Skip main navigation

Prognostic Value of Quantitative Contrast-Enhanced Cardiovascular Magnetic Resonance for the Evaluation of Sudden Death Risk in Patients With Hypertrophic Cardiomyopathy

Originally published 2014;130:484–495



    Hypertrophic cardiomyopathy (HCM) is the most common cause of sudden death in the young, although not all patients eligible for sudden death prevention with an implantable cardioverter-defibrillator are identified. Contrast-enhanced cardiovascular magnetic resonance with late gadolinium enhancement (LGE) has emerged as an in vivo marker of myocardial fibrosis, although its role in stratifying sudden death risk in subgroups of HCM patients remains incompletely understood.

    Methods and Results—

    We assessed the relation between LGE and cardiovascular outcomes in 1293 HCM patients referred for cardiovascular magnetic resonance and followed up for a median of 3.3 years. Sudden cardiac death (SCD) events (including appropriate defibrillator interventions) occurred in 37 patients (3%). A continuous relationship was evident between LGE by percent left ventricular mass and SCD event risk in HCM patients (P=0.001). Extent of LGE was associated with an increased risk of SCD events (adjusted hazard ratio, 1.46/10% increase in LGE; P=0.002), even after adjustment for other relevant disease variables. LGE of ≥15% of LV mass demonstrated a 2-fold increase in SCD event risk in those patients otherwise considered to be at lower risk, with an estimated likelihood for SCD events of 6% at 5 years. Performance of the SCD event risk model was enhanced by LGE (net reclassification index, 12.9%; 95% confidence interval, 0.3–38.3). Absence of LGE was associated with lower risk for SCD events (adjusted hazard ratio, 0.39; P=0.02). Extent of LGE also predicted the development of end-stage HCM with systolic dysfunction (adjusted hazard ratio, 1.80/10% increase in LGE; P<0.03).


    Extensive LGE measured by quantitative contrast enhanced CMR provides additional information for assessing SCD event risk among HCM patients, particularly patients otherwise judged to be at low risk.


    More than 50 years after its contemporary description, hypertrophic cardiomyopathy (HCM) remains the most common cause of sudden death in the young.16 Although several clinical markers have proved to be useful guides for risk stratification,35,7 current strategies do not identify all HCM patients at risk for sudden death.3,5,8,9 Over the last decade, implantable cardioverter-defibrillators (ICDs) have been effective in the primary prevention of sudden death in HCM,7,1012 underscoring the importance of more precise identification of those patients at highest risk.

    Editorial see p 455

    Clinical Perspective on p 495

    Ventricular tachyarrhythmias, emanating from regions of structurally abnormal myocardium (including areas of disorganized architecture and myocardial fibrosis), represent the likely mechanism of sudden death in HCM.3,7,10,1316 Contrast-enhanced cardiovascular magnetic resonance (CMR) imaging with late gadolinium enhancement (LGE) is capable of noninvasive identification of myocardial fibrosis in coronary artery disease and cardiomyopathies,1720 including HCM.2124 Although recent investigations in HCM have demonstrated an association between LGE and ambulatory ventricular tachyarrhythmias,14,2527 available data do not resolve the clinical utility of LGE in sudden death risk stratification.28,29 Therefore, we have assembled a large, multicenter HCM cohort, defined by eligibility for CMR, to investigate the prognostic value of LGE with respect to sudden cardiac death (SCD) events and other adverse disease consequences, including in those patients otherwise judged to be at low risk for sudden death in the context of current practice.


    Study Patients

    We evaluated 1669 HCM patients who were initially considered for CMR study at 7 HCM centers between November 2001 and February 2010. A total of 376 patients were excluded from the cohort on the basis of these criteria: prior implantation of an ICD (or other incompatible device), history of sustained ventricular tachycardia/ventricular fibrillation, claustrophobia, known associated obstructive coronary artery disease (including history of myocardial infarction or acute coronary event associated with increased cardiac enzymes or Q waves), other myocardial diseases, septal myectomy or alcohol ablation (before CMR), and incomplete follow-up (n=7). Therefore, the final study group comprised 1293 patients referred and eligible for CMR.

    The date of the first evaluation (ie, study entry) was the time of the initial CMR examination. Median follow-up from study entry to the most recent evaluation (clinic visit or telephone interview) or death (as of January 2012) was 3.3 years (quartile 1–3, 2.3–4.5 years). Selected data from 270 patients in the present study cohort have been part of previous analyses.30,31

    All patients signed statements approved by the Internal Review boards of participating institutions, agreeing to the use of their medical information for research. All authors had full access to and take responsibility for the integrity of the data and have agreed to the manuscript as written.


    HCM Diagnosis

    HCM diagnosis was defined as CMR documentation of a hypertrophied and nondilated left ventricle (LV; wall thickness ≥15 mm in adults and the equivalent relative to body surface area in children) at some point during their clinical course in the absence of another cardiac or systemic disease capable of producing similar magnitude of hypertrophy.2,4

    SCD Events

    Sudden death was defined as unexpected sudden collapse occurring within 1 hour from the onset of symptoms in patients with a previously stable or uneventful clinical course. Additionally, potentially lethal cardiovascular events in which patients were successfully resuscitated from cardiac arrest (with documented ventricular fibrillation) or received appropriate defibrillation interventions from an ICD were regarded as equivalent to sudden death and are included in all references to SCD events. Stored intracardiac electrograms were analyzed independently at each center by expert electrophysiologists blinded to CMR results; ICD discharges were characterized as appropriate if triggered by ventricular fibrillation or rapid ventricular tachycardia (rate, ≥180 bpm).7,10

    Risk Stratification

    At study entry, each patient was assessed for the conventional primary prevention sudden death risk factors described in HCM35,32: (1) history of HCM-related sudden death in ≥1 first-degree or other relatives <50 years of age; (2) massive LV hypertrophy (maximum wall thickness ≥30 mm); (3) nonsustained ventricular tachycardia (≥3 consecutive ventricular beats, ≥120 bpm) on 24-hour ambulatory (Holter) ECG monitoring; and (4) unexplained syncope, inconsistent with neurocardiogenic origin, occurring within 5 years before CMR evaluation. Hypotensive blood pressure response to exercise was excluded from this analysis because exercise testing for risk stratification was customary practice in only a minority of patients..7,10,12 The conventional risk factors were then collapsed into a score ranging from 0 to 4, depending on the number of risk factors.

    Ambulatory (Holter) ECGs were obtained in 1034 patients at the discretion of investigators at each of the participating HCM centers. In 259 patients, this test was not performed as a result of patient refusal or advanced age or was not judged necessary with high-risk status already established by other risk markers. Low-risk status was judged to be present in 598 study patients in whom all 4 risk factors were tested and were negative and in 184 patients in whom 3 risk factors were tested (exclusive of the Holter ECG) and were negative.

    Heart Failure

    Adverse heart failure–related events and mortality were defined as symptom progression during follow-up period to New York Heart Association functional class III or IV.33 Patients with refractory heart failure and heart transplantation were considered equivalent to HCM-related heart failure death. End-stage phase of HCM with LV remodeling was defined by a CMR-derived LV ejection fraction (EF) <50%.34

    CMR Imaging

    CMR imaging was performed with a 1.5-T scanner (Philips, Best, the Netherlands; or Siemens, Erlangen, Germany) using steady-state, free-precession breath-hold cines in 3 long-axis planes and sequential short-axis slices from the atrioventricular ring to the apex.19,29 LGE images were acquired 10 to 20 minutes after intravenous administration of 0.2 mmol/kg gadolinium-DTPA with breath-hold 2-dimensional segmented inversion-recovery sequence or phase-sensitive inversion-recovery sequences in identical planes as in cine images. Inversion time was optimized to null normal myocardial signal. For phase-sensitive sequences, uncorrected magnitude images were used.

    CMR Analysis

    Images from all centers were transferred to a core laboratory (PERFUSE, Boston, MA) for central, blinded analysis. LV volume, mass, and EF were measured by use of standard volumetric techniques and analyzed with commercially available software (QMASS version 7.4, Medis Inc). LV chamber was assessed according to the American Heart Association 17-segment model.35 LV endocardial and epicardial borders on cine images were manually planimetered to define the myocardium, taking care to exclude papillary muscles and the intertrabecular blood pool. Maximal LV wall thickness was defined as the greatest dimension at any site within the LV myocardium.

    The LV short-axis stack of LGE images was first assessed visually for the presence of LGE by 2 experienced readers (R.H.C. and E.A.) blinded to patient profiles and clinical outcome, with any disagreement adjudicated by a third expert reader (W.J.M.). Quantification of LGE was then performed by 1 expert reader (R.H.C.) on all LGE-positive studies by manually adjusting a gray-scale threshold to define areas of visually identified LGE (see the online-only Data Supplement text and video). These areas were then summed to generate a total volume of LGE and expressed as a proportion of total LV myocardium (%LGE). Therefore, in this study, LGE was used as an imaging risk marker for clinical outcome and events (presumably representing myocardial fibrosis).21,23,24

    At PERFUSE, the visual LGE quantitation method used here for assessment of LGE has been previously reported and validated with high reproducibility with strong correlation to the gray-scale threshold method of ≥6 SDs exceeding the mean of normal myocardium (r=0.9, P<0.001; see the online-only Data Supplement).36 In addition, high gray-scale thresholding methods have recently been validated by histopathology and have been shown to provide the best representation of total fibrosis burden (ie, replacement and interstitial).37,38 Therefore, we want to underscore that the results presented here pertaining to the extent of LGE in predicting SCD events are reliable only when the same quantification technique used in this study is applied. The time required for quantification using the visual gray-scale threshold method averaged 10 minutes per study. The software required to perform this analysis is now available on a number of commercial imaging platforms. For additional methods, please see the online-only Data Supplement.

    To test interobserver agreement (R.H.C. and E.A.), LGE was quantified using the visual thresholding on 24 randomly selected studies. Intraobserver agreement (R.H.C.) was analyzed 12 months after the initial image analysis. For interobserver and intraobserver agreement measurements, endocardial and epicardial borders were retraced, and the amount of LGE was recalculated using visual gray-scale method.

    Statistical Analysis

    Continuous and categorical data are expressed as mean±SD, median (quartiles 1–3), or n (%) as appropriate. Comparisons of characteristics between groups were made with the unpaired Student t test, χ2 test, or Fisher exact test as appropriate. All reported P values are 2 sided. The prespecified primary clinical end point was SCD events and included the composite of sudden death, aborted cardiac arrest, or appropriate ICD discharge for ventricular tachycardia/ventricular fibrillation.

    Survivor curves comparing patients with and without LGE were constructed by the Kaplan-Meier method, and differences between groups were examined by use of a log-rank test for equality of survivor functions. The relationship between the presence or amount of LGE and the likelihood of subsequent clinical events was further evaluated through the use of univariate and multivariable Cox proportional hazards model. The proportional hazards assumption was tested graphically and with time-dependent covariates before proceeding.

    The multivariable model was constructed to adjust for possible confounders using a stepwise selection method with an entrance and stay criteria of P<0.20, forcing the number of conventional sudden death risk factors into all models a priori. Variables entered into the multivariable model for SCD events thus include %LGE, conventional SCD risk factors (all 4 risk markers collapsed into analysis as 1 continuous variable), and maximal LV thickness. After the model was completed, the remaining candidate variables (ie, age, LV mass, EF, LV outflow tract gradient at rest, and septal reduction therapies performed after CMR) were retested individually with a sensitivity analysis to examine their influence on effect estimates. Separate multivariable models were constructed and retested in a similar fashion for death resulting from any cause and the development of end-stage HCM.

    The incremental value of LGE in predicting 5-year SCD event risk was assessed in the overall cohort and the prespecified subgroup without conventional sudden death risk factors using area under the receiver-operating characteristics curve, integrated discrimination improvement, and net reclassification improvement (NRI).39,40 For NRI, the 5-year predicted risk for SCD was divided into 3 risk categories, defined as low (≤0.5%/y, 2.5%/5 y), high (≥1.5%/y, 7.3%/5 y), and intermediate (0.6%–1.4%/y, 2.6%–7.2%/5 y).5 To account for sampling variability, confidence intervals (CIs) for all measures were obtained by bootstrapping with 1000 resamples. Results were internally validated with the bootstrap approach, and degree of overoptimism was calculated for each performance metric.41 The optimism for the performance of the final multivariable risk prediction model is estimated by calculating the average difference between model performance (as measured by the area under the receiver-operating characteristics curve) in 500 bootstrap samples and the model performance of the original sample (ie, the full data set). All analyses were performed with SAS 9.3 (SAS Institute, Cary, NC).


    Baseline Characteristics

    Clinical and demographic characteristics of the study population are summarized in Table 1. At CMR, the mean patient age was 46±17 years (range, 7–87 years), 815 (63%) were male, and initial New York Heart Association class was 1.6±0.7. Resting LV outflow tract obstruction (gradient ≥30 mm Hg) was present in 302 patients (23%), with no difference in the prevalence of rest obstruction between patients with and without LGE (n=119 [22%] versus n=183 [24%]; P=0.42).

    Table 1. Demographic and Clinical Characteristics of 1293 HCM Patients With CMR

    VariableAll Patients(n=1293)
    Age, y46±17
    Male, n (%)815 (63)
    Body surface area, g/m21.9±0.3
    NYHA class, n (%)
     I735 (57)
     II380 (29)
     III/IV178 (14)
    Atrial fibrillation, n (%)159 (12)
    Basal LVOT gradient ≥30 mm Hg302 (23)
    CMR variables
     LVED dimension, mm54±7
     LVEF %67±9
     Maximum LV thickness, mm20±5
     LV mass, g163±71
     LV mass index, g/m283±34
     LGE, n (%)548 (42)
     LGE, median (Q1–Q3), g9 (4–21)
     %LGE, median (Q1–Q3)5 (3–13)
    Location of LGE, n (%)
     Septum275 (51)
     LV free wall143 (26)
     Septum and LV free wall89 (16)
     Apex187 (34)
     Only at RV insertion into LV134 (25)
    Risk factors (0–4), n (%)
     0/1/2/3/4 risk factors782 (60)/415 (32)/90 (7)/6 (0.5)/0 (0)
     Nonsustained VT on ambulatory Holter*204 (20)
     Unexplained syncope122 (9)
     Family history of SCD219 (17)
     Maximum LV wall thickness ≥30 mm68 (5)
    Drugs, n (%)
     β-Blockers741 (57)
     Calcium channel antagonists257 (20)
     ACE-I/ATII194 (15)
     Amiodarone42 (3)
     Disopyramide73 (6)
     Diuretics146 (11)
    ICD implantation after initial CMR259 (20)
    Genetic testing, n (%)
     MYBPC399 (24)
     MYH777 (19)
     TNNT216 (4)
     TPM13 (0.7)
     ACTC1 (0.2)
     Others11 (5)
    Duration of follow-up, median (Q1–Q3), y3.3 (2.3–4.5)
    Major clinical events during follow-up, n (%)
     HCM-related Sudden Death14 (1.0)
     Aborted arrest6 (0.5)
     ICD discharge (VT/VF)17 (1.3)
     Heart failure death6 (0.5)
     Heart transplantation9 (0.7)
     End-stage HCM87 (7)
     Progression to NYHA class III/IV§99 (9)
     Noncardiac death21 (1.6)

    ACE-I indicates angiotensin-converting enzyme inhibitor; ACTC, α-cardiac actin; ATII, angiotensin receptor blocker; CMR, cardiovascular magnetic resonance; HCM, hypertrophic cardiomyopathy; ICD, implantable cardioverter-defibrillator; LGE, late gadolinium enhancement; LV, left ventricular; LVED, left ventricular end-diastolic dimension; LVEF, left ventricular ejection fraction; LVOT, left ventricular outflow tract; MYBPC3, cardiac myosin binding protein C; MYH7, β-myosin heavy chain; NYHA, New York Heart Association; Q1, quartile 1; Q3, quartile 3; RV, right ventricular; SCD, sudden cardiac death; TNNT2, troponin T; TPM1, α-tropomyosin; VF, ventricular fibrillation; and VT, ventricular tachycardia.

    *By convention, ambulatory 24-hour Holter monitors were ordered at the discretion of the investigators at each HCM center, with Holter monitoring performed in 1034 of 1293 study patients.

    Four hundred fourteen patients (32%) underwent genetic testing for HCM.

    Fifty-eight had end-stage HCM at study entry; 26 others developed end-stage HCM during follow-up.

    §Includes only 1115 patients who were in NYHA class I/II at study enrollment.

    Clinical Outcome

    During follow-up, SCD events occurred in 37 patients (3%): 14 died suddenly, 6 survived an aborted cardiac arrest, and 17 had appropriate primary prevention ICD therapy for ventricular tachycardia/ventricular fibrillation (cumulative SCD events incidence, 0.9%/y; Table 2). There was no difference in SCD event risk between patients taking cardioactive medications (ie, β-blockers, calcium channel blockers, disopyramide, or amiodarone) and those not taking these medications (P=0.37). In addition, 118 patients experienced adverse HCM-related heart failure events: 99 survivors with progressive heart failure symptoms to New York Heart Association class III/IV and 19 patients who died of heart failure or embolic stroke or underwent heart transplantation; 17 patients died of noncardiac causes, most commonly cancer and sepsis.

    Table 2. Clinical and Demographic Characteristics of 37 Patients Experiencing SCD Events

    PatientAge, ySexNYHA ClassLVEF, %Maximum Wall Thickness, mmLV Mass Index, g/m2LVOTGradientLGELGE,g (% of LV mass)24-Hour HolterRisk FactorsEvent
    Low-risk patients
     113F162226240 (0)YSudden Death
     214M172138300 (0)YSudden Death
     315M169168000 (0)YSudden Death
     419F231201060+80 (40)YSudden Death
     520M25927830+7 (5)YAlive, ICD shock
     622F15722576+1 (1)YAlive, aborted arrest
     730M26326710+17 (10)YAlive, aborted arrest
     835M171251290+15 (7)YAlive, aborted arrest
     935F25114485+53 (54)YAlive, ICD shock
     1037M26323968+41 (26)YAlive, ICD shock
     1139F166201030+6 (4)YSudden Death
     1247M160195450 (0)YAlive, aborted arrest
     1348M1542412035+3(1)NAlive, aborted arrest
     1455M260221048+87(36)YAlive, ICD shock
     1557F25216629+11(11)NSudden Death
     1659M279271420+8(3)YSudden Death
     1761F27522108600 (0)YSudden Death
     1864M3822111770+2(1)YSudden Death
     1966F374238000 (0)YSudden Death
     2072M1702212700 (0)YAlive, aborted arrest
     2173M2712091100 (0)YSudden Death
    Patients with ≥1 risk factor
     2215F15920950+5 (3)NFHAlive, ICD shock
     2321M173261750+21 (6)YNSVTAlive, ICD shock
     2427F16616650-0 (0)YSyncopeAlive, ICD shock
     2548F27022890+12 (7)YNSVTAlive, ICD shock
     2654M158281170+25 (10)YNSVTAlive, ICD shock
     2758M16517480+2 (2)YFHSudden Death
     2858M16519430+5 (2)YFHSudden Death
     2961F16616460+17 (23)YNSVTAlive, ICD shock
     3065M33520865+27 (17)YNSVTAlive, ICD shock
     3178M36622690-0 (0)NSyncopeAlive, ICD shock
     3225F16918710+10 (10)YFH,NSVTAlive, ICD shock
     3331M1655013816+43 (19)YNSVT,extreme LVHAlive, ICD shock
     3445F27732900+50 (31)YNSVT,extreme LVHAlive, ICD shock
     3549F37725780-0 (0)YSyncope,FHAlive, ICD shock
     3666M381221010+9 (5)YFH,NSVTSudden Death
     3736F162221078+28 (12)YSyncope, FH, NSVTAlive, ICD shock

    FH indicates family history of sudden death resulting from hypertrophic cardiomyopathy; HCM, hypertrophic cardiomyopathy; ICD, implantable cardioverter-defibrillator; LA, left atrium; LGE, late gadolinium enhancement; LV, left ventricle; LVH, left ventricular hypertrophy; LVOT, left ventricular outflow tract; NSVT, nonsustained ventricular tachycardia; and SCD, sudden cardiac death.

    Distribution of LGE

    Of the 1293 study patients, LGE was present in 548 (42%; Figure 1). Of those with LGE, the extent was 9±10% of the LV myocardial mass: ≤10% of the LV (n=381, 29%), 11% to 19% (n=94, 7%), and ≥20% (n=73, 6%). Among 37 patients with SCD events, LGE was present in 26 (70%; Table 2), occupying 13±14% of the LV myocardium.

    Figure 1.

    Figure 1. Contrast-enhanced cardiovascular magnetic resonance images in 4 patients with hypertrophic cardiomyopathy. A, Basal left ventricular (LV) short-axis image from an asymptomatic 29-year-old man without conventional risk factors. Focal areas of late gadolinium enhancement (LGE) are confined to the midmyocardial anterior wall (arrows), encompassing 4% of the LV mass. B, Mid-LV short-axis image from a 61-year-old woman with substantial LGE (23% of LV mass) involving the basal anterior septum and contiguous anterolateral free wall (thick arrows), as well as focally at the intersection of right ventricular (RV) free wall and posterior septum (thin arrow). A 12-beat nonsustained ventricular tachycardia (VT; 180 bpm) run on 24-hour ambulatory ECG was the only evidence of increased sudden cardiac death (SCD) risk. Extensive LGE was the arbitrator for the decision to implant a cardioverter-defibrillator (ICD) for primary prevention, which 5 months later terminated an episode of rapid VT. C, A 4-chamber long-axis image from mildly symptomatic 54-year-old man without conventional SCD risk factors and normal ejection fraction (EF; 60%) but with transmural LGE involving the distal posterior septum, apex, and lateral free wall (arrows) encompassing 36% of the LV mass. One year after ICD implantation, this patient received a shock for rapid monomorphic VT (180 bpm). D, A 4-chamber long-axis image from 29-year-old man with extensive LGE involving large portions of the ventricular septum (arrows) encompassing 32% of the overall LV mass. Over follow-up, he developed end stage with systolic dysfunction (EF, 40%) associated with progressive heart failure (New York Heart Association class III) and currently awaits heart transplantation.

    Association of %LGE and SCD

    During follow-up, SCD event risk was significantly greater among HCM patients with LGE compared with patients without any evidence of LGE (log-rank P=0.002). Notably, the unadjusted SCD event incidence per 1000 person-years increased in direct relation to the extent of LGE: 4 without LGE (95% CI, 2–8), 10 with LGE ≤10% (95% CI, 6–18), 18 with 11% to 19% (95% CI, 7–39), and 24 with ≥20% (95% CI, 9–51; P=0.001 for trend; Figure 2). The absence of LGE was associated with lower risk for SCD events (adjusted hazard ratio [HRadj], 0.39; 95% CI, 0.18–0.84; P=0.02; Figure 2). In addition, in a subgroup analysis of 1008 patients, SCD events were not significantly increased in HCM patients with minimal LGE (1%–5%) compared with those with no LGE (P=0.09).

    Figure 2.

    Figure 2. Relation between extent of late gadolinium enhancement (LGE) and sudden cardiac death (SCD) events in 1293 patients with hypertrophic cardiomyopathy. A, Hazard plot based on multivariable Cox regression analysis (P=0.008). B, Incidence of SCD events increased progressively and in direct relation to the extent of LGE (P<0.001).

    Prediction of SCD Events by %LGE

    Notably, adjusted SCD event risk increased in a continuous and direct manner with respect to the extent of LGE (Table 3 and Figure 3). The extent of LGE was a strong predictor of SCD events in that each 10% increase in LGE was associated with 40% increase in relative SCD events risk (HRadj, 1.46/10% increase in LGE; 95% CI, 1.12–1.92; Wald χ2=9.6; P=0.002; Table 4 and Figure 2), independently of patient age (P=0.89 for interaction). In addition, even when we consider those 184 HCM patients who did not undergo Holter ECG monitoring (and who had none of the other conventional risk factors) as theoretically having a positive Holter with nonsustained ventricular tachycardia, the relative risk of LGE in predicting SCD for the total cohort of 1293 patients remained essentially unchanged (HRadj, 1.45/10% LGE; 95% CI,1.11–1.90; Wald χ2, 7.5; P=0.006).

    Table 3. Adjusted HRs and Estimated 5-Year Sudden Cardiac Death Event Rates for the HCM Cohort and the Low-Risk Subgroup Without Conventional Risk Markers

    %LGEAdjusted HR Point Estimate*95% CIEstimated 5-y SCD event rate (%)95% CI
    Total cohort–4.8
    Low risk–4.6

    CI indicates confidence interval; HR, hazard ratio; LGE, late gadolinium enhancement; and SCD, sudden cardiac death.

    *Adjusted for number of conventional sudden death risk factors and left ventricular ejection fraction.

    Table 4. Results of Univariate and Multivariable Cox Proportional-Hazards Analyses of the Relation Between Baseline Clinical Variables and Outcome

    Sudden Death Event,Univariate AnalysisSudden Death Event, Multivariable Analysis*Death Resulting From Any Cause,Univariate AnalysisDeath Resulting From Any Cause, Multivariable AnalysisDevelopment of End-Stage HCM, Univariate AnalysisDevelopment of End-Stage HCM, Multivariable Analysis
    VariableHazard Ratio(95% CI)P ValueHazard Ratio(95% CI)P ValueHazard Ratio(95% CI)P ValueHazard Ratio(95% CI)P ValueHazard Ratio(95% CI)P ValueHazard Ratio(95% CI)P Value
    %LGE (per 10% increase)1.50 (1.22–1.85)0.00011.46 (1.12–1.92)0.0021.35 (1.07–1.71)0.011.51 (1.13–2.01)0.0061.89 (1.47–2.43)<0.00011.80 (1.40–2.40)0.03
    Age (per decade increase)0.93 (0.77–1.12)0.44NANA1.67 (1.37–2.05)<0.00011.67 (1.34–2.08)<0.00011.01 (0.82–1.26)0.91NANA
    Sudden death risk factors1.39 (0.89–2.16)0.151.17 (0.74–1.85)0.800.64 (0.37–1.09)0.100.48 (0.26–0.89)0.020.90 (0.50–1.63)0.411.04 (0.68–1.58)0.87
    LV mass (per 10 g increase)1.01 (0.97–1.05)0.76NANA1.00 (0.96–1.04)0.79NANA0.99 (0.94–1.05)0.79NANA
    LVEF (per 10% decrease)1.26 (0.82–1.72)0.140.99 (0.69–1.42)NA1.41 (1.06–1.84)0.021.22 (0.90–1.65)0.204.29 (2.46–7.46)<0.00012.63 (2.12–3.23)NA

    LGE indicates late gadolinium enhancement; LV, left ventricular; LVEF, left ventricular ejection fraction; left ventricular outflow tract obstruction; and NA, not applicable.

    *Adjusted for number of conventional sudden death risk factors and LVEF. Sensitivity analysis using age, LV mass, maximal LV wall thickness, left ventricular outflow tract obstruction, and septal reduction therapy did not change effect estimates.

    Adjusted for age, conventional sudden death risk factors, and LVEF. Sensitivity analysis using LV mass, maximal LV wall thickness, left ventricular outflow tract obstruction, and septal reduction therapy did not change effect estimates.

    Adjusted for conventional sudden death risk factors and LVEF. Sensitivity analysis using age, LV mass, left ventricular outflow tract obstruction, and septal reduction therapy did not change effect estimates.

    Figure 3.

    Figure 3. Predicted 5-year event rates relative to late gadolinium enhancement (LGE) by percent left ventricular mass for risk of end-stage HCM with systolic dysfunction, sudden cardiac death events, and total mortality.

    Compared with patients without LGE, the HRadj of SCD events related to %LGE was as follows: 10%, HRadj=1.46; 15%, HRadj=1.77; and 20%, HRadj=2.14 (Figure 3 and Table 3). The estimated risk of SCD events at 5 years increased incrementally with respect to %LGE, ranging from 4.9% in patients with 10% LGE to 6.9% in patients with 20% LGE (Figure 3 and Table 3).

    In addition, the extent of LGE remained a significant predictor of SCD events, even after the exclusion of HCM patients with an EF <50% at study entry (HRadj, 1.61/10% LGE; 95% CI, 1.21–2.16; P=0.002). The extent of LGE was also a predictor of SCD events when expressed as total grams (HRadj, 1.13 per 10 g LGE; 95% CI, 1.01–1.28; P=0.04). Furthermore, LGE was an independent predictor of all-cause mortality (P=0.006; Table 4).

    Relation of LGE to SCD Event Risk in Patients With Conventional Risk Factors

    Sixteen HCM patients with ≥1 risk factors experienced SCD events (Table 2). A strong trend was present between sudden death risk and extent of LGE (unadjusted hazard ratio, 1.32/10% LGE; 95% CI, 0.93–1.86). %LGE was a stronger predictor of SCD events compared with each of the individual risk factors (univariate analysis global Wald statistic for %LGE=13.8 versus 0.1 for massive LVH, 0.9 for syncope, 0.1 for family history of SCD, and 3.3 for nonsustained ventricular tachycardia; P≤0.001 for each comparison; Table 5). In addition, when %LGE is considered together with each of the risk markers, the incremental prognostic value in predicting SCD events is increased significantly (Table 5).

    Table 5. Univariate and Bivariate Analyses of the Extent of LGE Versus Conventional Sudden Death Risk Factors Among 1293 HCM Patients

    Univariate HR(95% CI)Model Global Wald χ2P ValueBivariate HR(95% CI)PModel Global Wald χ2P ValueP Value, Univariate vs Bivariate
    Massive LVH0.92 (0.22–3.84)0.01220.910.72 (0.17–2.99)0.6514.4780.00070.0001
    %LGE (per 10% increase)1.50 (1.22–1.85)13.894<0.0011.50 (1.22–1.85)<0.001
    Syncope1.18 (0.42–3.32)0.09450.761.07 (0.38–3.06)0.9014.0230.00090.0001
    %LGE (per 10% increase)1.50 (1.22–1.85)13.894<0.0011.49 (1.21–1.84)<0.001
    Family History of SCD1.19 (0.52–2.70)0.1640.6861.05 (0.45–2.43)0.9113.9660.00090.0002
    %LGE (per 10% increase)1.50 (1.22–1.85)13.894<0.0011.49 (1.20–1.85)<0.001
    Non sustained VT1.97 (0.95–4.07)3.3490.06721.61 (0.77–3.36)0.2114.7410.00060.0007
    %LGE (per 10% increase)1.50 (1.22–1.85)13.894<0.0011.46 (1.17–1.82)<0.001
    Sudden death risk factors (per risk factor)1.37 (0.88–2.14)1.8960.171.17 (0.74–1.85)0.5114.3730.00080.0004
    %LGE (per 10% increase)1.50 (1.22–1.85)13.894<0.0011.46 (1.17–1.82)0.0009

    CI indicates confidence interval; HCM, hypertrophic cardiomyopathy; HR, hazard ratio; LGE, late gadolinium enhancement; LVH, left ventricular hypertrophy; SCD, sudden cardiac death; and VT, ventricular tachycardia.

    LGE in Low-Risk Patients

    Of the 37 patients with SCD events, 21 (57%) were considered to be at lower risk for SCD using current clinical parameters.8 Among these patients, SCD event risk increased in direct proportion to extent of LGE (HRadj, 1.66/10% LGE; 95% CI, 1.24–2.23; Wald χ2=11.56; P=0.0007). Therefore, compared with patients without LGE, the relative risk of SCD event in patients judged at lower risk related to %LGE was 10% (HRadj, 1.66), 15% (HRadj, 2.14), and 20% (HRadj, 2.76; Table 3). The estimated risk of SCD event at 5 years also increased in an incremental manner with respect to %LGE, ranging from 4.9% in patients with 10% LGE to 8.1% in patients with 20% LGE (Table 3). In addition, when the analysis was restricted to the 598 low-risk patients with all 4 conventional risk factors assessed (and with negative results), the relative risk of SCD events remained essentially unchanged (HR per 10% LGE, 1.62; 95% CI, 1.21–2.37, Wald χ2=9.39; P=0.002).

    Enhanced SCD Event Risk Model by LGE

    The performance of the multivariate SCD event risk model was improved by the addition of LGE (likelihood ratio P=0.0075). The area under the receiver-operating characteristics curve increased from 0.710 (95% CI, 0.632–0.788) to 0.741 (95% CI, 0.664–0.818); the relative integrated discrimination improvement was 0.565 (95% CI, 0.019–3.564); and the NRI was 0.129 (95% CI, 0.003–0.383). This change in the model resulted from patients with SCD events being reclassified from a lower to higher SCD risk category (event NRI, 0.130; nonevent NRI, −0.001). When confined to the low-risk patient subgroup, the relative integrated discrimination improvement was 1.737 (95% CI, 0.044–19.19) and NRI was 0.295 (95% CI, 0.120–0.543; event NRI, 0.172 versus nonevent NRI, 0.123). Internal validation of the risk model using bootstrapping suggested limited degrees of optimism (<8% for all risk factor performance metrics).

    Relation of LGE to Systolic Dysfunction (End Stage)

    At study entry, 1235 patients had preserved systolic function (EF ≥50%) and 58 were in the end stage of HCM, characterized by systolic dysfunction (EF <50%). Twenty-six of those 1235 patients with normal EF evolved to end-stage HCM during follow-up, including 13 with progression to New York Heart Association class III/IV, transplantation, stroke, or heart-failure death.

    Amount of LGE at study entry was greater in the 26 patients who developed end-stage HCM during follow-up compared with the 1209 HCM patients in whom systolic function remained within the normal range (13±15% versus 3±6% LGE; P<0.0001; Figure 1). Therefore, the extent of LGE was a strong independent predictor of the development to end-stage HCM (HRadj, 1.80/10% increase in LGE; 95% CI, 1.40–2.40; P=0.03; Figure 3 and Table 4). %LGE was not a determinant of adverse heart failure events/mortality in HCM patients with preserved EF (≥50%; P=0.23).


    The visual gray-scale thresholding method was associated with good reproducibility: intraobserver coefficient of variation, 5.9±1.1%; interobserver coefficient of variation, 6.3±1.2%; and concordance correlation coefficient (ρc), 0.996, with minimal bias (bias, −0.1g; 95% CI, −3.5 to 3.3).


    Although effective in promoting the prevention of SCD, current risk stratification strategies in HCM patients do not identify all at-risk patients, largely as a result of the substantial heterogeneity in clinical and phenotypic expression of this genetic disease.2,3,5,8,42 Because SCD can occur in HCM patients considered to be at low risk, identification of additional markers to allow more precise selection of those patients who may benefit from primary prevention ICD therapy represents a major clinical aspiration.3,28 Recently, considerable interest in using contrast-enhanced CMR to improve the risk stratification model has emerged.4348 However, available data on the prognostic value of CMR do not provide an opportunity to specifically predict SCD risk for individual HCM patients.28,29 In the present large, multicenter HCM cohort, we have investigated LGE as a predictor of SCD and other adverse disease consequences among a predominantly lower-risk cohort of HCM patients.

    Our data show that in the overall HCM study cohort (n=1293), extensive LGE remained an important marker of increased risk for SCD, even after adjustment for other relevant disease variables, including EF. Notably, a continuous relationship between risk of SCD and amount of LGE emerged as a general principle. Compared with patients without LGE, SCD risk increased substantially across the range of LGE amounts, with LGE ≥15% of the LV mass conferring a >2-fold risk in patients otherwise considered low risk. The 3 model performance metrics we used (ie, area under the receiver-operating characteristics curve, integrated discrimination improvement, and NRI) demonstrated an improvement in the SCD risk model after the addition of LGE, substantiating LGE as a risk marker for SCD in HCM and providing information which exceeded that currently available.39,40 In addition, we found that the association between SCD risk and LGE was independent of patient age, although underrepresentation of young patients in our cohort could have influenced this observation.

    Perhaps the most important and novel finding of this multicenter study was the unique opportunity to identify SCD risk among an important but underrecognized subgroup of predominantly asymptomatic HCM patients previously considered (from current clinical criteria) to be at low risk for lethal ventricular tachyarrhythmias. Without the application of contrast-enhanced CMR to HCM, these patients would potentially remain unprotected against SCD, with no impetus to implant ICDs for primary prevention.7,10 Because a substantial portion of clinically identified HCM patients do not demonstrate acknowledged risk factors sufficient to be definitely regarded at increased risk,4,5 CMR alone could identify some of these unrecognized high-risk patients who could potentially benefit from this enhanced risk stratification model.7

    However, we would like to emphasize that an essential element of these data is the linear relation between %LGE and SCD event risk, which avoids the imposition of a single and rigid LGE cut point (eg, ≥15%). Indeed, using graded risk levels (depicted in Table 3) is a more realistic and clinically useful strategy for estimating relative risk. It allows prophylactic ICD decisions to be resolved in the context of the continuous relation between %LGE and SCD risk, in accord with the wishes of the fully informed and autonomous patient and the managing cardiologist and in agreement with what constitutes an unacceptable level of risk.49

    In addition, our results support a role for %LGE with contrast-enhanced CMR as a novel imaging marker to aid in more accurately identifying patients at risk for SCD who otherwise may have some evidence of enhanced risk.50 For example, 15% LGE was associated with an almost 2-fold increase in SCD event risk in patients with ≥1 risk factors compared with patients with risk factors (but without LGE). This consideration becomes a useful clinical tool for those HCM patients situated in the ambiguous gray zone of HCM risk stratification because, not uncommonly, a single risk marker may be poorly or incompletely defined. In such cases, extensive LGE can act as a potential arbitrator for resolving otherwise ambiguous ICD decisions.

    Previous contrast-enhanced CMR studies in HCM have focused largely on the association between the presence of LGE and SCD.4548 However, evidence of any amount of LGE per se cannot be regarded as a risk marker because this designation attributes equal predictive weight to a broad spectrum of LGE amounts (eg, from minimal to extensive). Furthermore, assigning increased risk to HCM patients on the basis of solely the presence of LGE per se conveys an impractical and clinically imprudent message, given that most CMR studies report some LGE in >50% of HCM patients.29,43,44,47 By inference, most such HCM patients could theoretically be regarded as potential candidates for primary prevention ICDs, including a very large proportion who would not benefit from this therapy and could be exposed only to potential device complications.3,4,7

    On the other hand, the absence of LGE itself was associated with lower risk of SCD events, which may serve to influence decision making against ICD implantation in those patients for whom high-risk status remains uncertain on the basis of conventional risk stratification. Nevertheless, we should note that the absence of LGE was not absolutely protective against SCD risk in this cohort, suggesting that susceptibility to potentially lethal ventricular tachyarrhythmias in HCM can be influenced by factors other than myocardial fibrosis.14 We also recognize that although focal LGE can be assessed reliably in the vast majority of patients, technical limitations occasionally make precise quantification of small amounts of LGE challenging. However, the incremental increase in absolute SCD event risk associated with very small amounts of LGE (ie, in the range of 1%–5%) is trivial and does not differ significantly from that in patients without LGE. Greater insights into this issue of LGE and SCD risk could be achieved through the emergence of novel CMR techniques (eg, T1 mapping), which could provide an even more robust assessment of the abnormal myocardial substrate in HCM.51

    Our data also identify an association between extensive LGE (presumably a marker for replacement fibrosis) and progressive heart failure with systolic dysfunction (ie, end-stage HCM).52 In the present study, we have also shown prospectively that in patients with preserved systolic function at study entry, extensive amounts of LGE can be predictive of subsequent remodeling and evolution to the end stage.34 Indeed, a continuous relationship between future development of end-stage HCM and %LGE was demonstrated, with ≥20% LGE conveying a >3-fold increase in risk (compared with patients without LGE). The capability to prospectively identify patients who will progress to the end stage is clinically relevant by permitting anticipation of changes in clinical course and management strategies, including tailored drug administration and early consideration for heart transplantation and prophylactic defibrillators.3,4,34


    A certain degree of preference with respect to patient selection was unavoidable within our study design because of the exclusion of some high-risk patients in whom ICDs were implanted before CMR. As a result, the present large, multicenter HCM cohort was made up predominantly of patients at low risk for sudden death. In addition, HCM is generally a low-event-rate disease,3,5,6 contributing to the relatively small number of patients with sudden death in this study cohort. These notable features of our study design and patient selection may also have influenced our analysis supporting LGE as statistically a stronger predictor of SCD events than each of the individual conventional risk factors used in HCM. Therefore, these data should not obscure the time-honored efficacy of the current conventional risk factor model for identifying high-risk patients documented to have substantial benefit for prevention of sudden death with prophylactic ICD therapy, which has served the HCM patient population so well over the last 15 years. Finally, although the present data were assembled by necessity in HCM referral centers, the results should nevertheless have implications for the broader HCM disease spectrum, given that the clinical profile of our cohort is similar to that reported in the HCM literature with respect to demographics, %LGE, and outcome rates.2,4,4345 Furthermore, with the growing penetration of CMR into clinical cardiovascular practice,17 our referral center–derived data should become increasingly applicable to patient decision making. Notably, in 17 of our 37 sudden death events, the ICD detected arrhythmias that may not have been fatal in the absence of an ICD, based on inferences from randomized defibrillator trial data in patients with coronary artery disease.53


    Although the present data do not resolve all remaining questions in the arena of risk stratification for the HCM patient population, the capability of contrast-enhanced CMR to identify extensive LGE advances the risk stratification strategy in this disease by providing the opportunity to potentially recognize additional patients at increased risk for SCD events. Conversely, the absence of LGE was associated with lower risk of SCD events. In addition, extensive LGE was predictive of adverse LV remodeling with systolic dysfunction (end-stage HCM) and therefore proved to be associated with 2 diverse consequences of HCM.


    Guest Editor for this article was Judith S. Hochman, MD.

    The online-only Data Supplement is available with this article at

    Correspondence to Martin S. Maron, MD, 800 Washington St, No. 70, Boston, MA 02111. E-mail


    • 1. Teare D. Asymmetrical hypertrophy of the heart in young adults.Br Heart J. 1958; 20:1–8.CrossrefMedlineGoogle Scholar
    • 2. Maron BJ, McKenna WJ, Danielson GK, Kappenberger LJ, Kuhn HJ, Seidman CE, Shah PM, Spencer WH, Spirito P, Ten Cate FJ, Wigle ED. American College of Cardiology/European Society of Cardiology clinical expert consensus document on hypertrophic cardiomyopathy: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines.J Am Coll Cardiol. 2003; 42:1687–1713.CrossrefMedlineGoogle Scholar
    • 3. Maron BJ. Contemporary insights and strategies for risk stratification and prevention of sudden death in hypertrophic cardiomyopathy.Circulation. 2010; 121:445–456.LinkGoogle Scholar
    • 4. Gersh BJ, Maron BJ, Bonow RO, Dearani JA, Fifer MA, Link MS, Naidu SS, Nishimura RA, Ommen SR, Rawkowski H, Seidman CE, Towbin JA, Udelson JE, Yancy CW. 2011 ACCF/AHA guidelines for the diagnosis and treatment of hypertrophic cardiomyopathy: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.J Am Coll Cardiol/Circulation. 2011; 58:212–260.CrossrefMedlineGoogle Scholar
    • 5. Elliott PM, Poloniecki J, Dickie S, Sharma S, Monserrat L, Varnava A, Mahon NG, McKenna WJ. Sudden death in hypertrophic cardiomyopathy: identification of high risk patients.J Am Coll Cardiol. 2000; 36:2212–2218.CrossrefMedlineGoogle Scholar
    • 6. Maron BJ, Maron MS. Hypertrophic cardiomyopathy.Lancet. 2013; 381:242–255.CrossrefMedlineGoogle Scholar
    • 7. Maron BJ, Spirito P, Shen WK, Haas TS, Formisano F, Link MS, Epstein AE, Almquist AK, Daubert JP, Lawrenz T, Boriani G, Estes NA, Favale S, Piccininno M, Winters SL, Santini M, Betocchi S, Arribas F, Sherrid MV, Buja G, Semsarian C, Bruzzi P. Implantable cardioverter-defibrillators and prevention of sudden cardiac death in hypertrophic cardiomyopathy.JAMA. 2007; 298:405–412.CrossrefMedlineGoogle Scholar
    • 8. Maron BJ, Maron MS, Lesser JR, Hauser RG, Haas TS, Harrigan CJ, Appelbaum E, Main ML, Roberts WC. Sudden cardiac arrest in hypertrophic cardiomyopathy in the absence of conventional criteria for high risk status.Am J Cardiol. 2008; 101:544–547.CrossrefMedlineGoogle Scholar
    • 9. Bongioanni S, Spirito P, Masi AS, Chiribiri A, Bonamini R, Conte MR. Extensive myocardial fibrosis in a patient with hypertrophic cardiomyopathy and ventricular tachycardia without traditional high-risk features.Circ Cardiovasc Imaging. 2009; 2:349–350.LinkGoogle Scholar
    • 10. Maron BJ, Shen WK, Link MS, Epstein AE, Almquist AK, Daubert JP, Bardy GH, Favale S, Rea RF, Boriani G, Estes NA, Spirito P. Efficacy of implantable cardioverter-defibrillators for the prevention of sudden death in patients with hypertrophic cardiomyopathy.N Engl J Med. 2000; 342:365–373.CrossrefMedlineGoogle Scholar
    • 11. Jayatilleke I, Doolan A, Ingles J, McGuire M, Booth V, Richmond DR, Semsarian C. Long-term follow-up of implantable cardioverter defibrillator therapy for hypertrophic cardiomyopathy.Am J Cardiol. 2004; 93:1192–1194.CrossrefMedlineGoogle Scholar
    • 12. Maron BJ, Spirito P, Ackerman MJ, Casey SA, Semsarian C, Estes NA, Shannon KM, Ashley EA, Day SM, Pacileo G, Formisano F, Devoto E, Anastasakis A, Bos JM, Woo A, Autore C, Pass RH, Boriani G, Garberich RF, Almquist AK, Russell MW, Boni L, Berger S, Maron MS, Link MS. Prevention of sudden cardiac death with implantable cardioverter-defibrillators in children and adolescents with hypertrophic cardiomyopathy.J Am Coll Cardiol. 2013; 61:1527–1535.CrossrefMedlineGoogle Scholar
    • 13. Monserrat L, Elliott PM, Gimeno JR, Sharma S, Penas-Lado M, McKenna WJ. Non-sustained ventricular tachycardia in hypertrophic cardiomyopathy: an independent marker of sudden death risk in young patients.J Am Coll Cardiol. 2003; 42:873–879.CrossrefMedlineGoogle Scholar
    • 14. Adabag AS, Casey SA, Kuskowski MA, Zenovich AG, Maron BJ. Spectrum and prognostic significance of arrhythmias on ambulatory Holter electrocardiogram in hypertrophic cardiomyopathy.J Am Coll Cardiol. 2005; 45:697–704.CrossrefMedlineGoogle Scholar
    • 15. Shirani J, Pick R, Roberts WC, Maron BJ. Morphology and significance of the left ventricular collagen network in young patients with hypertrophic cardiomyopathy and sudden cardiac death.J Am Coll Cardiol. 2000; 35:36–44.CrossrefMedlineGoogle Scholar
    • 16. Basso C, Thiene G, Corrado D, Buja G, Melacini P, Nava A. Hypertrophic cardiomyopathy and sudden death in the young: pathologic evidence of myocardial ischemia.Hum Pathol. 2000; 31:988–998.CrossrefMedlineGoogle Scholar
    • 17. Pennell DJ. Cardiovascular magnetic resonance.Circulation. 2010; 121:692–705.LinkGoogle Scholar
    • 18. Rochitte CE, Tassi EM, Shiozaki AA. The emerging role of MRI in the diagnosis and management of cardiomyopathies.Curr Cardiol Rep. 2006; 8:44–52.CrossrefMedlineGoogle Scholar
    • 19. Moon JC, McKenna WJ. The emerging role of cardiovascular magnetic resonance in refining the diagnosis of hypertrophic cardiomyopathy.Nat Clin Pract Cardiovasc Med. 2009; 6:166–167.MedlineGoogle Scholar
    • 20. Alpendurada F, O’Hanlon R, Prasad SK. Cardiovascular magnetic resonance of cardiomyopathies.Curr Cardiol Rep. 2009; 11:61–69.CrossrefMedlineGoogle Scholar
    • 21. Moon JC, Reed E, Sheppard MN, Elkington AG, Ho SY, Burke M, Petrou M, Pennell DJ. The histologic basis of late gadolinium enhancement cardiovascular magnetic resonance in hypertrophic cardiomyopathy.J Am Coll Cardiol. 2004; 43:2260–2264.CrossrefMedlineGoogle Scholar
    • 22. Kim RJ, Judd RM. Gadolinium-enhanced magnetic resonance imaging in hypertrophic cardiomyopathy: in vivo imaging of the pathologic substrate for premature cardiac death?J Am Coll Cardiol. 2003; 41:1568–1572.CrossrefMedlineGoogle Scholar
    • 23. Kwon DH, Smedira NG, Rodriguez ER, Tan C, Setser R, Thamilarasan M, Lytle BW, Lever HM, Desai MY. Cardiac magnetic resonance detection of myocardial scarring in hypertrophic cardiomyopathy: correlation with histopathology and prevalence of ventricular tachycardia.J Am Coll Cardiol. 2009; 54:242–249.CrossrefMedlineGoogle Scholar
    • 24. Papavassiliu T, Schnabel P, Schröder M, Borggrefe M. CMR scarring in a patient with hypertrophic cardiomyopathy correlates well with histological findings of fibrosis.Eur Heart J. 2005; 26:2395.CrossrefMedlineGoogle Scholar
    • 25. Kwon DH, Setser RM, Popović ZB, Thamilarasan M, Sola S, Schoenhagen P, Garcia MJ, Flamm SD, Lever HM, Desai MY. Association of myocardial fibrosis, electrocardiography and ventricular tachyarrhythmia in hypertrophic cardiomyopathy: a delayed contrast enhanced MRI study.Int J Cardiovasc Imaging. 2008; 24:617–625.CrossrefMedlineGoogle Scholar
    • 26. Fluechter S, Kuschyk J, Wolpert C, Doesch C, Veltmann C, Haghi D, Schoenberg SO, Sueselbeck T, Germans T, Streitner F, Borggrefe M, Papavassiliu T. Extent of late gadolinium enhancement detected by cardiovascular magnetic resonance correlates with the inducibility of ventricular tachyarrhythmia in hypertrophic cardiomyopathy.J Cardiovasc Magn Reson. 2010; 12:30.CrossrefMedlineGoogle Scholar
    • 27. Aquaro GD, Masci P, Formisano F, Barison A, Strata E, Pingitore A, Positano V, Spirito P, Lombardi M. Usefulness of delayed enhancement by magnetic resonance imaging in hypertrophic cardiomyopathy as a marker of disease and its severity.Am J Cardiol. 2010; 105:392–397.CrossrefMedlineGoogle Scholar
    • 28. Salerno M, Kramer CM. Prognosis in hypertrophic cardiomyopathy with contrast-enhanced cardiac magnetic resonance: the future looks bright.J Am Coll Cardiol. 2010; 56:888–889.CrossrefMedlineGoogle Scholar
    • 29. Maron MS. Clinical utility of cardiovascular magnetic resonance in hypertrophic cardiomyopathy.J Cardiovasc Magn Reson. 2012; 14:13.CrossrefMedlineGoogle Scholar
    • 30. Maron MS, Appelbaum E, Harrigan CJ, Buros J, Gibson CM, Hanna C, Lesser JR, Udelson JE, Manning WJ, Maron BJ. Clinical profile and significance of delayed enhancement in hypertrophic cardiomyopathy.Circ Heart Fail. 2008; 1:184–191.LinkGoogle Scholar
    • 31. Biagini E, Lorenzini M, Olivotto I, Rocchi G, Lovato L, Lai F, Rosmini S, Pazzi C, Pasquale F, Reggiani ML, Fattori R, Rapezzi C. Effects of myocardial fibrosis assessed by MRI on dynamic left ventricular outflow tract obstruction in patients with hypertrophic cardiomyopathy: a retrospective database analysis.BMJ Open. 2012; 2:e001267CrossrefMedlineGoogle Scholar
    • 32. Maron BJ. Hypertrophic cardiomyopathy: a systematic review.JAMA. 2002; 287:1308–1320.CrossrefMedlineGoogle Scholar
    • 33. Maron BJ, Olivotto I, Spirito P, Casey SA, Bellone P, Gohman TE, Graham KJ, Burton DA, Cecchi F. Epidemiology of hypertrophic cardiomyopathy-related death: revisited in a large non-referral-based patient population.Circulation. 2000; 102:858–864.LinkGoogle Scholar
    • 34. Harris KM, Spirito P, Maron MS, Zenovich AG, Formisano F, Lesser JR, Mackey-Bojack S, Manning WJ, Udelson JE, Maron BJ. Prevalence, clinical profile, and significance of left ventricular remodeling in the end-stage phase of hypertrophic cardiomyopathy.Circulation. 2006; 114:216–225.LinkGoogle Scholar
    • 35. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS; American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association.Circulation. 2002; 105:539–542.LinkGoogle Scholar
    • 36. Harrigan CJ, Peters DC, Gibson CM, Maron BJ, Manning WJ, Maron MS, Appelbaum E. Hypertrophic cardiomyopathy: quantification of late gadolinium enhancement with contrast-enhanced cardiovascular MR imaging.Radiology. 2011; 258:128–133.CrossrefMedlineGoogle Scholar
    • 37. Maron MS. Contrast-enhanced CMR in HCM: what lies behind the bright light of LGE and why it now matters.JACC Cardiovasc Imaging. 2013; 6:597–599.CrossrefMedlineGoogle Scholar
    • 38. Moravsky G, Ofek E, Rakowski H, Butany J, Williams L, Ralph-Edwards A, Wintersperger BJ, Crean A. Myocardial fibrosis in hypertrophic cardiomyopathy: accurate reflection of histopathological findings by CMR.JACC Cardiovasc Imaging. 2013; 6:587–596.CrossrefMedlineGoogle Scholar
    • 39. Pencina MJ, D’Agostino RB, Pencina KM, Janssens AC, Greenland P. Interpreting incremental value of markers added to risk prediction models.Am J Epidemiol. 2012; 176:473–481.CrossrefMedlineGoogle Scholar
    • 40. Pencina MJ, D’Agostino RB, D’Agostino RB, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond.Stat Med. 2008; 27:157–172.CrossrefMedlineGoogle Scholar
    • 41. Harrell FE, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors.Stat Med. 1996; 15:361–387.CrossrefMedlineGoogle Scholar
    • 42. Maron MS, Maron BJ, Harrigan C, Buros J, Gibson CM, Olivotto I, Biller L, Lesser JR, Udelson JE, Manning WJ, Appelbaum E. Hypertrophic cardiomyopathy phenotype revisited after 50 years with cardiovascular magnetic resonance.J Am Coll Cardiol. 2009; 54:220–228.CrossrefMedlineGoogle Scholar
    • 43. O’Hanlon R, Grasso A, Roughton M, Smith G, Alpendurada FD, Wong J, Dahl A, Oxborough D, Godfrey R, Sharma S, Roughton M, George K, Pennell DJ, Whyte G, Prasad SK. Prognostic significance of myocardial fibrosis in hypertrophic cardiomyopathy.J Am Coll Cardiol. 2010; 56:867–874.CrossrefMedlineGoogle Scholar
    • 44. Bruder O, Wagner A, Jensen CJ, Schneider S, Ong P, Kispert EM, Nassenstein K, Schlosser T, Sabin GV, Sechtem U, Mahrholdt H. Myocardial scar visualized by cardiovascular magnetic resonance imaging predicts major adverse events in patients with hypertrophic cardiomyopathy.J Am Coll Cardiol. 2010; 56:875–887.CrossrefMedlineGoogle Scholar
    • 45. Rubinshtein R, Glockner JF, Ommen SR, Araoz PA, Ackerman MJ, Sorajja P, Bos JM, Tajik AJ, Valeti US, Nishimura RA, Gersh BJ. Characteristics and clinical significance of late gadolinium enhancement by contrast-enhanced magnetic resonance imaging in patients with hypertrophic cardiomyopathy.Circ Heart Fail. 2010; 3:51–58.LinkGoogle Scholar
    • 46. Green JJ, Berger JS, Kramer CM, Salerno M. Prognostic value of late gadolinium enhancement in clinical outcomes for hypertrophic cardiomyopathy.JACC Cardiovasc Imaging. 2012; 5:370–377.CrossrefMedlineGoogle Scholar
    • 47. Moon JC, McKenna WJ, McCrohon JA, Elliott PM, Smith GC, Pennell DJ. Toward clinical risk assessment in hypertrophic cardiomyopathy with gadolinium cardiovascular magnetic resonance.J Am Coll Cardiol. 2003; 41:1561–1567.CrossrefMedlineGoogle Scholar
    • 48. Choudhury L, Mahrholdt H, Wagner A, Choi KM, Elliott MD, Klocke FJ, Bonow RO, Judd RM, Kim RJ. Myocardial scarring in asymptomatic or mildly symptomatic patients with hypertrophic cardiomyopathy.J Am Coll Cardiol. 2002; 40:2156–2164.CrossrefMedlineGoogle Scholar
    • 49. Hauser RG, Maron BJ. Lessons from the failure and recall of an implantable cardioverter-defibrillator.Circulation. 2005; 112:2040–2042.LinkGoogle Scholar
    • 50. O’Mahony C, Jichi F, Pavlou M, Monserrat C, Anastasakis A, Rapezzi C, Biagini E, Gimeno JR, Limongelli G, McKenna WJ, Omar RZ, Elliott PM. A novel clinical risk prediction model for sudden cardiac death in hypertrophic cardiomyopathy (HCM Risk-SCD) [published online ahead of print October 14, 2013].Eur Heart J. doi: 10.1093/eurheartj/eht439. Accessed September 26, 2013.CrossrefGoogle Scholar
    • 51. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA. Assessment of myocardial fibrosis with cardiovascular magnetic resonance.J Am Coll Cardiol. 2011; 57:891–903.CrossrefMedlineGoogle Scholar
    • 52. Olivotto I, Maron BJ, Appelbaum E, Harrigan CJ, Salton C, Gibson CM, Udelson JE, O’Donnell C, Lesser JR, Manning WJ, Maron MS. Spectrum and clinical significance of systolic function and myocardial fibrosis assessed by cardiovascular magnetic resonance in hypertrophic cardiomyopathy.Am J Cardiol. 2010; 106:261–267.CrossrefMedlineGoogle Scholar
    • 53. Tung R, Zimetbaum P, Josephson ME. A critical appraisal of implantable cardioverter-defibrillator therapy for the prevention of sudden cardiac death.J Am Coll Cardiol. 2008; 52:1111–1121.CrossrefMedlineGoogle Scholar


    Hypertrophic cardiomyopathy is the most common cause of sudden cardiac death (SCD) in young patients, although identifying all at-risk patients remains challenging. Recently, contrast-enhanced cardiovascular magnetic resonance with late gadolinium enhancement (LGE) has emerged as an important imaging marker of myocardial fibrosis, a structural nidus for the generation of potentially lethal ventricular tachyarrhythmias. Therefore, we assessed whether extent of LGE provided additional prognostic information in assessing cardiovascular outcome among a large cohort of 1293 patients with hypertrophic cardiomyopathy eligible for cardiovascular magnetic resonance. Amount of LGE was associated with an increased risk of SCD events (including appropriate defibrillator interventions), with every 10% increase in LGE associated with a 40% increase in relative risk of SCD events, even after adjustment for relevant clinical variables, including the conventional sudden death risk factors. In addition, among the subgroup of patients with hypertrophic cardiomyopathy judged otherwise to be at lower risk, extensive LGE (≥15% of the left ventricular mass) identified a 2-fold greater risk of SCD events with an estimated likelihood of SCD events of 6% at 5 years. In addition, the absence of LGE was associated with a lower risk of SCD events (adjusted hazard ratio, 0.39; P=0.02). These findings demonstrate that extensive LGE is a novel imaging marker that may identify patients with hypertrophic cardiomyopathy at increased risk for SCD events who otherwise would be not be considered high risk on the basis of the conventional risk stratification strategy and who may become candidates for primary prevention of SCD with an implantable defibrillator. The absence of LGE is associated with low risk, providing a measure of reassurance for patients.


    eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

    Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.