distribution, comorbidities such as hypertension, diabetes, renal impairment, and patient heart rate during image acquisition between the end point–negative and end point–positive groups. ECV ¼ extracellular volume index; FWHM ¼ full-width half-maximum; ICM ¼ ischemic cardiomyopathy; LV ¼ left ventricular; LVEF ¼ left ventricular ejection fraction; NICM ¼ non–ischemic cardiomyopathy. * Variable with a statistically significant difference between the 2 groups. † T1-derived indices (n ¼ 95).

noncontrast native T1 values (n ¼ 95) were significantly associated with appropriate ICD therapy (Table 3). In multivariable-adjusted analysis, secondary prevention, CECMR–derived indices of scar and gray zone, and noncontrast native T1 values were all significantly associated with ICD therapy (Table 3). To overcome limitations of missing data, only patients who had complete information on CE-CMR– derived indices and T1 mapping–derived indices were included in our multivariate analysis. The HR of secondary prevention was 1.70; a mean difference of 4% in Grayzone_2sd-3sd between the end point–positive and end point– negative groups gave a 2.4-fold excess risk of the primary outcome. Similarly, for noncontrast native T1, a mean difference of 90 ms between the end point–positive and end point–negative groups was associated with a 1.4-fold excess risk of the primary outcome. The respective HR with the corresponding confidence interval for each index is given in Table 3.

Risk reclassification Risk reclassification was assessed for Grayzone_2sd-3sd and T1_native values that allowed 90% sensitivity for detecting the primary end point. Based on ROC analyses, a cutoff value of Z1015 ms was used for T1_native and a value of Z5.5% of the LV mass was used for Grayzone_2sd-3sd to dichotomize the study participants into high-risk (ie, those requiring an ICD) and low-risk (ie, those not requiring an ICD) groups. In the primary prevention group, 50 of the 80 participants (62.5%) who had an ICD implanted but who never experienced the end point were correctly reclassified

into the low-risk category on the basis of Grayzone_2sd-3sd data, while none of the participants experiencing a positive end point were mistakenly reclassified into the low-risk group. By comparison, 25 of the 57 participants (43.9%) with a primary prevention ICD and a negative end point were correctly reclassified to the low-risk group on the basis of T1_native measures, while none of the patients with a positive end point were mistakenly reclassified into the low-risk group. The net reclassification improvement was 62.5% and 26.3%, respectively, as compared with the original clinical risk classification. However, in the secondary prevention group, the use of neither Grayzone_2sd-3sd nor T1_native was effective in risk reclassification (net reclassification improvement 26.3% and 7.9%, respectively; Figure 2A). The value of using risk reclassification is illustrated in Figure 2B by using Kaplan-Meier survival curves. Risk restratification using either the Grayzone_2sd-3sd or T1_native value to high- or low-risk categories resulted in significantly different cumulative event rates in the primary prevention cohort but not in the secondary prevention cohort.

Reproducibility of T1 measurement T1 measurements were repeated in a subgroup of randomly selected patients from the study cohort (10 with ICM and 10 with NICM). For T1_native measurement, the intraobserver average difference in values was 5 ⫾ 3 ms and the coefficient of variation (COV) was 0.3% ⫾ 0.2%; the interobserver average difference in values was 6 ⫾ 5 ms and the COV was 0.4% ⫾ 0.3%. For T1_contrast measurement, the intraobserver

Chen et al

T1 Mapping Predicts Ventricular Arrhythmia

Table 3 Findings from univariate and multivariate Cox regression analyses Variable

Hazard ratio

Confidence interval

P

0.99 1.53 2.16 2.23 1.89 2.84 1.48 1.37

0.96–1.03 0.45–5.16 0.84–5.55 0.87–5.68 0.83–4.29 0.83–9.69 0.61–3.59 0.75–2.24

NS NS NS NS NS NS NS NS

1.18 1.59 0.99 1.01 1.09 1.09 1.14

0.51–2.73 1.04–1.82 0.96–1.02 0.99–1.01 1.05–1.14 1.04–1.15 1.08–1.22

NS .039 NS NS o.001 .001 o.001

1.31 1.06 1.03 1.01

1.17–1.47 1.01–1.11 0.98–1.09 0.94–1.11

o.001 .021 NS NS

1.70

1.01–1.91

1.10 1.11 1.13

1.04–1.15 1.05–1.19 1.05–1.22

NS .048 NS o.001 .001 .001

1.36 1.10

1.15–1.61 1.04–1.16

o.001 .001

Univariate analyses Age (þ1 y) Sex: female COPD Diabetes Hypertension CVA AF eGFR o60 mL/min/1.73 m2 QRS duration Z120 ms 21 prevention LVEF (þ1%) LV mass (þ1 g) Scar_2sd (þ1%LV) Scar_FWHM (þ1%LV) Grayzone_2sd-FWHM (þ1% LV) Grayzone_2sd-3sd (þ1%LV) T1_native (þ10 ms) T1_contrast (10 ms) ECV (þ1%) Multivariate analyses*† QRS duration Z120 ms 21 prevention LVEF r35% Scar_2sd (þ1%LV) Scar_FWHM (þ1%LV) Grayzone_2sd-FWHM (þ1% LV) Grayzone_2sd-3sd (þ1%LV) T1_native (þ10 ms)

For multivariable-adjusted analyses, each of the factors listed in the table has been mutually adjusted for all the other factors mentioned for the corresponding regression analysis. ECV ¼ extracellular volume index; EF ¼ ejection fraction; FWHM ¼ fullwidth half-maximum; LV ¼ left ventricular; NS ¼ nonsignificant. * To overcome limitations of missing data, only patients who had complete information on CE-CMR–derived indices and T1 mapping–derived indices were included in our multivariate analysis (n ¼ 95). † To avoid spurious associations due to the inclusion of correlated variables in a single Cox model, the multivariate analysis studied each of the scar or gray zone indices separately, that is, for the analysis of Scar_2sd þ1%LV, the adjustment included QRS Z120 ms, 21 prevention, LVEF r35%, and T1_native þ10 ms; for the analysis of Scar_FWHM þ1%LV, the adjustment included QRS Z120 ms, 21 prevention, LVEF r35%, and T1_native þ10 ms; for the analysis of Grayzone_2sd-FWHM þ1%LV, the adjustment included QRS Z120 ms, 21 prevention, LVEF r35%, and T1_native þ10 ms; for the analysis of Grayzone_2sd-3sd þ1%LV, the adjustment included QRS Z120 ms, 21 prevention, LVEF r35%, and T1_native þ10 ms.

average difference in values was 5 ⫾ 4 ms and the COV was 0.8% ⫾ 0.6%; the interobserver average difference in values was 5 ⫾ 4 ms and the COV was 0.9% ⫾ 0.7%. There were good agreements in inter- and intrameasurements of native T1 and contrast T1 (Pearson r ¼ 0.99; P r .01).

797

Discussion The present study is the first to demonstrate that noncontrast native T1 estimated with T1 mapping is an independent predictor of ventricular arrhythmia in patients with ICM and NICM. It also provides further confirmation that the extent of scar and gray zone derived from CE-CMR techniques is independently associated with ventricular arrhythmia, and it highlights the importance of the gray zone in patients with NICM.

Diffuse fibrosis and ventricular arrhythmia

It has long been recognized that diffuse interstitial fibrosis occurs at various stages in the pathological progression of cardiomyopathies of different etiologies.10–12,19,20 Fibroblasts are not electrically “inert” and can modulate cellular ion channel remodeling, thereby affecting tissue electrical properties through a variety of mechanisms including mechanoelectrical feedback mediated via stretch-activated ion channels; gap junctions via connexin with close coupling of nearby cardiac myocytes; and altering the orientation of cardiac fibers. Collectively, these effects can provide a substrate for ventricular arrhythmia.21,22 An electroanatomic mapping study suggests that diffuse interstitial fibrosis plays a critical role in the maintenance of reentrant VT and propensity to degenerate to ventricular fibrillation.23 Studies using myocardial biopsy specimens to look for diffuse fibrosis in patients with deemed normal hearts and those with hypertrophic cardiomyopathy have also shown a strong association between interstitial fibrosis and ventricular arrhythmia.24,25 T1 mapping of the myocardium provides a true quantitative assessment of the voxel SI and is used to detect regions of diffuse interstitial fibrosis with higher native T1 values than normal myocardium. In comparison, conventional CECMR techniques require an arbitrary reference than that of the normal myocardium to generate an image with a signal contrast between fibrotic and apparently nonfibrotic regions. It is therefore less discriminative in “highlighting” the diffuse fibrosis that affects the myocardium in a more homogeneous way. Native T1 values reflecting the presence of diffuse interstitial fibrosis, not detectable by conventional CE-CMR techniques, have been found to be higher in myocardial regions free of delayed contrast enhancement in a number of different cardiomyopathies.26,27 The present study is the first to show that native myocardial T1 is independently associated with ventricular arrhythmia in patients with advanced cardiomyopathies of both ischemic and nonischemic etiologies. T1 values have also been measured after the administration of gadolinium contrast agents, and in addition, the patients’ hematocrit levels can be used to calculate ECV. Several studies have demonstrated a close correlation between ECV and histological measures of fibrosis in cardiomyopathies of various etiologies.13,28,29 Others have demonstrated a correlation between ECV and mortality.30–32Accurate ECV assessment depends on the time

798

Original clinical risk prediction Prediction Primary prevention Total 80 80 12 12 92 92

Total 27 27 11 11 38 38

Total 57 57 8 8 65 65

Total 21 21 9 9 30 30

Figure 2 Risk reclassification with Grayzone_2sd-3sd (high risk Z5.5% LV mass) or T1_native (high risk Z1015 ms) in the primary prevention and secondary prevention cohorts (A), with corresponding Kaplan-Meier curves for the reclassified high- and low-risk categories (B). LV ¼ left ventricular.

Heart Rhythm, Vol 12, No 4, April 2015

with Grayzone Low risk High risk 0 80 50* 30 Endpoint +ve 0 12 0 12 Total 0 92 50 42 Net reclassification improvement = 62.5% Secondary prevention Low risk High risk ICD event Endpoint -ve 0 27 12* 15 Endpoint +ve 0 11 2* 9 Total 0 38 14 24 Net reclassification improvement = 26.3% Original clinical risk prediction Prediction Primary prevention with T1 ICD event Low risk High risk Endpoint -ve 0 57 25* 32 Endpoint +ve 0 8 0 8 Total 0 65 25 40 Net reclassification improvement = 43.9% Secondary prevention ICD event Low risk High risk Endpoint -ve 0 21 3* 18 Endpoint +ve 0 9 2* 7 Total 0 30 5 25 Net reclassification improvement =-7.9% * denotes changes affected by risk reclassification. ICD event Endpoint -ve

Chen et al

T1 Mapping Predicts Ventricular Arrhythmia

delay of T1 mapping after contrast injection, the hematocrit level that affects the partition coefficient of the gadolinium contrast agent, the contrast kinetics between regions of fibrosis and normal myocardium, and the contrast washout rate that is influenced by the renal clearance rate. In the present study, ECV was not a significant independent predictor of ventricular arrhythmia that may be explained by several factors including the fact that the time delay after contrast injection for postcontrast T1 mapping varied between 10 and 25 minutes and that the hematocrit levels were measured using blood samples taken from a time range spanning 24 hours before and after the time of image acquisition owing to clinical constraints. It is known that hematocrit levels may vary significantly, depending on the time of the day, posture, physical activity, and postprandial state.33 Similarly, the formula from which ECV is derived holds true when the extracellular contrast concentration in the blood and myocardium reaches the equilibrium and there are uncertainties whether this equilibrium is achievable with the bolus technique we used rather than the slow contrast infusion technique.34 Finally, in our cohort of advanced cardiomyopathies, variations in diuretic therapy and varying degrees of renal function are likely to have an impact on renal clearance of gadolinium and therefore contrast kinetics.

Regional fibrosis and ventricular arrhythmia

Regional fibrosis provides an important substrate for ventricular arrhythmias and is an independent predictor of ventricular arrhythmia in both ICM and NICM, with several electroanatomic mapping studies demonstrating that the scar border zone (gray zone) is important in the initiation and maintenance of ventricular arrhythmia.35,36 Further characterization of the enhanced region with intermediate SI to further differentiate the gray zone from the scar core has provided incremental prediction of ventricular arrhythmia in ICM, but has not been studied so far in NICM.37,38 A criticism of gray zone assessment is the effect of partial volume and image acquisition voxel size that may influence its quantification. Furthermore, the multitude of SI-based quantitative methods used in various studies makes direct comparison of the predictive value of each index difficult. To overcome these limitations, automated computer analysis was used to improve the reproducibility of the quantification by standardizing the SI-based assessment; and a comparison was made between 2 gray zone indices on the same cohort of patients (Grayzone_2sd-3sd and Grayzone_2sd-FWHM). This study systematically tested the predictive values of each index expressed as %LV and found both to be independent predictors of ventricular arrhythmia in a multivariate model that included adjustments for established variables used for arrhythmia risk stratification. Grayzone_2sd-3sd gave the highest HR for every 1%LV mass increment as compared with other

799 indices, although ROC analysis revealed that the areas under the curve were not significantly different between the gray zone and other scar indices. This study is the largest cohort to date to demonstrate the incremental value of gray zone quantification for arrhythmia prediction and to highlight the potential of arrhythmia prediction by the gray zone in patients with NICM.

Clinical relevance Native T1 removes the dependence on contrast kinetics and makes its potential translation into routine clinical application attractive. In addition, it allows myocardial tissue characterization free of the potential risk–associated gadolinium contrast in patients with severe renal impairment. In the subanalysis assessing the impact of using Grayzone_2sd-3sd and T1_native values for risk restratification, additional risk assessment based on the cutoff values derived from this study could correctly reclassify patients in the primary prevention group from the high-risk to the low-risk category. In the secondary prevention group, some patients were incorrectly risk stratified to the low-risk group when tested using either Grayzone_2sd-3sd or T1_native alone; however, the specificity may be improved if both indices were considered simultaneously. These results are in line with other results that showed scar quantification with CE-CMR imaging as the only independent predictor of adverse ventricular arrhythmia outcome in patients with a history of nonsustained VT.39

Study limitations This study is limited by a relatively short follow-up and use of a surrogate end point (ie, appropriate ICD therapy), which does not necessarily equate to sudden arrhythmic death. The absolute T1 values of interest can be influenced by a number of technical parameters including magnetic field strength and acquisition sequence as well as postacquisition image processing including motion correction and image registration. T1 mapping was limited to just a single slice of myocardium and an assumption made that diffuse fibrosis imaged is truly diffuse/uniform in the remote myocardium. Conducting a single-center study allowed standardization of the imaging protocol across the entire cohort; however, it may have introduced systematic bias including the timing of postcontrast T1 measurement. Only 1 type of sequence was used for T1 image acquisition. A shorter T1 mapping acquisition sequence with shorter breath-holds during image acquisition may potentially reduce motion artifact and impact on the postcontrast T1 value analysis. In light of the results of the Multicenter Automatic Defibrillator Implantation Trial– Reduce Inappropriate Therapy (MADIT-RIT) study, optimized ICD programming aimed at reducing ICD therapy could have potentially reduced the event rate of the study.40 However, by standardizing the ICD programming within our cohort, it is unlikely that our programming as setup would have introduced any systemic bias in the associations

800 studied. This study included a heterogeneous patient population; although diffuse fibrosis may be present in advanced cardiomyopathies of various etiologies, a more homogeneous population group would have improved the power of detection of suggested association. A multicenter study with a larger sample would be of great value to further assess our findings.

Conclusion The study has shown for first time that quantitative myocardial tissue assessment using T1 mapping is an independent predictor of ventricular arrhythmia with an HR comparable to scar or gray zone quantification detected by conventional CE-CMR techniques in both ICM and NICM. Our results suggest a potential for the use of myocardial tissue characterization using CMR imaging to refine the current risk stratification decisions regarding ICD implantation, especially in the context of primary prevention of sudden arrhythmic death.

Acknowledgments We thank Siobhon Crichton, MSc, and Sreenivasa Rao Kondapally Seshasai MD PhD, for their helpful discussion on statistics and input in the manuscript preparation.

Appendix Supplementary data Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.hrthm. 2014.12.020.

References 1. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert JP, Higgins SL, Brown MW, Andrews ML. Multicenter Automatic Defibrillator Implantation Trial II Investigators. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877–883. 2. Bardy GH, Lee KL, Mark DB, et al. Amiodarone or an implantable cardioverterdefibrillator for congestive heart failure. N Engl J Med 2005;352:225–237. 3. Epstein AE, DiMarco JP, Ellenbogen KA, et al. ACC/AHA/HRS 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices) developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. J Am Coll Cardiol 2008;51:e1–e62. 4. Scott PA, Rosengarten JA, Curzen NP, Morgan JM. Late gadolinium enhancement cardiac magnetic resonance imaging for the prediction of ventricular tachyarrhythmic events: a meta-analysis. Eur J Heart Fail 2013;15:1019–1027. 5. Wu KC, Weiss RG, Thiemann DR, Kitagawa K, Schmidt A, Dalal D, Lai S, Bluemke DA, Gerstenblith G, Marbán E, Tomaselli GF, Lima JAC. Late gadolinium enhancement by cardiovascular magnetic resonance heralds an adverse prognosis in nonischemic cardiomyopathy. J Am Coll Cardiol 2008;51:2414–2421. 6. Gulati A, Jabbour A, Ismail TF, et al. Association of fibrosis with mortality and sudden cardiac death in patients with nonischemic dilated cardiomyopathy. JAMA 2013;309:896–908. 7. Neilan TG, Coelho-Filho OR, Danik SB, et al. CMR quantification of myocardial scar provides additive prognostic information in nonischemic cardiomyopathy. JACC Cardiovasc Imaging 2013;6:944–954.

Heart Rhythm, Vol 12, No 4, April 2015 8. Spach MS, Boineau JP. Microfibrosis produces electrical load variations due to loss of side-to-side cell connections: a major mechanism of structural heart disease arrhythmias. Pacing Clin Electrophysiol 1997;20:397–413. 9. Massare J, Berry JM, Luo X, Rob F, Johnstone JL, Shelton JM, Bassel-Duby R, Hill JA, Naseem RH. Diminished cardiac fibrosis in heart failure is associated with altered ventricular arrhythmia phenotype. J Cardiovasc Electrophysiol 2010;21:1031–1037. 10. Beltrami CA, Finato N, Rocco M, Feruglio GA, Puricelli C, Cigola E, Quaini F, Sonnenblick EH, Olivetti G, Anversa P. Structural basis of end-stage failure in ischemic cardiomyopathy in humans. Circulation 1994;89:151–163. 11. Jugdutt BI. Ventricular remodeling after infarction and the extracellular collagen matrix. Circulation 2003;108:1395–1403. 12. Iles L, Pfluger H, Phrommintikul A, Cherayath J, Aksit P, Gupta SN, Kaye DM, Taylor AJ. Evaluation of diffuse myocardial fibrosis in heart failure with cardiac magnetic resonance contrast-enhanced T1 mapping. J Am Coll Cardiol 2008;52: 1574–1580. 13. Flett AS, Hayward MP, Ashworth MT, Hansen MS, Taylor AM, Elliott PM, McGregor C, Moon JC. Equilibrium contrast cardiovascular magnetic resonance for the measurement of diffuse myocardial fibrosis: preliminary validation in humans. Circulation 2010;122:138–144. 14. Sibley CT, Noureldin RA, Gai N, Nacif MS, Liu S, Turkbey EB, Mudd JO, van der Geest RJ, Lima JAC, Halushka MK, Bluemke DA. T1 mapping in cardiomyopathy at cardiac MR: comparison with endomyocardial biopsy. Radiology 2012;265:724–732. 15. McMurray JJV, Adamopoulos S, Anker SD, et al. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2012. Eur Heart J 2012;33:1787–1847. 16. Ibáñez L, Schroeder W, Ng L, Cates J, Consortium TIS, Hamming R. The ITK Software Guide. Kitware, Inc; 2003. Available at: http://www.kitware.com. 17. Messroghli DR, Plein S, Higgins DM, Walters K, Jones TR, Ridgway JP, Sivananthan MU. Human myocardium: single-breath-hold MR T1 mapping with high spatial resolution—reproducibility study. Radiology 2006;238:1004–1012. 18. Rogers T, Dabir D, Mahmoud I, Voigt T, Schaeffter T, Nagel E, Puntmann VO. Standardization of T1 measurements with MOLLI in differentiation between health and disease—the ConSept study. J Cardiovasc Magn Reson 2013;15:78. 19. Brooks A, Schinde V, Bateman AC, Gallagher PJ. Interstitial fibrosis in the dilated non-ischaemic myocardium. Heart 2003;89:1255–1256. 20. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JAC. Assessment of myocardial fibrosis with cardiovascular magnetic resonance. J Am Coll Cardiol 2011;57:891–903. 21. Souders CA, Bowers SLK, Baudino TA. Cardiac fibroblast: the renaissance cell. Circ Res 2009;105:1164–1176. 22. Camelliti P, Borg TK, Kohl P. Structural and functional characterisation of cardiac fibroblasts. Cardiovasc Res 2005;65:40–51. 23. Pogwizd SM, McKenzie JP, Cain ME. Mechanisms underlying spontaneous and induced ventricular arrhythmias in patients with idiopathic dilated cardiomyopathy. Circulation 1998;98:2404–2414. 24. Sugrue DD, Holmes DR, Gersh BJ, Edwards WD, McLaran CJ, Wood DL, Osborn MJ, Hammill SC. Cardiac histologic findings in patients with lifethreatening ventricular arrhythmias of unknown origin. J Am Coll Cardiol 1984;4:952–957. 25. Almaas VM, Haugaa KH, Strøm EH, Scott H, Dahl CP, Leren TP, Geiran OR, Endresen K, Edvardsen T, Aakhus S, Amlie JP. Increased amount of interstitial fibrosis predicts ventricular arrhythmias, and is associated with reduced myocardial septal function in patients with obstructive hypertrophic cardiomyopathy. Europace 2013;15:1319–1327. 26. Dass S, Suttie JJ, Piechnik SK, Ferreira VM, Holloway CJ, Banerjee R, Mahmod M, Cochlin L, Karamitsos TD, Robson MD, Watkins H, Neubauer S. Myocardial tissue characterization using magnetic resonance noncontrast T1 mapping in hypertrophic and dilated cardiomyopathy. Circ Cardiovasc Imaging 2012;5: 726–733. 27. Puntmann VO, Voigt T, Chen Z, Mayr M, Karim R, Rhode K, Pastor A, CarrWhite G, Razavi R, Schaeffter T, Nagel E. Native T1 mapping in differentiation of normal myocardium from diffuse disease in hypertrophic and dilated cardiomyopathy. JACC Cardiovasc Imaging 2013;6:475–484. 28. Karamitsos TD, Piechnik SK, Banypersad SM, Fontana M, Ntusi NB, Ferreira VM, Whelan CJ, Myerson SG, Robson MD, Hawkins PN, Neubauer S, Moon JC. Noncontrast T1 mapping for the diagnosis of cardiac amyloidosis. JACC Cardiovasc Imaging 2013;6:488–497. 29. Sado DM, Flett AS, Banypersad SM, et al. Cardiovascular magnetic resonance measurement of myocardial extracellular volume in health and disease. Heart 2012;98:1436–1441. 30. Kellman P, Wilson JR, Xue H, Bandettini WP, Shanbhag SM, Druey KM, Ugander M, Arai AE. Extracellular volume fraction mapping in the myocardium, part 2: initial clinical experience. J Cardiovasc Magn Reson 2012;14:64.

Chen et al

T1 Mapping Predicts Ventricular Arrhythmia

31. Wong TC, Piehler KM, Kang IA, Kadakkal A, Kellman P, Schwartzman DS, Mulukutla SR, Simon MA, Shroff SG, Kuller LH, Schelbert EB. Myocardial extracellular volume fraction quantified by cardiovascular magnetic resonance is increased in diabetes and associated with mortality and incident heart failure admission. Eur Heart J 2014;35:657–664. 32. Moon JC, Messroghli DR, Kellman P, Piechnik SK, Robson MD, Ugander M, Gatehouse PD, Arai AE, Friedrich MG, Neubauer S, Schulz-Menger J, Schelbert EB. Myocardial T1 mapping and extracellular volume quantification: a Society for Cardiovascular Magnetic Resonance (SCMR) and CMR Working Group of the European Society of Cardiology consensus statement. J Cardiovasc Magn Reson 2013;15:92. 33. Mayer GA. Diurnal, postural and postprandial variations of hematocrit. Can Med Assoc J 1965;93:1006–1008. 34. White SK, Sado DM, Fontana M, Banypersad SM, Maestrini V, Flett AS, Piechnik SK, Robson MD, Hausenloy DJ, Sheikh AM, Hawkins PN, Moon JC. T1 mapping for myocardial extracellular volume measurement by CMR: bolus only versus primed infusion technique. JACC Cardiovasc Imaging 2013;6:955–962. 35. Estner HL, Zviman MM, Herzka D, Miller F, Castro V, Nazarian S, Ashikaga H, Dori Y, Berger RD, Calkins H, Lardo AC, Halperin HR. The critical isthmus sites of ischemic ventricular tachycardia are in zones of tissue heterogeneity, visualized by magnetic resonance imaging. Heart Rhythm 2011;8:1942–1949.

801 36. Perez-David E, Arenal A, Rubio-Guivernau JL, et al. Noninvasive identification of ventricular tachycardia-related conducting channels using contrast-enhanced magnetic resonance imaging in patients with chronic myocardial infarction comparison of signal intensity scar mapping and endocardial voltage mapping. J Am Coll Cardiol 2011;57:184–194. 37. Roes SD, Borleffs CJ, van der Geest RJ, Westenberg JJ, Marsan NA, Kaandorp TA, Reiber JH, Zeppenfeld K, Lamb HJ, de Roos A, Schalij MJ, Bax JJ. Infarct tissue heterogeneity assessed with contrast-enhanced MRI predicts spontaneous ventricular arrhythmia in patients with ischemic cardiomyopathy and implantable cardioverter-defibrillator. Circ Cardiovasc Imaging 2009;2: 183–190. 38. Schmidt A, Azevedo CF, Cheng A, Gupta SN, Bluemke DA, Foo TK, Gerstenblith G, Weiss RG, Marban E, Tomaselli GF, Lima JA, Wu KC. Infarct tissue heterogeneity by magnetic resonance imaging identifies enhanced cardiac arrhythmia susceptibility in patients with left ventricular dysfunction. Circulation 2007;115:2006–2014. 39. Dawson DK, Hawlisch K, Prescott G, Roussin I, Di Pietro E, Deac M, Wong J, Frenneaux MP, Pennell DJ, Prasad SK. Prognostic role of CMR in patients presenting with ventricular arrhythmias. JACC Cardiovasc Imaging 2013;6: 335–344. 40. Moss AJ, Schuger C, Beck C, et al. Reduction in inappropriate therapy and mortality through ICD programming. N Engl J Med 2012;367:2275–2283.

CLINICAL PERSPECTIVES Regional myocardial fibrosis, conventionally detected by contrast-enhanced (CE) cardiac magnetic resonance (CMR) imaging, provides a potential substrate for the initiation and maintenance of ventricular arrhythmia. Myocardial tissue characterization using T1 mapping has emerged as a new CMR application, which has the potential to overcome the limitations of conventional CE-CMR techniques and characterize diffuse fibrosis. The present study demonstrates that an incremental increase in noncontrast native T1 measurement (a marker of diffuse fibrosis) is independently associated with the incidence of ventricular arrhythmia in patients with implantable cardioverter-defibrillators (ICDs). The study also provides further confirmation that the extent of scar and gray zone derived from CE-CMR techniques is independently associated with ventricular arrhythmia, and it highlights the importance of the gray zone in patients with non–ischemic cardiomyopathy (NICM). Such CMR applications have the potential to improve risk stratification in selecting patients for ICD therapy and may be particularly useful for the primary prevention group in whom there is considerable risk heterogeneity. We acknowledge that there is currently a lack of standardization of quantitative myocardial tissue assessment with CMR imaging in clinical practice, and we are limited by our continuing understanding of this emerging T1 mapping technique. Larger cohort studies are needed to establish the “ideal” threshold values of diffuse and regional fibrosis to enable the translation of noninvasive myocardial fibrosis quantification using CMR T1 mapping into the clinical management of patients at risk of life-threatening arrhythmias.