Author + information
- Received January 4, 2018
- Revision received March 11, 2018
- Accepted March 30, 2018
- Published online May 16, 2018.
- Sabrina Nordin, MBBSa,b,
- Rebecca Kozor, PhDc,
- Katia Medina-Menacho, MDa,b,
- Amna Abdel-Gadir, MBBSa,b,
- Shanat Baig, MBBSd,
- Daniel M. Sado, MDe,
- Ilaria Lobascio, MDa,
- Elaine Murphy, MBBChf,
- Robin H. Lachmann, MD, PhDf,
- Atul Mehta, PhDg,
- Nicola C. Edwards, PhDd,
- Uma Ramaswami, MDg,
- Richard P. Steeds, MDd,
- Derralynn Hughes, PhDg and
- James C. Moon, MDa,b,∗ ()
- aCardiology Department, Barts Heart Centre, London, United Kingdom
- bInstitute of Cardiovascular Science, University College London, London, United Kingdom
- cSydney Medical School, University of Sydney, Sydney, Australia
- dCardiology Department, University Hospitals Birmingham, Birmingham, United Kingdom
- eKing’s College London, London, United Kingdom
- fCharles Dent Metabolic Unit, National Hospital for Neurology and Neurosurgery, London, United Kingdom
- gLysosomal Storage Disorder Unit, Royal Free Hospital, London, United Kingdom
- ↵∗Address for correspondence:
Professor James Moon, Barts Heart Centre, St Bartholomew’s Hospital, West Smithfield, London EC1A 7BE, United Kingdom.
Objectives The authors sought to explore the Fabry myocardium in relation to storage, age, sex, structure, function, electrocardiogram changes, blood biomarkers, and inflammation/fibrosis.
Background Fabry disease (FD) is a rare, x-linked lysosomal storage disorder. Mortality is mainly cardiovascular with men exhibiting cardiac symptoms earlier than women. By cardiovascular magnetic resonance, native T1 is low in FD because of sphingolipid accumulation.
Methods A prospective, observational study of 182 FD (167 adults, 15 children; mean age 42 ± 17 years, 37% male) who underwent cardiovascular magnetic resonance including native T1, late gadolinium enhancement (LGE), and extracellular volume fraction, 12-lead electrocardiogram, and blood biomarkers (troponin and N-terminal pro-brain natriuretic peptide).
Results In children, T1 was never below the normal range, but was lower with age (9 ms/year, r = −0.78 children; r = −0.41 whole cohort; both p < 0.001). Over the whole cohort, the T1 reduction with age was greater and more marked in men (men: −1.9 ms/year, r = −0.51, p < 0.001; women: −1.4 ms/year, r = −0.47 women, p < 0.001). Left ventricular hypertrophy (LVH), LGE, and electrocardiogram abnormalities occur earlier in men. Once LVH occurs, T1 demonstrates major sex dimorphism: with increasing LVH in women, T1 and LVH become uncorrelated (r = −0.239, p = 0.196) but in men, the correlation reverses and T1 increases (toward normal) with LVH (r = 0.631, p < 0.001), a U-shaped relationship of T1 to indexed left ventricular mass in men.
Conclusions These data suggest that myocyte storage starts in childhood and accumulates faster in men before triggering 2 processes: a sex-independent scar/inflammation regional response (LGE) and, in men, apparent myocyte hypertrophy diluting the T1 lowering of sphingolipid.
Fabry disease (FD) is a rare, x-linked lysosomal storage disorder caused by deficiency in the enzyme α-galactosidase A. This leads to slowly progressive sphingolipid accumulation affecting multiple organs including the heart, kidneys, and brain (1). Men are affected earlier than women (average age of cardiac symptoms: 32 vs. 40 years) (2,3). Cardiac manifestations include left ventricular hypertrophy (LVH), arrhythmias, chronic inflammation (4), myocardial fibrosis, and functional impairment (5). Cardiovascular death is the leading cause of death in both men and women (6,7), with LVH and myocardial fibrosis being among the suggested risk factors for ventricular arrhythmia and sudden cardiac death (8). Treatment options in FD include enzyme replacement therapy (ERT), which has been the mainstay therapy, and more recently oral chaperone therapy, which is available for amenable mutations (9). Sphingolipid storage over time appears to trigger myocardial processes including LVH and irreversible myocardial fibrosis. Although specific phases of cardiac involvement are ill defined, early initiation of treatment appears desirable to avoid irreversible and progressive phenotype alterations (10).
The cardiac phenotype in FD has been increasingly understood by cardiac magnetic resonance (CMR) over the past 15 years. Early data showed the presence of late gadolinium enhancement (LGE) in approximately 50% of patients, initially in the basal inferolateral wall in FD (11), sometimes occurring before LVH in women (12) or in some mutations in men (13). Recent advances in CMR allow quantitative assessment of the myocardium using fundamental magnetic tissue relaxation constants, for example T1, which are displayed as parametric color maps. Sphingolipid, or at least the specific configuration in lamellar bodies, appears to be the cause of the low T1 and occurs in 85% of FD with LVH and 40% to 50% of FD without LVH (14–16). Low T1 has also been observed in FD with right ventricle hypertrophy (17). The T1 signal, however, reflects the whole myocardium: although fat lowers T1, pseudonormalization has been previously hypothesized in advanced disease, either with triggered hypertrophy diluting the storage signal or with fibrosis (high T1) normalizing it (14). There are well-documented data on differences in Fabry cardiomyopathy between male and female FD by transthoracic echocardiography and CMR with LGE (12). We sought to refine further our understanding of myocardial disease development using a multiparametric CMR approach in a large cohort of adults and children, male and female FD.
This was a prospective observational study in 182 patients with FD. Ethical approval was obtained from the local research ethics committee. Participants were recruited from Fabry clinics at the Royal Free Hospital and National Hospital for Neurology and Neurosurgery in London. Inclusion criteria included gene-positive FD; both men and women; children (<18 years) and adults (≥18 years).
All participants underwent CMR and 12-lead electrocardiogram (ECG). Blood samples were performed on adults just before the scan and analyzed for estimated glomerular filtration rate (eGFR), high-sensitivity troponin T (Roche Diagnostic; normal range: 0 to 14 ng/l) and N-terminal pro-brain natriuretic peptide (NT-proBNP) analysis (Roche Diagnostics; normal range according to age and sex) (18).
CMR was performed (1.5 Tesla Avanto, Siemens Healthcare, Erlangen, Germany) using a standard clinical protocol. LGE images were acquired in the standard long axis and short axis stack using either fast long angle shot sequence or a respiratory motion-corrected free-breathing single short steady-state free precession, averaged sequence, both with phase-sensitive inversion recovery. Before contrast administration (0.1 mmol/kg body weight, gadoterate meglumine, Dotarem, Guerbet S.A., France), T1 mapping was performed using a shortened modified Look-Locker inversion recovery sequence on basal and mid left ventricular short axis slices. The resulting pixel-by-pixel T1 color maps were displayed using a customized 12-bit lookup table, in which normal myocardium was green, increasing T1 was red, and decreasing T1 was blue. Post-contrast T1 mapping was performed 15 min after gadolinium administration for assessment of extracellular volume fraction (ECV) quantification. Contrast was not administered in participants age <18 years, eGFR <30 ml/min/1.73 m2, or if the patient declined.
All imaging analysis was performed using CVI42 software (Circle Cardiovascular Imaging Inc., Calgary, Canada). A region of interest for native T1 and ECV was manually drawn in the septum, taking care to avoid the blood–myocardial boundary with a 20% offset. LVH was defined as maximum wall thickness (MWT) >12 mm in adults or increased indexed left ventricular mass (LVMi) on CMR according to age- and sex-matched normal reference ranges in adults and children (19,20). Normal T1 reference ranges (mean ± 2 SD) were defined using 73 adult healthy volunteers (mean age: 49 ± 14 years): normal range total population 958 ± 56 ms, lower limit 902 ms; male subgroup normal range 947 ± 46 ms, lower limit 901 ms and female subgroup normal range 972 ± 56 ms, lower limit 916 ms (21).
The ECGs were independently analyzed by 2 experienced observers (S.N. and S.B.). Recorded ECG variables included: heart rate, rhythm, PR interval duration (normal: 120 to 200 ms), and QRS complex duration (normal: <120 ms). The presence of complete left or right bundle branch block, T-wave inversion in at least 2 contiguous leads, multifocal ventricular ectopics, and Sokolow-Lyon voltage criteria for LVH (SV1 + RV5 or RV6 >35 mm) were also recorded.
Statistical analyses were performed using SPSS 24 (IBM, Armonk, New York). Continuous variables were expressed as mean ± SD or median (interquartile range) according to normality using Shapiro-Wilk test. Troponin and NT-proBNP values were natural log transformed for bivariate testing. Categorical variables were expressed as percentages. Group testing was independent sample t-test, Mann-Whitney U test, chi-square test, or Fisher exact test according to normality, categorical, or continuous data. Correlations between parameters were analyzed using Pearson (r) or Spearman’s rho (rs). Difference in regression slopes was determined using analysis of covariance. The relationship between LGE and clinically relevant variable was evaluated using logistic regression model to identify the independent associations of the dependent variables. A p value <0.05 was considered statistically significant.
There were 182 participants: 167 adults and 15 children. Baseline characteristics are shown in Table 1. Mean age of the FD cohort was 42 ± 17 years, male sex 37% (68 of 182). Mean eGFR was 82 ± 17 ml/min/1.73 m2. Eleven patients had eGFR <60 ml/min/1.73 m2. Thirty-seven patients did not receive contrast at CMR; this is mainly because of patient preference (3 patients had poor eGFR excluding contrast administration). Thirty percent (55 of 182) had the cardiac variant (22). Of this group, 91% (50 of 55) had N215S mutation, 5% (3 of 55) R301Q mutation, and 4% (2 of 55) I91T mutation (Online Table 1). A total of 51% (92 of 182) of patients were on ERT with a median ERT duration of 7.0 ± 6.4 years.
In the adult cohort (n = 167), mean age was 45 ± 15 years (range 18 to 81 years of age). Twenty-eight percent (47 of 167) were of the cardiac variant and 37% were male (62 of 167), with ERT in 54% (91 of 167). In children (n = 15), the median age was 11 ± 5 years (range 6 to 16 years of age), 53% (8 of 15) were of the cardiac variant (all N215S mutation), and 40% were male (6 of 15). Only 1 child was on ERT (for uncontrolled acroparesthesia).
Age-related trends of T1
All children were LVH-negative with normal function (mean left ventricular ejection fraction [LVEF]: 68 ± 6%; LVMi: 54 ± 12 g/m2; MWT: 7 ± 1 mm). The children did not receive gadolinium, therefore precluding assessment of LGE and ECV. No child had a low T1 (mean native T1: 971 ± 31 ms), but T1 fell linearly with increasing age (Figure 1A). Extrapolating the curve to adulthood, the curve cuts the lower limit of normal at the age of 18 (r = −0.78; p < 0.001) (Figure 1A).
Mean LVEF in FD adults was 73 ± 7%. Three adults had mildly impaired LV function with LVEF between 49% and 54%. Of the male FD with LVH, 98% (44 of 45) had low T1 (Figure 2A). The 1 patient with normal T1 had extensive LGE involving multiple segments including the septum. Of the female FD with LVH, 90% (28 of 31) had low T1 and those with normal T1 all had apical LVH.
With increasing age, T1 falls, but the trend is less strong than in children (r = −0.41; p < 0.001) (Figure 1B). Over the whole cohort, this was more marked in men (−1.9 ms/year, r = −0.51 men, p < 0.001; women −1.4 ms/year, r = −0.47 women, p < 0.001) (Figure 1B). When T1 is compared with LVH, however, the relationship was sex specific. Before LVH is present, with increasing LVMi, T1 falls in both sexes but more markedly in men (r = −0.54, p < 0.001 in men; r = −0.276, p = 0.01 in women) (Figure 2). After overt LVH is present, a major sex dimorphism is found. In men, the LVH is more extreme, and the correlation reverses (r = 0.631; p < 0.001), although almost all are below the lower limit of normal (pseudonormalization). In women, once LVH is present, T1 is not related to the degree of LVH (r = −0.239; p > 0.05).
Age- and sex-related trends for all parameters
LVH was more prevalent in men (66% [45 of 68] compared with 27% [31 of 114] in women; p < 0.001). In both sexes, the prevalence of LVH increased with age (Figure 3A: male; Figure 3B: female). There were 2 main differences between men and women: first, phenotype development of LVH was later in women (effectively complete penetrance by 40 to 49 years in men and 60 to 69 years in women). Second, when present, the LVH was more severe in men (Figure 2), with LVMi in men ranging up to 2.6× the average LVMi (based on normal reference ranges) and 1.7× the average LVMi for women. The highest MWT in men was 30 mm and was 26 mm in women.
LGE prevalence was higher in men than in women (59% [32 of 54] vs. 37% [34 of 91]; p = 0.015). Prevalence of LGE increased with age in both men and women (Figures 3C and 3D). Given that LVH mainly occurs later in women, this meant a much higher frequency of LGE in LVH-negative women. The only true LVH-negative male had LGE at the RV insertion points, a nonspecific pattern. The other 2 men had high normal LVMi with an MWT of 13 mm.
Male FD had lower ECV compared with female FD (ECV: 0.25 ± 0.03 vs. 0.28 ± 0.02; p < 0.001), as is found in healthy subjects. The correlation observed in ECV between age and sex, however, is stronger in men compared with women (r = 0.38 in men, p = 0.04; r = 0.12 in women, p > 0.05).
Blood biomarkers (troponin and NT-proBNP)
Troponin increased with age in men and women (Figures 4A and 4B). Troponin correlated with age (rs = 0.58; p < 0.001) as well as LVMi (rs = 0.69), T1 (rs = −0.51), and LGE (LGE-positive: 83% [34 of 41] vs. LGE-negative: 4% [2 of 57]); all p < 0.001.
NT-proBNP increased with age (Figures 4C and 4D). NT-proBNP correlated with age (rs = 0.47; p < 0.001) and was associated with LVMi rs = 0.58, T1 rs = −0.34, and LGE (LGE-positive: 68% [28 of 41] vs. LGE-negative: 9% [5 of 58]); all p < 0.001. Troponin and native T1 values were independently related to LGE, with troponin having the strongest association with presence of LGE (odds ratio: 1.191; p = 0.008). The odds ratio of NT-proBNP, LVMi, eGFR, sex, and age did not reach statistical significance.
The prevalence of ECG abnormalities increased with age in men and women (Figures 4E and 4F), occurring earlier in men (age 18 to 19 years) than in women (age 20 to 29 years).
In general, patients on ERT had higher MWT (14 ± 5 mm vs. 10 ± 3 mm) and LVMi (98 ± 42 g/m2 vs. 69 ± 26 g/m2) and lower septal T1 value (876 ± 43 ms vs. 916 ± 60 ms); all p < 0.001. ERT use was increasingly common with age in both men and women, but the age group of ERT initiation was similar between sexes (10 to 19 years). In men, 78% (36 of 46) with LVH were on ERT; 74% (23 of 31) of women were on ERT; both p < 0.005. There is no significant correlation between T1 value and ERT duration in the total cohort (rs = 0.004; p = NS). No strong correlation was found between ERT duration and T1 values in men and women (rs = 0.290 male and −0.029 female; both p = NS). There is a downward trend between age of ERT initiation and T1 values (rs = −0.30 male and −0.44 female; both p < 0.05). The trend of LVMi and T1 between men and women were similar in patients on ERT compared with patients who are ERT-naïve (Figure 5) and in patients with cardiac variant compared with classical variant (Online Figure 1).
We sought insight into myocardial phenotype development in FD by looking at a large cohort of patients (including children) and measuring multiple parameters: LVH, scar (LGE), blood biomarkers, and, importantly, T1, a quantitative myocardial signal that is reduced by sphingolipid storage. The data show that no child had overt T1 lowering of storage, but that T1 falls through childhood, suggesting progressive subclinical accumulation. In children and adults, the fall in T1 with age is steeper with men compared with women, suggesting storage is faster in men. Men had earlier ECG abnormalities, blood biomarker increase, LVH, and slightly earlier LGE; with LGE in LVH-negative subjects only occurring in women. Once LVH occurs, male hypertrophy is far more extreme than in women (even when indexed), and the relationship of LVH to T1 changes: in women, T1 falls until LVH is present, when it is broadly flat, but in men, after LVH is present, T1 is higher (more normal) with increased LVH. What does the apparent T1 rise in men mean? Storage cannot be the cause (or T1 would be falling). T1 is a composite signal from myocardial interstitium (fibrosis, edema, high T1), capillary blood (high T1), myocyte sarcomeric protein (presumed normal T1), and sphingolipid storage (low T1). With this framework, there are multiple possible explanations: diffuse fibrosis, edema, or capillary vasodilatation pseudo-normalizing T1 appear unlikely because all of these increase ECV, which was normal in this study; focal fibrosis would explain the 1 patient with extensive LGE and normal T1, but not the others; removal of storage by ERT is a plausible scenario, but first, this would likely cause at least some LV mass regression, and second, given that most men and women with LVH are on ERT, the enzyme would need to be more efficient in men, which appears unlikely. We believe the answer is that, in men, storage is triggering sarcomeric protein expression via myocyte hypertrophy (23,24) and usual LVH rather than storage LVH, which are diluting the T1 lowering of sphingolipid. This would mean that there are 2 types of sex dimorphism in FD: that affecting all tissues (related to the second functional copy of the alpha galactosidase gene in women) and a cardiac-specific sex dimorphism related to the male myocyte response to insult, which here is storage. Sex dimorphism in myocyte hypertrophy is familiar from other cardiac diseases: the hypertrophy in aortic stenosis is far more marked in men than women (25), and probably observed in other conditions (e.g., hypertrophic cardiomyopathy, a non–sex-linked disease that has a male predominance in most large studies) (26). Whether FD, AS, and HCM hypertrophy of sex dimorphism derive from common mechanisms is unknown at this time. Differential expression of androgen and estrogen receptors and differences in the renin-angiotensin system, nitric oxide activity, and norepinephrine release may contribute to sex differences observed in LV remodeling (25,27,28).
Based on these findings, our prior work on LVH-negative patients and on LGE including inflammation imaging (4), we hypothesize a model of myocardial phenotype evolution in FD consisting of an accumulation phase, a hypertrophy and inflammation phase, and a fibrosis and impairment (late) phase (Figure 6). This later phase is underrepresented in our study, partly because these patients have devices (e.g., permanent pacemaker or implantable cardioverter defibrillator) or have a significant mortality rate.
The silent storage phase starts in childhood and is subclinical. Myocardial T1 is normal but falling. Minor architectural changes in cardiac morphology may be present. Overt storage phase indicates T1 is now low and progressing faster in men than women and is associated with LV mass within normal limits and ECG changes.
Myocyte hypertrophy and inflammation phase
LGE and inflammation appears mainly in the basal inferolateral wall and is associated with persistent chronic troponin elevation but no thinning. This may occur before LVH in women (and Taiwan IVS4 subjects) (13).
LVH, demonstrating sex dimorphism appears in women and consists of likely balanced sphingolipid and myocyte hypertrophy in proportion. In men, this consists mainly of increasing myocyte contractile protein, a true hypertrophy with the T1 fall becoming less prominent.
Fibrosis and impairment phase
Persistent LVH and troponin elevation is present, but now fibrosis (myocyte death) and thinning occur. LGE can be found extensively outside the basal inferolateral wall together with NT-proBNP elevation, LV impairment, and clinical heart failure.
Our group and others have shown that ECV is normal in FD (15,29). The difference in ECV between men and women is most likely explained by normal sex differences where ECV in FD (15) and in healthy controls is known to be higher in women than in men (15,29,30), with more change over time observed in men, which could be related to higher prevalence of myocyte hypertrophy or myocardial fibrosis occurring in men with increasing age (Figure 3). We have previously shown that LGE in established FD is chronic inflammation strongly correlating with troponin levels (4), supporting findings by positron emission tomography/magnetic resonance imaging study (31) and endomyocardial biopsy (32). Here again, troponin was independently and strongly related to LGE, although we do not report T2 values in this paper.
Limitations of this study include no histological validation of the presence of storage with low native T1 in this study. The shape (linear/nonlinear) of the relationship of sphingolipid to storage is unknown. There were no controls for the cohort of children although we note that T1 mapping in healthy controls in children might be comparative to healthy controls value in adults with no age effect in T1 values in the pediatric group (30,33). In addition, this is also a single center study with single time point data, so our findings are hypothesis generating only based on our experiences and knowledge collaborated from all of our published works. We acknowledge that further longitudinal studies are needed and are currently being undertaken.
Sphingolipid accumulation potentially starts in childhood, proceeds more markedly in men than women before triggering a sex-independent scar or inflammation response but a sexually dimorphic myocyte hypertrophic response in men. Multi-time point long-term follow-up studies are needed to explore this further.
COMPETENCY IN MEDICAL KNOWLEDGE: Fabry disease is an x-linked storage disease and the leading cause of death is cardiomyopathy. Storage can now be detected by measuring T1. This study proposes phases of cardiac involvement, with storage apparently starting in childhood and accumulating faster in men before triggering new processes: regional scar and inflammation (men and women), and apparent myocyte hypertrophy (more in men) before advanced disease (extensive scar and LV impairment).
TRANSLATIONAL OUTLOOK: Fabry disease is currently treated with enzyme replacement therapy. Treating cardiac disease will be key to improving outcomes. This paper suggests cardiac disease starts early, in childhood, and has different stages and activated pathways. Additional pathways to storage that could be therapeutic targets include myocardial inflammation and, particularly in men, myocyte hypertrophy.
We are grateful for the contributions of patients, and the staff members at the Lysosomal Storage Disorder Unit, Royal Free Hospital and Charles Dent Metabolic Unit.
This research is supported by researchers at the National Institute for Health Research University College London Hospitals Biomedical Research Centre. This study is funded by an investigator-led research grant from Genzyme, now Sanofi-Genzyme. Sanofi-Genzyme had no role in the study beyond the initial funding and requiring recruitment milestones as part of financial governance. Dr. Nordin is supported by investigator-led research grant by Sanofi-Genzyme; and has received honoraria from Shire. Dr. Kozor has received honoraria for presenting and advisory board participation as well as support for investigator-led research from Sanofi-Genzyme. Dr. Murphy has received unrestricted educational grants from Sanofi-Genzyme and Shire; and clinical trial funding from Sanofi-Genzyme, Shire, Amicus, and Biomarin. Dr. Lachmann has received honoraria and travel support from Sanofi-Genzyme. Dr. Mehta has received honoraria and grant support for research and educational activities from Shire, Sanofi-Genzyme, and Amicus. Dr. Moon is the principal investigator of the Sanofi-Genzyme research grant that funded this study. All other authors have reported that they have no disclosures relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac magnetic resonance
- extracellular volume fraction
- estimated glomerular filtration rate
- enzyme replacement therapy
- Fabry disease
- late gadolinium enhancement
- left ventricular ejection fraction
- left ventricular hypertrophy
- indexed left ventricular mass
- maximum wall thickness
- N-terminal pro-brain natriuretic peptide
- Received January 4, 2018.
- Revision received March 11, 2018.
- Accepted March 30, 2018.
- 2018 American College of Cardiology Foundation
- Nordin S.,
- Kozor R.,
- Bulluck H.,
- et al.
- Patel M.R.,
- Cecchi F.,
- Cizmarik M.,
- et al.
- Baig S.,
- Edwards N.,
- Kotecha D.,
- et al.
- Germain D.P.,
- Hughes D.A.,
- Nicholls K.,
- et al.
- Niemann M.,
- Herrmann S.,
- Hu K.,
- et al.
- Hsu T.R.,
- Hung S.C.,
- Chang F.P.,
- et al.
- Sado D.M.,
- White S.K.,
- Piechnik S.K.,
- et al.
- Thompson R.B.,
- Chow K.,
- Khan A.,
- et al.
- Pagano J.,
- Chow K.,
- Khan A.,
- et al.
- Kozor R.,
- Nordin S.,
- Treibel T.A.,
- et al.
- Patel V.,
- O'Mahony C.,
- Hughes D.,
- et al.
- Barbey F.,
- Brakch N.,
- Linhart A.,
- et al.
- Treibel T.A.,
- Kozor R.,
- Fontana M.,
- et al.
- Olivotto I.,
- Maron M.S.,
- Adabag A.S.,
- et al.
- Marsh J.D.,
- Lehmann M.H.,
- Ritchie R.H.,
- Gwathmey J.K.,
- Green G.E.,
- Schiebinger R.J.
- Sado D.M.,
- Flett A.S.,
- Banypersad S.M.,
- et al.
- Nappi C.,
- Altiero M.,
- Imbriaco M.,
- et al.
- Frustaci A.,
- Verardo R.,
- Grande C.,
- et al.
- ↵Pagano J, Yim D, Lam C, Seed M, Yoo S-J, Grosse-Wortmann L. Normal native myocardial T1 and ECV values in healthy children. Paper presented at: SCMR 20th Annual Scientific Sessions; March 2, 2017; Washington DC.