Author + information
- Received June 18, 2019
- Revision received August 19, 2019
- Accepted August 23, 2019
- Published online November 4, 2019.
- Stefan Neubauer, MDa,
- Paul Kolm, PhDb,
- Carolyn Y. Ho, MDc,
- Raymond Y. Kwong, MD, MPHc,
- Milind Y. Desai, MDd,
- Sarahfaye F. Dolman, MPHb,
- Evan Appelbaum, MDe,
- Patrice Desvigne-Nickens, MDf,
- John P. DiMarco, MD, PhDg,
- Matthias G. Friedrich, MDh,
- Nancy Geller, PhDf,
- Andrew R. Harper, MBBSa,
- Petr Jarolim, PhDc,
- Michael Jerosch-Herold, PhDc,
- Dong-Yun Kim, PhDf,
- Martin S. Maron, MDi,
- Jeanette Schulz-Menger, MDj,
- Stefan K. Piechnik, PhDa,
- Kate Thomson, PhDa,
- Cheng Zhang, PhDb,
- Hugh Watkins, MD, PhDa,
- William S. Weintraub, MDb,
- Christopher M. Kramer, MDg,∗ (, )@ChrisKramerMD,
- on behalf of the HCMR Investigators
- aDivision of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, United Kingdom
- bMedStar Heart and Vascular Institute, Washington, DC
- cCardiovascular Division, Department of Medicine and Department of Radiology, Brigham and Women’s Hospital, Boston, Massachusetts
- dCardiovascular Institute, Cleveland Clinic, Cleveland, Ohio
- eDivision of Cardiology, Beth Israel Deaconess Medical Center, Boston, Massachusetts
- fNational Heart, Lung, and Blood Institute, Bethesda, Maryland
- gCardiovascular Division, University of Virginia Health System, Charlottesville, Virginia
- hDepartments of Medicine and Diagnostic Radiology, McGill University, Montreal, Quebec, Canada
- iDivision of Cardiology, Tufts New England Medical Center, Boston, Massachusetts
- jCardiology Department, Charite' Experimental Clinical Research Center and Helios Clinics Berlin-Buch, Berlin, Germany
- ↵∗Address for correspondence:
Dr. Christopher M. Kramer, University of Virginia Health System, Cardiovascular Division, Department of Medicine, Lee Street, Box 800158, Charlottesville, Virginia 22908.
Background The HCMR (Hypertrophic Cardiomyopathy Registry) is a National Heart, Lung, and Blood Institute–funded, prospective registry of 2,755 patients with hypertrophic cardiomyopathy (HCM) recruited from 44 sites in 6 countries.
Objectives The authors sought to improve risk prediction in HCM by incorporating cardiac magnetic resonance (CMR), genetic, and biomarker data.
Methods Demographic and echocardiographic data were collected. Patients underwent CMR including cine imaging, late gadolinium enhancement imaging (LGE) (replacement fibrosis), and T1 mapping for measurement of extracellular volume as a measure of interstitial fibrosis. Blood was drawn for the biomarkers N-terminal pro–B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac troponin T (cTnT), and genetic analysis.
Results A total of 2,755 patients were studied. Mean age was 49 ± 11 years, 71% were male, and 17% non-white. Mean ESC (European Society of Cardiology) risk score was 2.48 ± 0.56. Eighteen percent had a resting left ventricular outflow tract (LVOT) gradient ≥30 mm Hg. Thirty-six percent had a sarcomere mutation identified, and 50% had any LGE. Sarcomere mutation–positive patients were more likely to have reverse septal curvature morphology, LGE, and no significant resting LVOT obstruction. Those that were sarcomere mutation negative were more likely to have isolated basal septal hypertrophy, less LGE, and more LVOT obstruction. Interstitial fibrosis was present in segments both with and without LGE. Serum NT-proBNP and cTnT levels correlated with increasing LGE and extracellular volume in a graded fashion.
Conclusions The HCMR population has characteristics of low-risk HCM. Ninety-three percent had no or only mild functional limitation. Baseline data separated patients broadly into 2 categories. One group was sarcomere mutation positive and more likely had reverse septal curvature morphology, more fibrosis, but less resting obstruction, whereas the other was sarcomere mutation negative and more likely had isolated basal septal hypertrophy with obstruction, but less fibrosis. Further follow-up will allow better understanding of these subgroups and development of an improved risk prediction model incorporating all these markers.
The HCMR (Hypertrophic Cardiomyopathy Registry) is a prospective National Heart, Lung, and Blood Institute–funded registry of 2,755 hypertrophic cardiomyopathy (HCM) patients recruited across Europe and North America (1). The primary goal of the study is to improve risk prediction for important adverse clinical outcomes in HCM by integrating cardiac magnetic resonance (CMR) imaging, biomarker, and genetic data with standard clinical and echocardiographic findings. Insights gained by HCMR will directly impact patient care by providing a systematic evidence base to inform and advance management guidelines (2) and develop predictive models (3). In current practice, risk stratification for sudden cardiac death (SCD) remains poorly resolved, particularly for patients at low and intermediate risk, limiting optimal use of implantable cardioverter-defibrillators (ICDs) (4,5). In addition, models have not yet been developed to predict other key adverse outcomes such as incident heart failure or atrial fibrillation.
Previous large cohorts of HCM patients were gathered retrospectively and/or from 1 or a handful of specialist centers (5–7), and in general, CMR has not been systematically included (5,7). An ongoing registry in 69 centers from 18 European countries is collecting patients with HCM (n = 1,739), but also includes other nonischemic cardiomyopathies, and is only collecting variables acquired at the discretion of the clinical sites (8). For example, only 34% of patients in the latter registry underwent CMR, 46% had genetic testing, and biomarkers were not routinely collected (8). HCMR is the first large prospective registry to include rigorous CMR imaging, genetic testing, and prospective collection of blood for biomarker analysis.
Myocardial fibrosis measured by CMR has gained attention as a potential determinant of risk in patients with HCM. The presence of substantial late gadolinium enhancement (LGE), a marker of replacement fibrosis, has been associated with a 2-fold increase in SCD risk (6) and 3-fold increase in composite events (9) if present in >15% of left ventricular (LV) mass. A meta-analysis of nearly 3,000 patients from several studies demonstrated that the presence of LGE was associated with a 3.4-fold increased risk of SCD/ICD discharge, and a 1.8-fold increase in all-cause mortality (10). The extent of LGE was also associated with an increased risk of SCD/ICD discharge (1.36/10% LGE; p = 0.005) in a continuous fashion. A recent study suggests that adding LGE to American College of Cardiology Foundation (ACCF)/American Heart Association (AHA) risk stratification, along with apical aneurysm morphology and multiple runs of nonsustained ventricular tachycardia (NSVT) improved identification of indications for ICD placement (11). Interstitial rather than replacement fibrosis may be an additional risk marker in HCM (12). HCMR is the first large multicenter study to use T1 mapping to assess extracellular volume as a surrogate for interstitial fibrosis in HCM. Integrating these markers of fibrosis and other CMR findings with clinical information, echocardiography, genotyping, and biomarker analysis may further inform risk prediction in HCM. Baseline characteristics of 2,755 patients with HCM are presented in the present paper.
The study design for HCMR has been previously published (1); the relevant methods are summarized here. HCMR is a prospective observational study. After written informed consent, all patients underwent standard clinical evaluation and CMR, and had blood drawn for genetic and biomarker analysis. Longitudinal follow-up is being conducted to determine the incidence of cardiovascular events, adjudicated by a clinical events committee.
Patients included were 18 to 65 years of age with an established diagnosis of HCM defined as unexplained LV hypertrophy (wall thickness >15 mm) without cavity dilatation or known predisposing cause (uncontrolled hypertension, aortic stenosis, and so on) (2). Patients known to have other causes of infiltrative/hypertrophic cardiomyopathies such as amyloidosis, sarcoidosis, Fabry disease, Danon disease, or Noonan’s syndrome, or discovered to have these diagnoses through HCMR genotyping, were excluded. Patients older than 65 years of age were excluded because they have high competing mortality risks, in particular from coronary artery disease and cancer.
Additional exclusion criteria were: 1) prior septal myectomy or alcohol septal ablation; 2) prior myocardial infarction or known coronary artery disease; 3) incessant ventricular arrhythmias; 4) inability to lie flat; 5) contraindication to contrast-enhanced CMR including pacemakers, defibrillators, intraocular metal, certain types of intracranial aneurysm clips, severe claustrophobia, and Stage IV/V chronic kidney disease; 6) diabetes mellitus with end organ damage; 7) ongoing pregnancy; or 8) inability to provide informed consent.
Patients were enrolled from 44 sites in the United States (n = 18), Canada (n = 4), United Kingdom (n = 13), Italy (n = 4), Germany (n = 3), and the Netherlands (n = 2) between April 2014 and April 2017 (Online Table 1). Participating sites are experienced centers with focused care of HCM patients as well as state-of-the-art CMR capabilities. Emphasis was placed on recruiting HCM patients across the risk spectrum including higher-risk patients referred for subsequent ICD insertion. Data regarding baseline demographics and clinical variables were recorded from clinical records including data from clinically performed echocardiographic, Holter and exercise testing studies closest to the time of enrollment. ESC (European Society of Cardiology) risk score was calculated using baseline clinical and echocardiographic data (3).
CMR was performed at 1.5-T or 3.0-T on MR systems from the 3 primary vendors (General Electric, Philips Medical Systems, and Siemens Healthineers) using a standardized protocol and multichannel phased-array chest coils and electrocardiographic gating. After rapid localization of the heart, short-axis cine steady-state free precession imaging (SSFP) was performed covering the whole heart in 8-mm-thick slices (no gap). Typical cine SSFP parameters included TR/TE 3.1/1.2 ms, in-plane resolution of 2 to 2.5 mm, temporal resolution of 40 to 50 ms. Baseline T1 mapping was performed in 3 short-axis slices centered in the mid-LV, representing 16 of the 17 AHA segments in the nearly 80% of the sites that had appropriate software. The Shortened Modified Look-Locker Inversion recovery technique (ShMOLLI), using a 5(1)1(1)1 Look-Locker scheme with conditional image processing (13), was used as the recommended standard on Philips and Siemens systems, both for native and post-contrast T1 mapping acquisitions. Gadolinium contrast was administered intravenously as a bolus dose of 0.15 mmol/kg. Long-axis function by SSFP cine imaging was then obtained. Post-contrast T1 mapping acquisitions were performed in the same 3 short-axis slices as pre-contrast, starting at 5, 14, and 29 min post-contrast. LGE imaging was acquired in the same long-axis and short-axis stack locations beginning at minute 17 post-contrast with a 2-dimensional breath-hold, segmented inversion-recovery sequence (inversion time [TI] optimized by the Look-Locker sequence [TI scout] to null normal myocardium). Total imaging time was approximately 60 min.
CMR image analysis
Commercially available software (MedisSuite 3.0 and QMassMR, Medis, Leiden, the Netherlands) was used for analysis of all CMR images (cine, T1 maps, and LGE) in a core laboratory. LV mass, volumes, wall thickness, and thickening was measured according to Society for Cardiovascular Magnetic Resonance standards (14). Cine images in short-axis contiguous cuts were evaluated for LV and right ventricular volumes and myocardial mass by manually tracing endocardial and epicardial borders. Papillary muscles were included in LV volumes and excluded from LV mass. Cine images were also evaluated for morphology (15) and defined as: 1) localized basal septal hypertrophy; 2) reverse curvature septal hypertrophy; 3) apical HCM; 4) concentric HCM; 5) mid-cavity obstruction with apical aneurysm; or 6) other, that is, did not fit into the preceding 5 categories (Figure 1) (16). Cine images in 2- and 4-chamber long-axis cuts were evaluated for left atrial volumes using the biplane area–length method, at end-ventricular systole, before atrial contraction, and end-ventricular diastole (17).
Quantification of LGE was performed according to Society for Cardiovascular Magnetic Resonance standards (14) using both the 6 SD quantitative threshold, as well as visually (18). LGE was categorized as none, >0% to 5%, >5% to 10%, >10% to 15%, and >15% of LV mass. T1 quantification was performed on a segmental basis by nonlinear least-squares fitting of the segmental inversion recovery curves, resulting in multiple T1 measurements (1 pre- and 3 post-contrast) calculated. Gadolinium partition coefficient λ was calculated segmentally and globally by linear regression of pre- and post-contrast R1 (=1/T1) relaxation rates in myocardium, against the corresponding R1s in the blood pool of the same short-axis slice. The linear regression slope was converted to extracellular volume (ECV) using the patient’s fractional blood volume of distribution (1 − hematocrit) (19). ECV index was calculated as ECV (%) times LV mass.
Amplicon-based sequencing for 36 cardiomyopathy-associated genes was undertaken using the Illumina MiSeq platform. Bioinformatic analysis was performed using the Genome Analysis Toolkit version 4 best practice guidelines. Variants were visually confirmed through inspection of BAM files. Variant annotation was performed using SNPEff and Ensembl’s Variant Effect Predictor (VEP version 95). Data from publicly available resources (ClinVar [version 20190211] and gnomAD r2.1) and the Oxford Regional Genetics Laboratory in-house mutation database was used to inform variant classification. Following quality control, 2,636 individuals (99.1%) were deemed suitable for subsequent genetic analyses.
Blood samples were transported on ice, processed within 60 min of phlebotomy to obtain serum and EDTA-anticoagulated plasma, aliquoted, and stored at −70°C until they were batched tested at the end of the study period in the Biomarker Research and Clinical Trials Laboratory at Brigham and Women's Hospital. Cardiac troponin T (cTnT) was tested using the Roche (Roche Diagnostics Corporation, Indianapolis, Indiana) TnT STAT Gen 5 assay to assess for myocardial injury (19,20). The analytical measurement range for the assay is 6 to 10,000 ng/l, and coefficients of variation were 4.1% at 15.6 ng/l, 4.0% at 27.6 ng/l, and 2.5% at 1,893 ng/l. NT-proBNP was measured using the Roche proBNP II assay to assess hemodynamic or myocardial wall stress. Analytical measurement range of the assay is 5 to 35,000 pg/ml, and total imprecision of the assay was 2.5% at both 138 pg/ml and 4,578 pg/ml.
Data management and statistical analysis
Clinical data were entered in an online data management system. Upon entry, data underwent a series of range and quality checks. Baseline arrhythmias were defined as a history of NSVT and/or atrial fibrillation. Summary statistics for continuous variables include mean ± SD and median (interquartile range). Most data were nonnormally distributed, so the nonparametric Kruskal-Wallis rank test was used to compare independent groups. The exceptions were age and body mass index (BMI) where Student's t-tests or analysis of variance were used. Categorical variables are summarized by number and proportion of valid (nonmissing) values and analyzed by contingency table analysis (chi-square). Odds ratios were calculated for 2 × 2 tables. Where multiple comparisons were made within tables, Bonferroni corrections were used to control Type I error rate (20). The association of morphology categories with demographic and clinical variables was assessed by contingency table analysis (chi-square) for categorical variables and 1-way analysis of variance for continuous variables. Savage Scores test was used to compare LGE distribution by morphology categories in a singly ordered (LGE) contingency table analysis (21). Statistical testing was performed with Stata, v15 (Stata Corp., College Station, Texas) and StatXact 7 (Cytel, Cambridge, Massachusetts).
The number of patients enrolled at each site is shown in Online Table 1. Of 2,762 patients initially enrolled, 1,362 were enrolled in North America and 1,400 in Europe. Seven patients were subsequently excluded because they were demonstrated to be phenocopies genetically and not have HCM, leaving 2,755 for analysis. Baseline demographic and clinical information are shown in Table 1.
Mean maximal wall thickness was 18.6 ± 4.8 mm. Eighteen percent of participants had a peak gradient >30 mm Hg, and these patients’ average gradient was 69 ± 31 mm Hg. Fifty-nine percent had mitral regurgitation, and 12% were graded as moderate or severe. Mean pulmonary artery pressure was 28 ± 11 mm Hg. Maximum left atrial dimension was 4.2 ± 0.8 cm.
Holter monitoring and exercise testing
Among 1,672 patients who had undergone clinically performed 24-h Holter monitoring, AF was seen in 4% and NSVT in 12%. The 1,520 participants who underwent clinically performed exercise treadmill testing achieved 9.7 METS on average, and 12% had a hypotensive response to exercise or failed to increase systolic blood pressure by 20 mm Hg.
CMR cine data
A total of 2,651 patients completed the CMR because 38 (1.4%) had studies aborted due to claustrophobia and 52 (2%) for other reasons. The contrast dose used was 0.15 mmol/l/kg (mean 20 ± 9 ml). The rhythm at the time of the CMR was normal sinus in 93%, atrial fibrillation in 2%, and other, that is, premature ventricular contractions, bigeminy, and so on, in 5%. LV and right ventricular structure and function results derived from SSFP cine CMR images are shown in Table 2, and examples are shown in Figure 1.
There were 2,628 studies available for morphological evaluation, with the remaining 27 incomplete for morphological assessment. A total of 1,197 (46%) had isolated basal septal hypertrophy, 1,059 (38%) reverse septal curvature, 224 (8%) apical HCM, 36 (1%) concentric HCM, 79 (3%) mid-cavity obstruction with apical aneurysm, and 33 (1%) were classified as other. Demographic and clinical characteristics associated with specific morphologies are presented in Table 3. Patients with reverse septal curvature morphology were, in general, younger, had lower BMI, more likely minority, and had thicker walls, more arrhythmias, less hypertension, and less left ventricular outflow tract (LVOT) obstruction as compared with those with isolated basal septal curvature.
Maximal LV wall thickness of any segment was 20.6 ± 4.8 mm. Results comparing maximal LV wall thickness by baseline variables are presented in Online Table 2. Significant relationships with wall thickness were found for age, BMI, male sex, LVOT gradient ≥30 mm Hg, and sarcomere mutation positive. Left atrial width from the 3-chamber long-axis view was 4.8 ± 0.8 cm. Left atrial area from the 4-chamber long-axis view was 28.9 ± 7.6 cm2.
Of 2,755 patients, 2,534 (92%) had valid LGE values to allow assessment of replacement fibrosis. LGE was present in 50% of patients based on visual criteria (Central Illustration), and in 60% based on >6 SD signal criteria. In the 50% of patients who had LGE by visual analysis, mean LGE mass was 3.7 ± 5.2% of LV mass. In patients with LGE present, ESC risk score was higher than those without LGE (2.61 ± 0.59 vs. 2.33 ± 0.49; p < 0.001). Only 2% of patients (n = 46) had LGE >15% of LV mass. Morphological correlates of LGE are shown in Table 4. A high proportion of patients with reverse septal curvature hypertrophy and apical aneurysm patterns had LGE, whereas isolated basal septal hypertrophy demonstrated LGE less frequently than other morphologies. The reverse septal curvature pattern was associated with the majority (79%) of cases with >10% LGE.
Comparison of the presence of LGE with baseline variables is shown in Table 5. BMI, family history of HCM, maximal wall thickness, reduced LV ejection fraction (LVEF), baseline arrhythmias, hypertension, and sarcomere mutation positive were all significantly associated with LGE presence. Patients with a family history of HCM were 1.2 times more likely to have LGE present than those without. Patients with LVEF <55% were nearly 1.3 times more likely to have LGE present. Patients with baseline arrhythmias were 1.4 times more likely to have LGE present, and those with a sarcomere mutation were 1.8 times more likely to have LGE present than those without.
There were 2,082 patients (76%) with analyzable native T1 and 2,013 (73%) with valid ECV measures. Mean native T1 of the entire LV myocardium was 972 ± 74 at 1.5-T and 1,170 ± 84 at 3.0-T. Native T1 in segments without LGE was 969 ± 74 at 1.5-T and 1,157 ± 86 at 3.0-T compared with 976 ± 74 at 1.5-T and 1,179 ± 81 at 3.0-T (p < 0.001 for both) in segments with LGE. There were no statistically significant differences in native T1 between the MR vendors. Pooled across field strengths, native T1 was 2% higher in women than men (p < 0.001) and showed modest statistically significant correlations with LGE and wall thickness, but not with age.
ECV was greater in regions with LGE (0.30 ± 0.05) than in those without (0.28 ± 0.04; p < 0.001). Mean ECV was greater in women (0.31 ± 0.04) than in men (0.28 ± 0.04; p < 0.001). For comparison purposes, ECV in normal volunteers ranges between 0.25 and 0.28, and tends to rise with age and be higher in females (22,23). Patients with higher ECV had a smaller BMI, were less likely to have a family history of HCM, had greater wall thickness, had more baseline arrhythmias, and were more likely to have a sarcomere mutation (Table 6). When evaluated by morphology, ECV was lowest in isolated basal septal hypertrophy compared with reverse septal curvature, apical and mid-cavity obstruction subtypes (Online Table 3). ECV index, a measure of mass of interstitium, was 49.2 ± 20.2 g in the cohort as a whole.
HCM risk factors
The mean ESC risk score (3) was 2.48 ± 0.56, suggesting that the study group is low risk. Of the enhanced ACCF/AHA risk factors (11), 12% had a family history of SCD, 13% had a history of syncope, 9% had sustained ventricular tachycardia or NSVT, 4% had wall thickness >30 mm, 2% had >15% LGE, and 3% had an apical aneurysm.
DNA samples were obtained from 2,661 individuals. Genetic analyses for genes that can be reliably interpreted in HCM comprise: the core sarcomeric genes (MYH7, MYBPC3, TNNT2, TNNI3, MYL2, MYL3, ACTC1, and TPM1) and the well-established “phenocopy” genes (GLA, PRKAG2, LAMP2, and TTR). Overall, 29.5% (n = 774) of individuals were found to have a variant classified as “pathogenic” or “likely pathogenic” in a sarcomere gene, with variants in the MYBPC3 (18.5%) and MYH7 (8.0%) genes accounting for the majority. Only 3 individuals (0.11%) demonstrated a combination of 2 likely pathogenic or pathogenic variants in confirmed sarcomere genes. Seven individuals were found to harbor pathogenic variants within either GLA (n = 4) or TTR (n = 3), indicating a diagnosis of Fabry’s disease or hereditary amyloidosis, respectively; these individuals were removed from all subsequent phenotypic analyses. In 12.3% of individuals (n = 325), “variants of uncertain significance” were detected in the sarcomere genes. See the Online Appendix for the approach to variants of uncertain significance. Using this approach, on the basis of gene-specific interpretations, we dichotomized the HCMR cohort into individuals carrying a sarcomere variant, that is, sarcomere mutation positive (n = 943; 35.8%) and those who did not, that is, sarcomere mutation negative (n = 1,693; 64.2%). Using this dichotomous criterion, just under 1% of probands carried 2 sarcomere variants (and none >2).
Those who were sarcomere mutation positive were younger, had a lower BMI, were more often female and white, had a family history of HCM, and had less hypertension (Table 7), consistent with prior findings (24). However, they also were less likely to have a significant LVOT gradient, which may, in part, reflect differences in morphology because more of the sarcomere mutation–positive group demonstrated reverse curvature asymmetric septal hypertrophy (58.1%) relative to isolated basal septal hypertrophy (33.8%), ratios that were reversed in the sarcomere mutation–negative group (30.7% and 51.8%, respectively; p < 0.0001). In addition, fewer sarcomere mutation–positive individuals demonstrated apical hypertrophy (4.5% vs. 10.7%), concentric hypertrophy (0.2% vs. 2.0%), and “other” forms of hypertrophy (0.8% vs. 1.6%). Incidence of mid-cavity obstruction with apical aneurysm was similar (2.6% vs. 3.2%). LVEF was similar between groups. Sarcomere mutation–positive patients were much more likely to have any LGE as well as more extensive LGE (Table 8). Native T1 was higher at 1.5-T in sarcomere mutation–positive individuals (978 ± 76 vs. 968 ± 74; p < 0.02), but similar at 3.0-T (1,175 ± 89 and 1,167 ± 81, respectively; p = 0.21), likely due to lower number at 3.0-T and thus lower power.
NT-proBNP and cTnT were obtained in 2,665 (97%) of the 2,755 patients in the HCMR analysis database. Online Table 4 presents comparisons of demographic and clinical variables and NT-proBNP. Because of the extreme skewness of the NT-proBNP distribution, median (interquartile range) are presented, as well as mean ± SD. Increasing age was associated with increasing NT-proBNP. Women had higher values than men as expected, obese patients had lower levels, and patients with baseline arrhythmias had higher levels. Patients with a resting LVOT gradient ≥30 mm Hg and those with a reduced LVEF had higher values. NT-proBNP levels increased as maximal wall thickness increased. The relationship between NT-proBNP and categories of LGE is presented in Figure 2. A similar relationship was seen with increasing ECV (by quartile) (Online Figure 1). NT-proBNP was significantly higher in sarcomere mutation-positive than –negative individuals (594 ± 842 vs. 520 ± 1,073; p < 0.001).
Normal values for cTnT for males were ≤22 ng/l and ≤14 ng/l for women (per Roche Diagnostics Corporation). Of the 2,665 patients with valid values, 282 men (15%) and 186 women (24%) had elevated cTnT. Online Table 5 presents comparisons of demographic and imaging data, and cTnT, abnormal versus normal. Women were over 1.6 times more likely to have abnormal cTnT levels than men. Those with a history of hypertension were 1.3 times more likely to have abnormal cTnT levels, and those with LVEF<55% were over 2.2 times more likely to have abnormal values. Minorities were 1.6 times more likely to have abnormal values, and patients with baseline arrhythmias 1.7 times more likely. As maximal wall thickness increased, so did abnormal cTnT. The relationship between cTnT and categories of LGE is presented in Figure 2. In both sexes, there was a stepwise increase in cTnT with categories of increasing LGE. A similar relationship was seen with increasing ECV (by quartile), although only in men (Online Figure 2). The incidence of elevated cTnT was similar in sarcomere mutation–positive and –negative groups (18% and 17%, respectively).
HCMR is the largest systematic, prospective natural history study in HCM to date which includes comprehensive CMR data in addition to other clinical metrics, genotyping, and biomarker analysis. Prior and ongoing registries are retrospective in nature and/or do not include systematic acquisition of these data (5–7). The 2,755 patients participating in HCMR reflect a broad sampling of North American and European sites, and 17% minority enrollment. A third had a family history of HCM, and a third had hypertension. Eighteen percent of patients had a LVOT gradient ≥30 mm Hg. Only 12% of patients had moderate or more mitral regurgitation. A sarcomere variant carrier yield of 35.8% is comparable to that seen for HCM in usual routine diagnostic service laboratories, confirming that the case population on which HCMR is based is representative of “real world” HCM practice. On the basis of the ESC risk scores, the patient group is of low risk.
The major contribution of the present study lies in the CMR, genetic, and biomarker findings and their interrelationships in this population. Two relatively distinct populations were identified in HCMR (Central Illustration). One was sarcomere mutation positive and more likely to demonstrate reverse septal curvature morphology, have more extensive LGE, but less resting LVOT obstruction. The second group was sarcomere mutation negative and more likely to demonstrate isolated basal septal hypertrophy, with less LGE, but more LVOT obstruction. The first of these groups represents the Mendelian form of familial HCM, whereas the second group presumably has multifactorial disease (25), as evidenced by the higher burden of causes of secondary LV hypertrophy (hypertension, high BMI, male sex, older age, and so on). The finding that significant resting outflow obstruction indicates a lower likelihood of the familial form of HCM was not suspected. It is also notable that apical HCM is also less likely to reflect sarcomeric HCM.
Myocardial replacement fibrosis is prevalent (50%), although the frequency of extensive LGE is less than that noted in the study by Chan et al. (6) in which the 4 sites included were highly specialized referral centers. Patients with LGE had thicker walls, more baseline arrhythmias, and were more likely to be sarcomere mutation positive, in keeping with the concept of a higher burden of LGE in clearly identified genetic disease. One prior study of 82 patients showed that the extent of LGE had an odds ratio of 2.1 to predict mutation-positive HCM (26). The ESC risk score increases modestly with any LGE compared with no LGE. Whether the presence and/or extent of LGE improves risk stratification compared to the ESC risk score remains to be determined with longer follow-up. One recent study does suggest it adds to ACCF/AHA guidelines for identification of patients who subsequently require an ICD (11).
The vast majority of patients in this study (86%) have a form of asymmetric septal hypertrophy, either isolated basal or reverse curvature. Patients with the reverse curvature form were younger, less commonly had hypertension, and were more likely sarcomere mutation positive. They represented most of the cases of >10% LGE, providing further data for a link between genetics, morphology, and fibrosis. Follow-up will test whether this morphological subtype with its link to sarcomere mutation positivity and increased fibrosis is a risk factor for outcome events. The other morphological subgroup with extensive LGE was the mid-cavity obstruction with apical aneurysm subtype, which has been shown to be associated with higher risk and adds to ACCF/AHA risk stratification (11,27).
In the HCMR cohort, there was evidence of interstitial fibrosis indicated by the elevated mean ECV when compared with prior measurements using the same techniques in normal controls (28,29). ECV was mildly elevated even in regions without LGE, suggesting that interstitial fibrosis is a characteristic of HCM. Similar to LGE, increased wall thickness, baseline arrhythmias, and sarcomere mutation positivity were associated with interstitial fibrosis. Unlike LGE, there was more interstitial fibrosis in women.
The 2 biomarkers that were measured were both elevated in subsets of patients in this cohort. NT-proBNP was elevated in subsets with resting LVOT gradient ≥30 mm Hg, reduced LVEF, more baseline arrhythmias, and sarcomere mutation positive. CTnT was higher in minorities and patients with hypertension, LVOT gradient ≥30 mm Hg, increased wall thickness, and reduced LVEF. Whether elevated biomarkers are predictive of worse outcome in HCM will only be clarified with further follow-up. This is an understudied area, especially for NT-proBNP (30). One smaller study in Japan demonstrated worse outcomes with increasing levels of troponin (31). cTnT has been shown to improve risk prediction in women, but only in the setting of coronary heart disease (30). Both biomarkers demonstrated stepwise increases in relationship to LGE extent and ECV. Because LGE extent is a marker of SCD/ICD discharge in HCM, it may be that elevated biomarkers are a synergistic risk marker with either or both replacement and interstitial fibrosis. This points to the importance of developing a multivariable model using all of these potential risk markers to predict outcome events once follow-up is long enough to allow sufficient numbers of events to occur.
This cohort excluded patients with HCM who had prior invasive septal therapy or ICD placement. Although minority recruitment was less than planned, the overall numbers may allow analysis of subpopulation differences and will likely be hypothesis-generating. The echocardiographic data were derived from clinical echo reports, and thus protocols were not standardized. Therefore, reporting of provoked obstruction was incomplete. Stress testing and Holter monitoring at entry were also not protocol-driven and thus were not performed in every patient. Although the use of ECV reduces the impact of magnetic field choice for T1 mapping, all T1 mapping techniques may be method and vendor-dependent.
The HCMR study population is characteristic of patients with low-risk HCM by ESC risk score. Ninety-three percent had no or only mild functional limitation. These patients have predominantly septal hypertrophy, 18% have resting LVOT gradient ≥30 mm Hg, and one-half have LGE. Over a third are sarcomere mutation positive. Interstitial fibrosis is prevalent even in segments without LGE. Serum biomarkers are elevated and relate to both replacement and interstitial fibrosis in a graded fashion. Two relatively distinct populations were identified. One group was sarcomere mutation positive and more likely had reverse septal curvature morphology, more fibrosis, and less resting obstruction, whereas the other was sarcomere mutation negative and more likely had isolated basal septal hypertrophy with resting obstruction and less fibrosis. Further follow-up will allow development of a model inclusive of the demographic, clinical, echocardiographic, CMR, biomarker, and genetic variables that best predict risk of major adverse cardiac events in HCM.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with hypertrophic cardiomyopathy and a sarcomere mutation associated with reversed septal curvature, more extensive myocardial fibrosis, and less resting ventricular outflow tract obstruction can be distinguished from others without this sarcomere mutation, who more often have isolated basal septal hypertrophy with less fibrosis and greater resting obstruction.
TRANSLATIONAL OUTLOOK: Further analysis of this registry and other sources should facilitate patient-specific risk profiling on the basis of demographic, echocardiographic, cardiac magnetic resonance, genetic, and biomarker data to identify those at risk of heart failure, arrhythmias, and other adverse outcomes, including mortality.
The authors acknowledge the technical assistance of Michael Bowman, PhD, Shapour Jalilzadeh, PhD, and image analysis skills of Kathleen Cheng, BA, Sebastian Estrada, BA, Hannah King, BA, and Evan Yang.
Funding by the National Heart, Lung, and Blood Institute grant U01HL117006-01A1 (to Drs. Kramer and Neubauer) and the Oxford NIHR Biomedical Research Centre (to Drs. Neubauer and Watkins). Dr. Neubauer has been a consultant for Pfizer and Cytokinetics; and has received research grants from Boehringer Ingelheim. Dr. Ho has been a consultant for and received research support from MyoKardia. Dr. Kwong has received research support from Siemens Healthineers, Bayer AG, and MyoKardia. Dr. DiMarco has been a consultant to Novartis, Celgene, and Daiichi-Sankyo. Dr. Friedrich has been a board member, shareholder, and consultant for Circle Cardiovascular Imaging. Dr. Jarolim has received research support from Abbott Laboratories, AstraZeneca, LP, Daiichi-Sankyo, GlaxoSmithKline, Merck and Co., Roche Diagnostics Corporation, Takeda Global Research and Development Center, and Waters Technologies Corporation; and has received consulting fees from Roche Diagnostics Corporation. Dr. Maron has been a consultant to iRhythm, Celltrion, and Cytokinetics; and has received research support from iRhythm. Dr. Schulz-Menger has been a consultant for Bayer; and has received research grants from Bayer, Siemens Healthineers, and Circle Cardiovascular Imaging. Dr. Piechnik has patent authorship rights for a U.S. patent held by Siemens Healthcare and managed by Oxford University Innovations. Dr. Watkins has been a consultant for Cytokinetics. Dr. Weintraub has been a consultant for Amarin, Janssen, AstraZeneca, and SC Pharma; and has received research grants from Amarin. Dr. Kramer has been a consultant for Cytokinetics; and has received research grants from Biotelemetry and MyoKardia. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. P.K. Shah, MD, served as Guest Editor-in-Chief for this paper.
- Abbreviations and Acronyms
- American College of Cardiology Foundation
- American Heart Association
- body mass index
- cardiac magnetic resonance
- high-sensitivity cardiac troponin T
- extracellular volume
- hypertrophic cardiomyopathy
- implantable cardioverter-defibrillator
- late gadolinium enhancement
- left ventricular
- left ventricular ejection fraction
- left ventricular outflow tract
- nonsustained ventricular tachycardia
- N-terminal pro–B-type natriuretic peptide
- sudden cardiac death
- steady-state free precession imaging
- Received June 18, 2019.
- Revision received August 19, 2019.
- Accepted August 23, 2019.
- 2019 American College of Cardiology Foundation
- Gersh B.J.,
- Maron B.J.,
- Bonow R.O.,
- et al.
- O’Mahony C.,
- Jichi F.,
- Ommen S.R.,
- et al.
- Chan R.H.,
- Maron B.J.,
- Olivotto I.,
- et al.
- Ho C.Y.,
- Day S.M.,
- Ashley E.A.,
- et al.
- Mentias A.,
- Raeisi-Giglou P.,
- Smedira N.G.,
- et al.
- Weng Z.,
- Yao J.,
- Chan R.H.,
- et al.
- Maron M.S.,
- Rowin E.J.,
- Wessler B.S.,
- et al.
- Ho C.Y.,
- Abbasi S.A.,
- Neilan T.G.,
- et al.
- Schulz-Menger J.,
- Bluemke D.A.,
- Bremerich J.,
- et al.
- Dodson J.A.,
- Neilan T.G.,
- Shah R.V.,
- et al.
- Appelbaum E.,
- Maron B.J.,
- Adabag S.,
- et al.
- Flett A.S.,
- Hayward M.P.,
- Ashworth M.T.,
- et al.
- Hochberg Y.,
- Tamhane A.
- Lehmann E.
- Liu C.-Y.,
- Liu Y.-C.,
- Wu C.,
- et al.
- Neilan T.G.,
- Coelho-Filho O.R.,
- Shah R.V.,
- et al.
- Ingles J.,
- Burns C.,
- Bagnall R.D.,
- et al.
- Teramoto R.,
- Fujino N.,
- Konno T.,
- et al.
- Rowin E.J.,
- Maron B.J.,
- Haas T.S.,
- et al.
- Rosmini S.,
- Bulluck H.,
- Treibel T.A.,
- et al.
- Manson J.E.,
- Bassuk S.S.
- Kubo T.,
- Kitaoka H.,
- Yamanaka S.,
- et al.