Key Points
Question
In patients with suspected cardiac sarcoidosis (CS), are cardiovascular magnetic resonance (CMR) phenotypes based on left ventricular ejection fraction (LVEF) and late gadolinium enhancement (LGE) associated with long-term adverse outcomes?
Findings
In this cohort study of 504 patients with histologically proven sarcoidosis, a CMR phenotype involving pathology-frequent LGE features was associated with a high risk of ventricular arrhythmic and heart failure events, while the absence of the phenotype was associated with a low risk of arrhythmic events, even in the presence of LGE or abnormal LVEF.
Meaning
These findings suggest that CMR phenotypes could be used to optimize clinical decision-making for treatment options, such as implantable cardioverter-defibrillators, and thus improve outcomes in patients with suspected CS.
This cohort study evaluates the association of cardiovascular magnetic resonance phenotypes based on left ventricular ejection fraction and late gadolinium enhancement with adverse outcomes in patients with suspected cardiac sarcoidosis.
Abstract
Importance
In patients with sarcoidosis with suspected cardiac involvement, late gadolinium enhancement (LGE) on cardiovascular magnetic resonance imaging (CMR) identifies those with an increased risk of adverse outcomes. However, these outcomes are experienced by only a minority of patients with LGE, and identifying this subgroup may improve treatment and outcomes in these patients.
Objective
To assess whether CMR phenotypes based on left ventricular ejection fraction (LVEF) and LGE in patients with suspected cardiac sarcoidosis (CS) are associated with adverse outcomes during follow-up.
Design, Setting, and Participants
This cohort study included consecutive patients with histologically proven sarcoidosis who underwent CMR for the evaluation of suspected CS from 2004 to 2020 with a median follow-up of 4.3 years at an academic medical center in Minnesota. Demographic data, medical history, comorbidities, medications, and outcome data were collected blinded to CMR data.
Exposures
CMR phenotypes were identified based on LVEF and LGE presence and features. LGE was classified as pathology-frequent or pathology-rare based on the frequency of cardiac damage features on gross pathology assessment of the hearts of patients with CS who had sudden cardiac death or cardiac transplant.
Main Outcomes and Measures
Composite of ventricular arrhythmic events and composite of heart failure events.
Results
Among 504 patients (mean [SD] age, 54.1 [12.5] years; 242 [48.0%] female and 262 [52.0%] male; 2 [0.4%] American Indian or Alaska Native, 6 [1.2%] Asian, 90 [17.9%] Black or African American, 399 [79.2%] White, 5 [1.0%] of 2 or more races (including the above-mentioned categories and Native Hawaiian or Other Pacific Islander), and 2 [0.4%] of unknown race; 4 [0.8%] Hispanic or Latino, 498 [98.8%] not Hispanic or Latino, and 2 [0.4%] of unknown ethnicity), 4 distinct CMR phenotypes were identified: normal LVEF and no LGE (n = 290; 57.5%), abnormal LVEF and no LGE (n = 53; 10.5%), pathology-frequent LGE (n = 103; 20.4%), and pathology-rare LGE (n = 58; 11.5%). The phenotype with pathology-frequent LGE was associated with a high risk of arrhythmic events (hazard ratio [HR], 12.12; 95% CI, 3.62-40.57; P < .001) independent of LVEF and extent of left ventricular late gadolinium enhancement (LVLGE). It was also associated with a high risk of heart failure events (HR, 2.49; 95% CI, 1.19-5.22; P = .02) independent of age, pulmonary hypertension, LVEF, right ventricular ejection fraction, and LVLGE extent. Risk of arrhythmic events was greater with an increasing number of pathology-frequent LGE features. The absence of the pathology-frequent LGE phenotype was associated with a low risk of arrhythmic events, even in the presence of LGE or abnormal LVEF.
Conclusions and Relevance
This cohort study found that a CMR phenotype involving pathology-frequent LGE features was associated with a high risk of arrhythmic and heart failure events in patients with sarcoidosis. The findings indicate that CMR phenotypes could be used to optimize clinical decision-making for treatment options, such as implantable cardioverter-defibrillators.
Introduction
Cardiac sarcoidosis (CS) is a high-risk manifestation of sarcoidosis associated with a poor prognosis, as demonstrated by an impaired quality of life and a high risk of death.1,2 Cardiovascular magnetic resonance imaging (CMR) is recommended and frequently used for the diagnosis and prognostication of patients with suspected CS.3,4
In studies of suspected CS evaluated by CMR, cardiac damage identified as late gadolinium enhancement (LGE) has been associated with ventricular arrhythmic events.5,6,7,8 However, only a minority of patients with LGE experience these events. There is a critical need to identify this subset of patients because of the clinical management implications, particularly for decisions involving implantable cardioverter-defibrillators (ICDs). LGE in these patients has been described in a variety of patterns, including subendocardial,9,10,11 patchy intramural,10 midmyocardial,11,12,13,14 and isolated lateral wall involvement.15 Whether the clinical outcomes vary by the specific CMR phenotype defined by the left ventricular ejection fraction (LVEF) and the LGE pattern is unknown.
We previously demonstrated16 that certain features of myocardial damage were frequent on gross pathology evaluation of patients with histologically diagnosed CS who had autopsy or cardiac transplant, while other features were rare or never seen. The features seen in more than 90% of cases were left ventricular (LV) subepicardial, LV multifocal, septal, and right ventricular (RV) free-wall involvement, while infrequent features included the absence of gross myocardial involvement, isolated LV midmyocardial involvement, isolated LV subendocardial involvement, and the absence of septal involvement.16 These findings led us to hypothesize that outcomes among patients with suspected CS depend on CMR phenotype. Specifically, we hypothesized that cardiac damage with the features seen frequently on the pathology of patients with CS who died or underwent cardiac transplant (referred to as pathology-frequent LGE) would be associated with a higher risk of CS-related adverse outcomes compared with features that were rare or never seen on pathology (referred to as pathology-rare LGE). In this study, we tested our hypothesis in a large cohort of patients clinically evaluated for CS with CMR.
Methods
Study Cohort
We studied consecutive patients at the University of Minnesota, Minneapolis, with histologically proven sarcoidosis who underwent CMR with LGE imaging for the evaluation of suspected CS between 2004 and 2020. Cardiac involvement was suspected based on symptoms (such as palpitations, presyncope/syncope, dyspnea unexplained by sarcoidosis-associated pulmonary disease or pulmonary hypertension, and chest pain) or electrocardiographic abnormalities (such as left or right bundle branch block, pathological Q waves, atrioventricular block, and ventricular tachycardia) suggestive of CS.3 Demographic data, medical history, comorbidities, medications, and outcome data were collected blinded to CMR data as previously described.7,17,18 Race and ethnicity data were collected by self-report according to the categories used in the US Census. Race and ethnicity were reported because sarcoidosis is more prevalent in Black individuals, making race and ethnicity potential variables of interest. This retrospective cohort study was approved by institutional review board at the University of Minnesota with a waiver of informed consent owing to the retrospective nature of the study.
CMR Acquisition
CMR was performed on clinical 1.5-T Siemens scanners (Avanto or Aera) using phased-array receiver coils according to standard recommendations.19,20 A typical protocol was as follows. First, localizers were acquired to identify the cardiac position. Next, cine CMR images were acquired in multiple short-axis views (every 10 mm to cover the entire LV from the mitral valve plane through the apex) and 3 long-axis views (2-, 3-, and 4-chamber) using a steady-state free-precession sequence. Standard LGE imaging was performed 10 to 15 minutes after administration of gadolinium contrast (0.15 mmol/kg) using a 2-dimensional segmented inversion-recovery gradient-echo sequence in identical views as cine CMR imaging. Typical inversion delay times were 280 to 360 milliseconds.
CMR Analyses
CMR analyses were performed in a core laboratory fashion blinded to all other patient information by the consensus of 2 investigators (F.K. and C.S.) with expertise in CMR as previously described.7,17,18 LVEF and RVEF were determined by quantitative analysis according to standard recommendations.21 Abnormal LVEF was defined as less than 51% in men and less than 52% in women and abnormal RVEF was defined as less than 42% in men and less than 46% in women, based on published normal ranges.22
LGE was identified visually. LGE features were identified using similar criteria as Okasha et al16:
Location within LV wall: subepicardial (involving the outer portion, including the RV aspect of the interventricular septum), midmyocardial (involving the middle portion), or subendocardial (involving the inner portion).
Focality within the LV: unifocal (1 lesion) or multifocal (more than 1 discrete lesion).
LV walls involved: septal or nonseptal (anterior, inferior, and/or lateral walls).
Involvement of the RV free wall: yes or no.
Transmural LGE was determined to be subendocardial or subepicardial in origin based on the edges of the LGE and classified accordingly. Midmyocardial LGE at the anterior and inferior RV insertion points with or without extension to the interventricular septum was counted as a single focus of involvement.
Features of myocardial damage seen frequently on gross pathology evaluation of patients with histologically diagnosed CS who underwent autopsy or cardiac transplant (ie, LV subepicardial, LV multifocal, septal, and/or RV free wall involvement16) were classified as pathology-frequent LGE. All other LGE features were classified as pathology-rare LGE.
The extent of left ventricular late gadolinium enhancement (LVLGE) was semiquantitatively assessed on a 17-segmental basis from the area of hyperenhanced myocardium on LGE images on a 5-point scale as previously described23: 0, no hyperenhancement; 1, 1% to 25% hyperenhancement; 2, 26% to 50% hyperenhancement; 3, 51% to 75% hyperenhancement; and 4, 76% to 100% hyperenhancement. The global LGE extent was assessed as a percentage of the LV myocardium by adding the segmental scores weighted by the midpoint of the range of hyperenhancement and dividing by 17 for the total number of segments.23,24 This method of LGE quantification has been shown to have a high correlation with24 and similar reproducibility as automated quantification methods.25 While the presence of RV free-wall LGE was assessed, the extent of RVLGE was not quantified, since quantification of the RV mass has been noted to have high variability due to its thin nature.17,26
Variability in the Identification of LGE Features and LVLGE Extent
The interobserver variability in identifying LGE features and LVLGE extent was tested in 25 patients with LGE by comparing the expert consensus reads with that of a third independent expert investigator (P.H.B.). To analyze intraobserver variability, the 2 expert investigators repeated the blinded LGE interpretations in 25 patients with LGE at an interval of 3 months after the original interpretations.
Clinical Follow-up and End Points
Follow-up data were assembled through a review of electronic medical records. The study end points were a ventricular arrhythmic composite end point and a heart failure composite end point. The arrhythmic end point included sudden (arrhythmic) cardiac death, resuscitated cardiac arrest with documented ventricular tachycardia or ventricular fibrillation (VT/VF), and serious ventricular arrhythmia, defined as sustained VT (duration more than 30 seconds), or appropriate ICD therapy (shock or antitachycardia pacing). The appropriateness of ICD therapies was adjudicated by cardiac electrophysiologists during the patients’ clinical care using intracardiac electrograms recorded by the ICD and based on tachycardia rate, onset, stability, atrioventricular association, and QRS morphology. Follow-up for the arrhythmic end point was censored for orthotopic heart transplant, death, or the end of follow-up. The heart failure end point included cardiac death due to heart pump failure, ventricular assist device placement, orthotopic heart transplant, or heart failure hospitalization. Follow-up for the heart failure end point was censored for death or the end of follow-up. Median follow-up for the arrhythmic end point was defined as the time until the end point was reached, orthotopic heart transplant, death, or the end of follow-up. Median follow-up for the heart failure end point was defined as the time until the end point was reached, death, or the end of follow-up. Follow-up was limited to 10 years after the CMR. Mortality status and death dates were also independently cross-verified with the Minnesota State Department of Health’s Office of Vital Records. For patients who died outside the hospital, death certificates were reviewed to determine the cause of death. The cause of death was classified using standard criteria.27 Data on events were obtained and classified without knowledge of the CMR findings.
Statistical Analyses
Categorical variables were expressed as counts with percentages. Normally distributed continuous variables were expressed as means with SDs, and nonnormally distributed continuous variables were presented as medians with IQRs. Cohen κ was used to assess interobserver and intraobserver agreements in identifying LGE features, and intraclass correlation coefficient was used for LGE extent. The correlation between LVEF and LVLGE extent was tested using the Pearson correlation coefficient test. The cumulative incidence of the study end points was estimated using the Kaplan-Meier method and compared using the log-rank test. Univariable and multivariable analyses of event-free survival were performed using Cox proportional hazards regression. Covariates for inclusion in the multivariable models were chosen a priori based on prior published literature and our clinical experience with CS. The frequencies of the pathology-based LGE features and LVLGE extent as dichotomous variables were compared using the McNemar test. For comparisons involving LVLGE extent as a dichotomous variable, a threshold of 5.7% was used based on our prior study.7 Statistical significance was defined as a 2-tailed P value <.05. Analyses were done in RStudio 2021.09.0 Build 351 (R Foundation).
Results
Patient Characteristics
This study included 504 patients with histologically proven sarcoidosis and suspected cardiac involvement, 242 of whom were female (48.0%) and 262 of whom were male (52.0%). By patient self-report, 2 individuals (0.4%) were American Indian or Alaska Native, 6 (1.2%) were Asian, 90 (17.9%) were Black or African American, 399 (79.2%) were White, 5 (1.0%) were of 2 or more races (including the above-mentioned categories and Native Hawaiian or Other Pacific Islander), and 2 (0.4%) were of unknown race; 4 (0.8%) were Hispanic or Latino, 498 (98.8%) were not Hispanic or Latino, and 2 (0.4%) were of unknown ethnicity. The mean (SD) age of the cohort was 54.1 (12.5) years (Table 1).
Table 1. Characteristics for All Patients and Patients Stratified by Cardiovascular Magnetic Resonance (CMR) Phenotype.
| No. (%) | |||||
|---|---|---|---|---|---|
| All patients | LVEF and no LGE | Pathology-LGE | |||
| Normal | Abnormal | Frequent | Rare | ||
| No. | 504 | 290 | 53 | 103 | 58 |
| Demographic characteristics | |||||
| Age, mean (SD), y | 54.1 (12.5) | 51.7 (12.6) | 55.3 (13.3) | 58.7 (10.2) | 56.7 (12.2) |
| Female | 242 (48.0) | 166 (57.2) | 25 (47.2) | 29 (28.2) | 22 (37.9) |
| Male | 262 (52.0) | 124 (42.8) | 28 (52.8) | 74 (71.8) | 36 (62.1) |
| Body mass index, mean (SD)a | 31.2 (7.2) | 31.6 (7.4) | 32.3 (7.0) | 30.1 (6.7) | 30.4 (7.2) |
| Raceb | |||||
| American Indian or Alaska Native | 2 (0.4) | 0 | 1 (1.9) | 0 | 1 (1.7) |
| Asian | 6 (1.2) | 3 (1.0) | 0 | 2 (1.9) | 1 (1.7) |
| Black | 90 (17.9) | 52 (17.9) | 11 (20.8) | 15 (14.6) | 12 (20.7) |
| White | 399 (79.2) | 231 (79.7) | 41 (77.4) | 84 (81.6) | 43 (74.1) |
| ≥2 Racesc | 5 (1.0) | 3 (1.0) | 0 | 1 (1.0) | 1 (1.7) |
| Unknown | 2 (0.4) | 1 (0.3) | 0 | 1 (1.0) | 0 |
| Ethnicityb | |||||
| Hispanic or Latino | 4 (0.8) | 3 (1.0) | 1 (1.9) | 0 | 0 |
| Not Hispanic or Latino | 498 (98.8) | 286 (98.6) | 52 (98.1) | 102 (99.0) | 58 (100.0) |
| Unknown | 2 (0.4) | 1 (0.3) | 0 | 1 (1.0) | 0 |
| Comorbidities | |||||
| Hypertension | 274 (54.4) | 143 (49.3) | 34 (64.2) | 64 (62.1) | 33 (56.9) |
| Diabetes | 114 (22.6) | 57 (19.7) | 14 (26.4) | 27 (26.2) | 16 (27.6) |
| Dyslipidemia | 248 (49.2) | 126 (43.4) | 32 (60.4) | 60 (58.3) | 30 (51.7) |
| Tobacco use | |||||
| Current | 44 (8.7) | 26 (9.0) | 6 (11.3) | 10 (9.7) | 2 (3.4) |
| Former | 181 (35.9) | 107 (36.9) | 20 (37.7) | 27 (26.2) | 27 (46.6) |
| Pulmonary hypertension | 87 (17.3) | 28 (9.7) | 11 (20.8) | 21 (20.4) | 27 (46.6) |
| Heart failure | 66 (13.1) | 9 (3.1) | 16 (30.2) | 29 (28.2) | 12 (20.7) |
| Extracardiac sarcoidosis involvement | |||||
| Lung | 469 (93.1) | 277 (95.5) | 52 (98.1) | 85 (82.5) | 55 (94.8) |
| Skin | 57 (11.3) | 33 (11.4) | 7 (13.2) | 9 (8.7) | 8 (13.8) |
| Eyes | 42 (8.3) | 31 (10.7) | 1 (1.9) | 9 (8.7) | 1 (1.7) |
| Liver | 34 (6.7) | 21 (7.2) | 2 (3.8) | 8 (7.8) | 3 (5.2) |
| Central nervous system | 28 (5.6) | 19 (6.6) | 4 (7.5) | 3 (2.9) | 2 (3.4) |
| Clinical symptoms | |||||
| Chest pain | 162 (32.1) | 106 (36.6) | 13 (24.5) | 29 (28.2) | 14 (24.1) |
| Palpitations | 154 (30.6) | 101 (34.8) | 11 (20.8) | 33 (32.0) | 9 (15.5) |
| Dyspnea | 251 (49.8) | 125 (43.1) | 32 (60.4) | 56 (54.4) | 38 (65.5) |
| Presyncope | 89 (17.7) | 53 (18.3) | 11 (20.8) | 22 (21.4) | 3 (5.2) |
| Syncope | 34 (6.7) | 14 (4.8) | 2 (3.8) | 15 (14.6) | 3 (5.2) |
| Arrhythmia | |||||
| Ventricular arrhythmia | |||||
| PVCs | 143 (28.4) | 57 (19.7) | 21 (39.6) | 45 (43.7) | 20 (34.5) |
| Nonsustained VT | 52 (10.3) | 10 (3.4) | 4 (7.5) | 36 (35.0) | 2 (3.4) |
| Sustained VT | 17 (3.4) | 2 (0.7) | 1 (1.9) | 13 (12.6) | 1 (1.7) |
| VF/cardiac arrest | 9 (1.8) | 1 (0.3) | 0 (0.0) | 8 (7.8) | 0 (0.0) |
| Supraventricular tachycardia | 67 (13.3) | 32 (11.0) | 6 (11.3) | 24 (23.3) | 5 (8.6) |
| Atrial fibrillation/flutter | 101 (20.0) | 35 (12.1) | 22 (41.5) | 31 (30.1) | 13 (22.4) |
| AV block ≥2nd degree | 25 (5.0) | 4 (1.4) | 3 (5.7) | 16 (15.5) | 2 (3.4) |
| Medications | |||||
| Aspirin | 176 (35.2) | 82 (28.4) | 16 (30.2) | 54 (53.5) | 24 (42.1) |
| Statins | 179 (35.6) | 77 (26.6) | 24 (45.3) | 52 (50.5) | 26 (44.8) |
| ACEI/ARB | 163 (32.4) | 67 (23.2) | 23 (43.4) | 55 (53.4) | 18 (31.0) |
| β-Blockers | 168 (33.4) | 64 (22.1) | 20 (37.7) | 61 (59.2) | 23 (39.7) |
| Steroids | 164 (32.6) | 96 (33.2) | 16 (30.2) | 26 (25.2) | 26 (44.8) |
| Nonsteroid immunomodulators | 80 (15.9) | 48 (16.6) | 11 (20.8) | 7 (6.8) | 14 (24.1) |
| CMR findings | |||||
| LVEDVI, median (IQR), mL/m2 | 68.5 (57.0-82.0) | 66.0 (56.0-75.8) | 75.0 (65.0-96.0) | 83.0 (66.0-101.5) | 63.0 (51.3-76.3) |
| LVESVI, median (IQR), mL/m2 | 29.0 (23.0-38.0) | 26.5 (21.0-31.0) | 40.0 (34.0-58.0) | 42.0 (31.0-66.0) | 26.5 (20.3-34.8) |
| Abnormal LVEF | 140 (27.8) | - | 53 (100.0) | 68 (66.0) | 19 (32.8) |
| LVEF, median (IQR), % | 57.0 (50.0-62.0) | 60.0 (57.0-64.0) | 44.0 (39.0-48.0) | 46.0 (32.0-54.0) | 56.0 (47.0-62.0) |
| RVEDVI, median (IQR), mL/m2 | 65.0 (53.0-76.0) | 63.0 (52.0-73.0) | 62.0 (48.0-77.0) | 70.0 (56.0-82.0) | 67.0 (59.0-88.3) |
| RVESVI, median (IQR), mL/m2 | 27.0 (21.0-35.0) | 25.0 (19.0-29.0) | 30.0 (23.0-37.0) | 31.0 (23.5-42.5) | 38.5 (26.0-53.5) |
| Abnormal RVEF | 76 (15.1) | 6 (2.1) | 13 (24.5) | 31 (30.1) | 26 (44.8) |
| RVEF, median (IQR), % | 58.0 (50.0-64.0) | 61.0 (56.0-66.0) | 51.0 (45.0-58.0) | 52.0 (39.5-59.5) | 45.0 (32.0-58.0) |
| Any LGE | 161 (31.9) | NA | NA | 103 (10.0) | 58 (10.0) |
| LVLGE extent, median (IQR), % | 0.0 (0.0-2.0) | NA | NA | 11.0 (6.0-22.5) | 2.0 (2.0-3.0) |
Abbreviations: ACEI, angiotensin converting enzymeinhibitor; ARB, angiotensin receptor blocker; AV, atrioventricular; LGE, late gadolinium enhancement; LVEDVI, left ventricular end-diastolic volume indexed; LVEF, left ventricular ejection fraction; LVESVI, left ventricular end-systolic volume indexed; LVLGE, left ventricular late gadolinium enhancement; NA, not applicable; PVCs, premature ventricular complexes; RVEDVI, right ventricular end-diastolic volume indexed; RVEF, right ventricular ejection fraction; RVESVI, right ventricular end-systolic volume indexed; VF, ventricular fibrillation; VT, ventricular tachycardia.
Calculated as weight in kilograms divided by height in meters squared.
Race and ethnicity data were collected via self-report using the multiple-choice categories used in the US Census. Race and ethnicity were reported because sarcoidosis is more prevalent in Black individuals, making race and ethnicity potential variables of interest.
May also include Native Hawaiian or Other Pacific Islander.
Based on the LVEF and LGE presence and features, 4 distinct phenotypes were identified (Figure 1): normal LVEF and no LGE (n = 290; 57.5%), abnormal LVEF and no LGE (n = 53; 10.5%), pathology-frequent LGE (n = 103; 20.4%), and pathology-rare LGE (n = 58; 11.5%). Characteristics of patients with each of the 4 phenotypes are listed in Table 1. Examples of patients with each of the 4 phenotypes are provided in Figure 2.
Figure 1. Identification of Cardiovascular Magnetic Resonance Phenotypes Based on Left Ventricular Ejection Fraction (LVEF), Late Gadolinium Enhancement (LGE) Presence, and LGE Features.
Figure 2. Representative Examples of Patients With Each of the 4 Distinct Cardiovascular Magnetic Resonance (CMR) Phenotypes.
ATP indicates antitachycardia pacing; ICD, implantable cardioverter-defibrillator; LGE, late gadolinium enhancement; LVEF, left ventricular ejection fraction; RVEF, right ventricular ejection fraction; VT, ventricular tachycardia.
Among the 53 patients with abnormal LVEF and no LGE, the median (IQR) LVEF was 44% (39-48) with a range of 17% to 51%. These patients accounted for 37.9% of patients with abnormal LVEF. Ten patients had LVEF less than 35%.
Among the 103 patients with pathology-frequent LGE, 60 (58.3%) had subepicardial LGE, 88 (85.4%) had multifocal LGE, 88 (85.4%) had septal LGE, and 30 (29.1%) had RV free-wall LGE. Nineteen patients (18.4%) had 1, 33 (32.0%) had 2, 23 (22.3%) had 3, and 28 (27.2%) had all 4 pathology-frequent LGE features, with a median (IQR) number of features of 2 (2-4). Thirty-five patients (34.0%) had a normal LVEF.
Among patients with pathology-frequent LGE, there was a negative correlation between LVEF and LVLGE extent, with a Pearson correlation coefficient r of −0.58 (95% CI, −0.69 to −0.43; P < .001).
Among the 58 patients with pathology-rare LGE, 42 (72.4%) had only insertion-site LGE, 4 (6.9%) had midmyocardial LGE without subepicardial LGE, and 12 (20.7%) had subendocardial LGE without subepicardial LGE. Thirty-nine patients (67.2%) had a normal LVEF.
Among patients with pathology-rare LGE, there was a positive correlation between LVEF and LGE extent, with a Pearson correlation coefficient r of 0.05 (95% CI, −0.21 to 0.30; P = .72).
Reproducibility of LGE Assessment
Reproducibility analysis of visual scoring demonstrated excellent agreement for LGE presence (κ statistic of 0.93 for interobserver and 0.98 for intraobserver agreement) and LVLGE extent (intraclass correlation coefficient of 0.91 for interobserver and 0.95 for intraobserver agreement).
Outcomes Stratified by CMR Phenotypes
Arrhythmic End Point
At a median (IQR) follow-up of 4.3 (2.4-6.9) years, 30 patients reached the arrhythmic end point. Of these, 22 had appropriate ICD shocks, 22 had appropriate antitachycardia pacing, 3 had resuscitated arrhythmic sudden cardiac death (SCD), and 1 had arrhythmic SCD. The incidence rates of the arrhythmic end point by the CMR phenotypes are provided in eTable 1 in the Supplement.
On Kaplan-Meier analyses, the estimated cumulative incidence of the arrhythmic end point was 28.2% (95% CI, 18.0-37.1) for pathology-frequent LGE, 0.0% for pathology-rare LGE, 1.9% (95% CI, 0.0-5.5) for no LGE and abnormal LVEF, and 1.9% (95% CI, 0.0-4.4) for no LGE and normal LVEF. Patients with pathology-frequent LGE had a significantly higher cumulative incidence of the arrhythmic end point compared with the 3 other phenotypes and there were no differences in the cumulative incidence of the arrhythmic end point between the other 3 phenotypes (Figure 3A).
Figure 3. Kaplan-Meier Cumulative Incidence Curves for the Arrhythmic and Heart Failure End Points for Patients Stratified by Cardiovascular Magnetic Resonance (CMR) Phenotype.
LGE indicates late gadolinium enhancement; LVEF, left ventricular ejection fraction.
Each of the 4 pathology-frequent LGE features was associated with a significantly higher cumulative incidence of the arrhythmic end point (eFigure 1 in the Supplement). The cumulative incidence of the arrhythmic end point increased with the number of pathology-frequent LGE features, demonstrating a dose-response association (eFigure 2 in the Supplement).
On Cox univariable analyses (Table 2) with CMR phenotype studied as a nominal variable, the phenotype of pathology-frequent LGE was associated with a significantly higher rate of the arrhythmic end point (hazard ratio [HR], 30.45 relative to the phenotype with normal LVEF and no LGE; 95% CI, 9.19-100.91; P < .001) while the other phenotypes were not. On Cox multivariable analyses (Table 2; eFigure 5 in the Supplement), pathology-frequent LGE was associated with the arrhythmic end point (HR, 12.12; 95% CI, 3.62-40.57; P < .001) independent of LVEF and LVLGE extent.
Table 2. Cox Proportional Hazards Regression Analyses for the Arrhythmic and Heart Failure End Points.
| Variable | Arrhythmic end point | Heart failure end point | ||||||
|---|---|---|---|---|---|---|---|---|
| Univariable analyses | Multivariable analyses | Univariable analyses | Multivariable analyses | |||||
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| Age, per 1-y increase | 0.99 (0.96-1.02) | .46 | NA | NA | 1.04 (1.02-1.06) | <.001 | 1.03 (1.00-1.05) | .02 |
| Male | 3.19 (1.37-7.43) | .007 | NA | NA | 1.44 (0.86-2.41) | .17 | NA | NA |
| Pulmonary hypertension | 0.95 (0.36-2.47) | .91 | NA | NA | 4.16 (2.50-6.92) | <.001 | 2.58 (1.49-4.46) | <.001 |
| Abnormal LVEF | 6.76 (3.09-14.76) | <.001 | NA- | NA | 4.77 (2.84-8.00) | <.001 | NA | NA |
| LVEF, per 1% decrease | 1.08 (1.05-1.10) | <.001 | 1.01 (0.98-1.05) | .39 | 1.07 (1.05-1.08) | <.001 | 1.04 (1.01-1.06) | .01 |
| Abnormal RVEF | 5.4 (2.63-11.07) | <.001 | NA | NA | 4.22 (2.51-7.11) | <.001 | NA | NA |
| RVEF, per 1% decrease | 1.07 (1.04-1.09) | <.001 | NA | NA | 1.06 (1.04-1.08) | <.001 | 1.02 (0.99-1.04) | .21 |
| Any LGE | 16.07 (5.60-46.11) | <.001 | NA | NA | 3.71 (2.22-6.21) | <.001 | NA | NA |
| LVLGE extent, per 1% increase | 1.08 (1.07-1.10) | <.001 | 1.04 (1.01-1.07) | .003 | 1.05 (1.03-1.06) | <.001 | 0.99 (0.96-1.02) | .54 |
| CMR phenotype | ||||||||
| LVEF and no LGE | ||||||||
| Normal | 1 [Reference] | 1 [Reference] | NA | NA | 1 [Reference] | 1 [Reference] | NA | NA |
| Abnormal | 1.86 (0.19-17.88) | .59 | NA | NA | 3.42 (1.51-7.75) | .003 | NA | NA |
| Pathology | ||||||||
| Frequent LGE | 30.45 (9.19-100.91) | <.001 | NA | NA | 6.33 (3.40-11.79) | <.001 | NA | NA |
| Rare LGE | 0 | NA | NA | NA | 2.92 (1.25-6.83) | .013 | NA | NA |
| Frequent, LGE compared with no pathology-frequent LGE | 31.25 (10.87-89.78) | <.001 | 12.12 (3.62-40.57) | <.001 | 4.08 (2.45-6.81) | <.001 | 2.49 (1.19-5.22) | .02 |
Abbreviations: CMR, cardiovascular magnetic resonance imaging; HR, hazard ratio; LGE, late gadolinium enhancement; LVEF, left ventricular ejection fraction; LVLGE, left ventricular late gadolinium enhancement; NA, not applicable; RVEF, right ventricular ejection fraction.
Heart Failure End Point
At a median (IQR) follow-up of 4.2 (2.3-6.9) years, 60 patients reached the heart failure end point. Of these, 5 had cardiac death due to heart pump failure, 3 had ventricular assist device placement, 6 had orthotopic heart transplant, and 59 had heart failure hospitalization. The incidence rates of the heart failure end point by the CMR phenotypes are provided in eTable 1 in the Supplement.
On Kaplan-Meier analyses, the estimated cumulative incidence of the heart failure end point was 37.7% (95% CI, 22.5-49.9) for pathology-frequent LGE, 30.7% (95% CI, 5.6-49.2) for pathology-rare LGE, 23.8% (95% CI, 6.2-38.1) for no LGE and abnormal LVEF, and 7.4% (95% CI, 3.6-11.1) for no LGE and normal LVEF. Patients with pathology-frequent LGE had a significantly higher cumulative incidence of the heart failure end point compared with patients with normal LVEF and no LGE but not compared with patients with abnormal LVEF and no LGE or pathology-rare LGE. Other than patients with pathology-frequent LGE, those with abnormal LVEF but no LGE also had a higher cumulative incidence of the heart failure end point compared with patients with normal LVEF and no LGE (Figure 3B).
Each of the 4 pathology-frequent LGE features was associated with a significantly higher cumulative incidence of the heart failure end point (eFigure 3 in the Supplement). The cumulative incidence of the heart failure end point did not change significantly with the number of pathology-frequent LGE features (eFigure 4 in the Supplement).
On Cox univariable analyses (Table 2) with the CMR phenotype studied as a nominal variable, relative to the phenotype with normal LVEF and no LGE, the other 3 phenotypes were associated with a higher cumulative incidence of the heart failure end point with the highest hazard ratio for the pathology-frequent LGE phenotype (HR, 6.33; 95% CI, 3.40-11.79; P < .001). On Cox multivariable analyses (Table 2; eFigure 5 in the Supplement), pathology-frequent LGE was associated with the heart failure end point (HR, 2.49; 95% CI, 1.19-5.22; P = .02) independent of age, pulmonary hypertension, LVEF, RVEF, and LVLGE extent.
Comparison Between Pathology-Based LGE Features and LVLGE Extent
On a group level, patients with pathology-frequent LGE had a greater LVLGE extent compared with patients with pathology-rare LGE (median extent, 11.0% vs 2.0%; P < .001). When the LVLGE extent was dichotomized using the 5.7% threshold, the 2 methods of stratifying patients with LGE were discordant in 26 of 161 patients (16.1%) with LGE. Of these, 22 had pathology-frequent LGE with LGE extent less than 5.7%, while 4 had pathology-rare LGE with LGE extent greater than 5.7% (eTable 2 in the Supplement). A comparison of the 2 methods of stratifying patients by the McNemar test showed a significant difference. Both pathology-frequent LGE and LVLGE extent were independently associated with the arrhythmic end point, while only pathology-frequent LGE was independently associated with the heart failure end point and LVLGE extent was not.
Discussion
Key Findings
In a large cohort of patients with histology-proven sarcoidosis and suspected cardiac involvement, a CMR phenotype incorporating the features of cardiac damage frequently seen on gross pathology images of patients who died of CS or underwent cardiac transplant for CS identified patients at high risk of arrhythmic and heart failure events. The association between the pathology-frequent LGE phenotype and arrhythmic events was greater with an increasing number of pathology-frequent LGE features. Conversely, the absence of the pathology-frequent LGE phenotype was associated with a low risk of arrhythmic events, even in the presence of LGE or abnormal LVEF.
Strengths
Our study has several strengths. First, we studied a large number of patients. To our knowledge, ours is the largest study to date of consecutive patients with suspected CS evaluated by CMR. Second, unlike in several studies of suspected CS, all patients in our study had histology-proven sarcoidosis. Third, we studied hard clinical outcomes rather than surrogate outcomes. Fourth, we studied both arrhythmic and heart failure outcomes.
Prior Literature on the Risk Associated With LGE in Suspected CS
LGE has been associated with incident ventricular arrhythmia, cardiovascular death, and all-cause death.5,6 However, only a minority of patients with LGE experience events. In a meta-analysis by Hulten et al,5 21% (41 of 199) and 10% (10 of 100) of patients with LGE experienced ventricular arrhythmia and cardiovascular death, respectively. In a recent study28 of 80 patients with CS, 73 of whom (91.3%) had LGE and all of whom were at low risk of SCD as deemed by a multidisciplinary team, none had arrhythmic SCD or sustained ventricular arrhythmia on an implantable loop recorder over a mean follow-up of 33 months. Thus there is a critical need to risk-stratify patients with suspected CS beyond simply identifying the presence or absence of LGE.
In a previous study7 of 290 patients, we established that LVLGE extent greater than 5.7% in patients with LVEF greater than 35% was as sensitive as and significantly more specific than the presence of any LGE for the identification of patients at risk for the composite end point. In this study, we demonstrated that pathology-frequent LGE features provided additional risk stratification beyond the LVLGE extent.
Previous studies have consistently demonstrated that the absence of LGE is associated with very low rates of arrhythmic events.5 The incidence rate of the arrhythmic end point in patients without LGE in our cohort was 0.22 per 100 patient-years, reaffirming the negative predictive value of LGE for arrhythmic events.
Clinical Implications
A key component of the clinical treatment of patients with CS is risk-stratification for SCD and consideration of ICD implantation.7 ICDs entail risks, including infection, inappropriate shocks, the potential for proarrhythmia, device malfunction, and procedural complications, all of which can adversely affect quality of life.29,30
The 2017 AHA/ACC/HRS Guideline for Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death31 recommends ICD implantation in patients with CS and LVEF 35% or less (class of recommendation I). It also states that it is reasonable to implant an ICD in patients with CS and LVEF greater than 35% who have extensive LGE on CMR (class of recommendation IIa).31 However, extensive LGE is not defined.
Pathology-frequent LGE provides a novel method, alternative to and independent of LGE extent greater than 5.7%, for identifying patients with LGE who have a higher risk and would thus benefit from ICD implantation. Conversely, the absence of this CMR phenotype is associated with a very low risk of ventricular arrhythmias and SCD, raising the possibility that ICD implantation could be avoided in these patients, even in those with an LVEF of less than 35%. Identification of pathology-frequent LGE is quick and does not require any scoring, calculation, or analytic software; thus, it could be applied easily in clinical practice.
Pathology-frequent LGE features also identify patients with suspected CS who are at higher risk of heart failure events. However, unlike with arrhythmic outcomes, the absence of pathology-frequent LGE features is not associated with a very low risk of heart failure events. This is because, in patients with sarcoidosis, pulmonary hypertension-related right heart failure is an important cause of heart failure events independent of CS4; this was corroborated by the association we found between pulmonary hypertension and the heart failure end point.
The absence of a negative correlation between LVEF and LVLGE extent in patients with pathology-rare LGE suggests that the LGE may be related to myocardial pathologies other than CS. We also identified a CMR phenotype that has not previously been described in patients with suspected CS, of abnormal LVEF without LGE. This phenotype accounted for 10.5% of our overall cohort and 37.9% of patients with an abnormal LVEF. Patients with this CMR phenotype had a low incidence of ventricular arrhythmias during follow-up but, not surprisingly, experienced heart failure events. Whether these patients have CS or unrelated nonischemic cardiomyopathies remains to be prospectively investigated.
Limitations
This study has limitations. As a retrospective study involving a single-center cohort limited to patients with histologically-proven sarcoidosis who underwent CMR with LGE imaging for clinically suspected CS, the generalizability of our findings may be limited. We also recognize the modest number of events in the study due to our emphasis on including only hard clinical events. Most of the cohort (79%) was White; thus, our findings need to be replicated in a larger, more racially diverse, multicenter cohort. Appropriate ICD therapy is an imperfect surrogate for SCD and may overestimate the risk; some ventricular arrhythmias that lead to ICD therapies may have terminated spontaneously without resulting in death.32 Cardiac monitoring was not universally used; thus, self-limiting ventricular arrhythmias could have been underrecognized. For deaths that occurred outside the hospital setting (26%), we determined the cause of death using death certificates.
Conclusions
In this study of a large cohort of patients with histologically proven sarcoidosis and suspected cardiac involvement, we identified distinct CMR phenotypes. Patients with LGE in cardiac damage patterns seen frequently on gross pathology had a high risk of ventricular arrhythmia and heart failure events, independent of LVEF and LVLGE extent. Conversely, the absence of pathology-frequent LGE was associated with a low risk of arrhythmic events, even in the presence of LGE or abnormal LVEF. If replicated in other cohorts, these CMR phenotypes could help optimize treatment decisions, including ICD implantation. After replication, subsequent studies should prospectively test whether clinical management guided by CMR phenotypes improves outcomes in patients with suspected CS.
eTable 1. Incidence rates of the arrhythmic and heart failure endpoints by CMR phenotypes
eTable 2. Comparison between LGE prevalence by pathology-based features and extent
eFigure 1. Kaplan-Meier cumulative incidence curves for the arrhythmic endpoint stratified by individual pathology-frequent LGE features
eFigure 2. Kaplan-Meier cumulative incidence curves for the arrhythmic endpoint stratified by the number of pathology-frequent LGE features
eFigure 3. Kaplan-Meier cumulative incidence curves for the heart failure endpoint stratified by individual pathology-frequent LGE features
eFigure 4. Kaplan-Meier cumulative incidence curves for the heart failure endpoint stratified by the number of pathology-frequent LGE features
eFigure 5. Forest Plots of the multivariable analyses for the arrhythmic and heart failure endpoints
References
- 1.Sauer WH, Stern BJ, Baughman RP, Culver DA, Royal W. High-risk sarcoidosis. current concepts and research imperatives. Ann Am Thorac Soc. 2017;14(suppl 6):S437-S444. doi: 10.1513/AnnalsATS.201707-566OT [DOI] [PubMed] [Google Scholar]
- 2.Perlman DM, Sudheendra MT, Furuya Y, et al. Clinical presentation and treatment of high-risk sarcoidosis. Ann Am Thorac Soc. 2021;18(12):1935-1947. doi: 10.1513/AnnalsATS.202102-212CME [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Birnie DH, Sauer WH, Bogun F, et al. HRS expert consensus statement on the diagnosis and management of arrhythmias associated with cardiac sarcoidosis. Heart Rhythm. 2014;11(7):1305-1323. doi: 10.1016/j.hrthm.2014.03.043 [DOI] [PubMed] [Google Scholar]
- 4.Crouser ED, Maier LA, Wilson KC, et al. Diagnosis and detection of sarcoidosis. an official american thoracic society clinical practice guideline. Am J Respir Crit Care Med. 2020;201(8):e26-e51. doi: 10.1164/rccm.202002-0251ST [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hulten E, Agarwal V, Cahill M, et al. Presence of late gadolinium enhancement by cardiac magnetic resonance among patients with suspected cardiac sarcoidosis is associated with adverse cardiovascular prognosis: a systematic review and meta-analysis. Circ Cardiovasc Imaging. 2016;9(9):e005001. doi: 10.1161/CIRCIMAGING.116.005001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Coleman GC, Shaw PW, Balfour PC Jr, et al. Prognostic value of myocardial scarring on CMR in patients with cardiac sarcoidosis. JACC Cardiovasc Imaging. 2017;10(4):411-420. doi: 10.1016/j.jcmg.2016.05.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kazmirczak F, Chen KA, Adabag S, et al. Assessment of the 2017 AHA/ACC/HRS guideline recommendations for implantable cardioverter-defibrillator implantation in cardiac sarcoidosis. Circ Arrhythm Electrophysiol. 2019;12(9):e007488. doi: 10.1161/CIRCEP.119.007488 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Franke KB, Marshall H, Kennewell P, et al. Risk and predictors of sudden death in cardiac sarcoidosis: a systematic review and meta-analysis. Int J Cardiol. 2021;328:130-140. doi: 10.1016/j.ijcard.2020.11.044 [DOI] [PubMed] [Google Scholar]
- 9.Patel AR, Klein MR, Chandra S, et al. Myocardial damage in patients with sarcoidosis and preserved left ventricular systolic function: an observational study. Eur J Heart Fail. 2011;13(11):1231-1237. doi: 10.1093/eurjhf/hfr099 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Greulich S, Deluigi CC, Gloekler S, et al. CMR imaging predicts death and other adverse events in suspected cardiac sarcoidosis. JACC Cardiovasc Imaging. 2013;6(4):501-511. doi: 10.1016/j.jcmg.2012.10.021 [DOI] [PubMed] [Google Scholar]
- 11.Watanabe E, Kimura F, Nakajima T, et al. Late gadolinium enhancement in cardiac sarcoidosis: characteristic magnetic resonance findings and relationship with left ventricular function. J Thorac Imaging. 2013;28(1):60-66. doi: 10.1097/RTI.0b013e3182761830 [DOI] [PubMed] [Google Scholar]
- 12.Flamée L, Symons R, Degtiarova G, et al. Prognostic value of cardiovascular magnetic resonance in patients with biopsy-proven systemic sarcoidosis. Eur Radiol. 2020;30(7):3702-3710. doi: 10.1007/s00330-020-06765-1 [DOI] [PubMed] [Google Scholar]
- 13.Nagai T, Kohsaka S, Okuda S, Anzai T, Asano K, Fukuda K. Incidence and prognostic significance of myocardial late gadolinium enhancement in patients with sarcoidosis without cardiac manifestation. Chest. 2014;146(4):1064-1072. doi: 10.1378/chest.14-0139 [DOI] [PubMed] [Google Scholar]
- 14.Nadel J, Lancefield T, Voskoboinik A, Taylor AJ. Late gadolinium enhancement identified with cardiac magnetic resonance imaging in sarcoidosis patients is associated with long-term ventricular arrhythmia and sudden cardiac death. Eur Heart J Cardiovasc Imaging. 2015;16(6):634-641. doi: 10.1093/ehjci/jeu294 [DOI] [PubMed] [Google Scholar]
- 15.Kuo L, Han Y, Mui D, et al. Diagnostic specificity of basal inferoseptal triangular late gadolinium enhancement for identification of cardiac sarcoidosis. JACC Cardiovasc Imaging. 2019;12(12):2574-2576. doi: 10.1016/j.jcmg.2019.06.024 [DOI] [PubMed] [Google Scholar]
- 16.Okasha O, Kazmirczak F, Chen KA, Farzaneh-Far A, Shenoy C. Myocardial involvement in patients with histologically diagnosed cardiac sarcoidosis: a systematic review and meta-analysis of gross pathological images from autopsy or cardiac transplantation cases. J Am Heart Assoc. 2019;8(10):e011253. doi: 10.1161/JAHA.118.011253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Velangi PS, Chen KA, Kazmirczak F, et al. Right ventricular abnormalities on cardiovascular magnetic resonance imaging in patients with sarcoidosis. JACC Cardiovasc Imaging. 2020;13(6):1395-1405. doi: 10.1016/j.jcmg.2019.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kalra R, Malik S, Chen KA, et al. Sex differences in patients with suspected cardiac sarcoidosis assessed by cardiovascular magnetic resonance imaging. Circ Arrhythm Electrophysiol. 2021;14(9):e009966. doi: 10.1161/CIRCEP.121.009966 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kramer CM, Barkhausen J, Flamm SD, Kim RJ, Nagel E; Society for Cardiovascular Magnetic Resonance Board of Trustees Task Force on Standardized Protocols . Standardized cardiovascular magnetic resonance imaging (CMR) protocols, society for cardiovascular magnetic resonance: board of trustees task force on standardized protocols. J Cardiovasc Magn Reson. 2008;10:35. doi: 10.1186/1532-429X-10-35 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kramer CM, Barkhausen J, Flamm SD, Kim RJ, Nagel E; Society for Cardiovascular Magnetic Resonance Board of Trustees Task Force on Standardized Protocols . Standardized cardiovascular magnetic resonance (CMR) protocols 2013 update. J Cardiovasc Magn Reson. 2013;15:91. doi: 10.1186/1532-429X-15-91 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Schulz-Menger J, Bluemke DA, Bremerich J, et al. Standardized image interpretation and post processing in cardiovascular magnetic resonance: Society for Cardiovascular Magnetic Resonance (SCMR) board of trustees task force on standardized post processing. J Cardiovasc Magn Reson. 2013;15:35. doi: 10.1186/1532-429X-15-35 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kawel-Boehm N, Hetzel SJ, Ambale-Venkatesh B, et al. Reference ranges (“normal values”) for cardiovascular magnetic resonance (CMR) in adults and children: 2020 update. J Cardiovasc Magn Reson. 2020;22(1):87. doi: 10.1186/s12968-020-00683-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wagner A, Mahrholdt H, Holly TA, et al. Contrast-enhanced MRI and routine single photon emission computed tomography (SPECT) perfusion imaging for detection of subendocardial myocardial infarcts: an imaging study. Lancet. 2003;361(9355):374-379. doi: 10.1016/S0140-6736(03)12389-6 [DOI] [PubMed] [Google Scholar]
- 24.Klem I, Klein M, Khan M, et al. Relationship of LVEF and myocardial scar to long-term mortality risk and mode of death in patients with nonischemic cardiomyopathy. Circulation. 2021;143(14):1343-1358. doi: 10.1161/CIRCULATIONAHA.120.048477 [DOI] [PubMed] [Google Scholar]
- 25.Klem I, Heiberg E, Van Assche L, et al. Sources of variability in quantification of cardiovascular magnetic resonance infarct size—reproducibility among three core laboratories. J Cardiovasc Magn Reson. 2017;19(1):62. doi: 10.1186/s12968-017-0378-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mooij CF, de Wit CJ, Graham DA, Powell AJ, Geva T. Reproducibility of MRI measurements of right ventricular size and function in patients with normal and dilated ventricles. J Magn Reson Imaging. 2008;28(1):67-73. doi: 10.1002/jmri.21407 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hicks KA, Tcheng JE, Bozkurt B, et al. ; American College of Cardiology; American Heart Association . 2014 ACC/AHA key data elements and definitions for cardiovascular endpoint events in clinical trials: a report of the American College of Cardiology/American Heart Association task force on clinical data standards (Writing Committee to Develop Cardiovascular Endpoints Data Standards). Circulation. 2015;132(4):302-361. doi: 10.1161/CIR.0000000000000156 [DOI] [PubMed] [Google Scholar]
- 28.Bakker A, Mathijssen H, Dorland G, et al. Long-term monitoring of arrhythmias with cardiovascular implantable electronic devices in patients with cardiac sarcoidosis. Heart Rhythm. 2022;19(3):352-360. doi: 10.1016/j.hrthm.2021.11.025 [DOI] [PubMed] [Google Scholar]
- 29.Alter P, Waldhans S, Plachta E, Moosdorf R, Grimm W. Complications of implantable cardioverter defibrillator therapy in 440 consecutive patients. Pacing Clin Electrophysiol. 2005;28(9):926-932. doi: 10.1111/j.1540-8159.2005.00195.x [DOI] [PubMed] [Google Scholar]
- 30.Koneru JN, Jones PW, Hammill EF, Wold N, Ellenbogen KA. Risk factors and temporal trends of complications associated with transvenous implantable cardiac defibrillator leads. J Am Heart Assoc. 2018;7(10):e007691. doi: 10.1161/JAHA.117.007691 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Al-Khatib SM, Stevenson WG, Ackerman MJ, et al. 2017 AHA/ACC/HRS guideline for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines and the Heart Rhythm Society. Circulation. 2018;138(13):e272-e391. doi: 10.1161/CIR.0000000000000549 [DOI] [PubMed] [Google Scholar]
- 32.Ellenbogen KA, Levine JH, Berger RD, et al. ; Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) Investigators . Are implantable cardioverter defibrillator shocks a surrogate for sudden cardiac death in patients with nonischemic cardiomyopathy? Circulation. 2006;113(6):776-782. doi: 10.1161/CIRCULATIONAHA.105.561571 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Incidence rates of the arrhythmic and heart failure endpoints by CMR phenotypes
eTable 2. Comparison between LGE prevalence by pathology-based features and extent
eFigure 1. Kaplan-Meier cumulative incidence curves for the arrhythmic endpoint stratified by individual pathology-frequent LGE features
eFigure 2. Kaplan-Meier cumulative incidence curves for the arrhythmic endpoint stratified by the number of pathology-frequent LGE features
eFigure 3. Kaplan-Meier cumulative incidence curves for the heart failure endpoint stratified by individual pathology-frequent LGE features
eFigure 4. Kaplan-Meier cumulative incidence curves for the heart failure endpoint stratified by the number of pathology-frequent LGE features
eFigure 5. Forest Plots of the multivariable analyses for the arrhythmic and heart failure endpoints
