Increased Myocardial Oxygen Consumption Precedes Contractile Dysfunction in Hypertrophic Cardiomyopathy Caused by Pathogenic TNNT2 Gene Variants.

Background Hypertrophic cardiomyopathy is caused by pathogenic sarcomere gene variants. Individuals with a thin-filament variant present with milder hypertrophy than carriers of thick-filament variants, although prognosis is poorer. Herein, we defined if decreased energetic status of the heart is an early pathomechanism in TNNT2 (troponin T gene) variant carriers. Methods and Results Fourteen individuals with TNNT2 variants (genotype positive), without left ventricular hypertrophy (G+/LVH-; n=6) and with LVH (G+/LVH+; n=8) and 14 healthy controls were included. All participants underwent cardiac magnetic resonance and [11C]-acetate positron emission tomography imaging to assess LV myocardial oxygen consumption, contractile parameters and myocardial external efficiency. Cardiac efficiency was significantly reduced compared with controls in G+/LVH- and G+/LVH+. Lower myocardial external efficiency in G+/LVH- is explained by higher global and regional oxygen consumption compared with controls without changes in contractile parameters. Reduced myocardial external efficiency in G+/LVH+ is explained by the increase in LV mass and higher oxygen consumption. Septal oxygen consumption was significantly lower in G+/LVH+ compared with G+/LVH-. Although LV ejection fraction was higher in G+/LVH+, both systolic and diastolic strain parameters were lower compared with controls, which was most evident in the hypertrophied septal wall. Conclusions Using cardiac magnetic resonance and [11C]-acetate positron emission tomography imaging, we show that G+/LVH- have an initial increase in oxygen consumption preceding contractile dysfunction and cardiac hypertrophy, followed by a decline in oxygen consumption in G+/LVH+. This suggests that high oxygen consumption and reduced myocardial external efficiency characterize the early gene variant-mediated disease mechanisms that may be used for early diagnosis and development of preventive treatments.

body surface area

cardiac magnetic resonance

external work

genotype positive/left ventricular hypertrophy positive

genotype positive/left ventricular hypertrophy negative

hypertrophic cardiomyopathy

late gadolinium enhancement

left ventricle

left ventricular mass

myocardial external efficiency

myocardial oxygen consumption

myosin‐binding protein C gene

β‐myosin heavy chain gene

[11C]‐acetate positron emission tomography

systolic circumferential strain

troponin T gene

Clinical Perspective

What Is New?

This study shows reduced cardiac efficiency at preclinical and hypertrophic cardiomyopathy disease stage in individuals carrying a TNNT2 (troponin T gene) variant.

At a regional level, analysis showed significantly higher myocardial oxygen consumption in the septal and lateral left ventricular wall of gene variant carriers without left ventricular hypertrophy and the lateral wall of gene variant carriers with left ventricular hypertrophy compared with controls, indicating that the presence of a TNNT2 gene variant increases local oxygen consumption and reduces efficiency of cardiac contraction.

What Are the Clinical Implications?

This study shows that the increase in myocardial oxygen consumption in TNNT2 gene variant carriers precedes changes in global and regional myocardial contractility, indicating that the energetic status rather than contractile parameters reflects the initial variant‐induced pathomechanism that can be used for early diagnosis and preventive therapy.

Hypertrophic cardiomyopathy (HCM) is the most common inherited cardiac disease and occurs with an estimated prevalence of 1:200 in the general population.1 HCM is characterized by isolated left ventricular hypertrophy (LVH), which cannot be explained by abnormal loading conditions.2 In ≈50% to 60% of all cases, a sarcomere gene variant is identified.3 Although most variants are present in genes encoding thick‐filament proteins of the sarcomere, a subset of patients with HCM have thin‐filament gene variants.4 The most frequently affected thin‐filament gene, TNNT2 (troponin T gene), encoding cardiac troponin T,5 accounts for 2% to 5% of all HCM cases.6

Patients with HCM harboring TNNT2 gene variants present with a relatively mild form of hypertrophy compared with patients with thick‐filament gene variants, while they have a poorer prognosis.5, 7, 8, 9 A study by Coppini and colleagues showed that individuals with thin‐filament variants are characterized by more adverse remodeling and more severe diastolic dysfunction compared with patients with thick‐filament gene variants.5 Another HCM patient cohort study showed that thin‐filament gene variant carriers had a greater probability of heart failure–related death than individuals carrying thick‐filament gene variants.9 Studies in HCM rodent models and human cardiac tissue have consistently shown that myofilaments with TNNT2 gene variants are characterized by an increased myofilament Ca2+ sensitivity, perturbed length‐dependent myofilament activation, and increased cross‐bridge kinetics and energetics.10, 11, 12, 13 To understand how these TNNT2 gene variant–mediated myofilament changes translate to changes in cardiac phenotype, a better understanding of the cardiac changes at preclinical and HCM disease stage is warranted.

Noninvasive imaging studies have demonstrated impaired cardiac energetics in both animals and humans with HCM,14, 15, 16 and even in carriers (genotype positive) without hypertrophy (G+/LVH).17 Individuals with thick‐filament gene variants showed reduced myocardial efficiency compared with healthy controls.18 In addition, carriers harboring MYH7 (β‐myosin heavy chain gene) variants demonstrated a more prominent reduction of myocardial efficiency compared with MYBPC3 (myosin‐binding protein C gene) carriers, indicative for a gene‐specific effect.19 Moreover, myocardial efficiency was further decreased in patients with obstructive HCM at the time of myectomy.19

As energetic alterations may play an important role in the preclinical stage of HCM pathophysiological characteristics, and may serve as a target for future therapy,13, 14, 20 herein we investigated if myocardial efficiency is altered at preclinical (genotype positive/LVH negative [G+/LVH−]) and HCM disease stage (genotype positive/LVH positive [G+/LVH+]) in TNNT2 gene variant carriers. Myocardial external efficiency (MEE) (ie the ratio between external work [EW]/myocardial oxygen consumption [MVO2]), was assessed in vivo by state‐of‐the‐art [11C]‐acetate positron emission tomography (PET) and cardiac magnetic resonance (CMR) imaging in 6 G+/LVH− and 8 G+/LVH+ individuals.

Methods

The data that support the findings of this study are available from the corresponding author on reasonable request.

Study Population

The study was approved by the local Ethics Committee and was performed in agreement with the Declaration of Helsinki. All participants gave written informed consent before inclusion. All preclinical gene variant carriers (G+/LVH−; n=6) and genotype‐positive HCM (G+/LVH+; n=8) patients were prospectively enrolled between March 2017 and October 2018. All participants were genetically tested positive for variants in the TNNT2 gene. Classification into the G+/LVH− and G+/LVH+ groups was based on the criteria for septal thickness, as proposed in the current European Society of Cardiology guidelines (ie, HCM is defined by a wall thickness ≥15 mm [≥13 mm in case of first‐degree family members] in ≥1 LV myocardial segments in the absence of any other cardiac or systemic condition likely to cause LV hypertrophy). In addition, maximal wall thickness for the G+/LVH− group was set at ≤10 mm. Exclusion criteria were septal reduction therapy or heart transplantation in their medical history, renal impairment (<30 mL·min−1·1.73 m−2), hypertension, and any relative or absolute contraindication to undergo a CMR scan. Fourteen healthy individuals served as the control group (data of the control group have been published before).19 Before the imaging protocols, from all participants, blood samples were drawn and NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide; expressed in ng·L−1), hemoglobin, creatinine, glucose, free fatty acid, and lactate levels were determined.

Cardiac Imaging Studies

Transthoracic echocardiography

To derive LV outflow tract gradients and diastolic function, continuous‐wave Doppler was applied, according to the American Society of Echocardiography guidelines.21 Septal diastolic mitral annular velocity and peak early diastolic mitral inflow velocity were measured in the apical 4‐chamber view.

CMR and positron emission tomography imaging

All participants underwent CMR imaging on a 1.5‐T whole body scanner (Avanto; Siemens, Erlangen, Germany), using a 6‐channel phased‐array body coil. Cine images were obtained using a standard retrospective gated, single breath‐hold segmented k‐space balanced steady‐state free sequence, with contiguous short axis slices to cover the whole LV from base to apex.

PET was performed to noninvasively assess oxygen metabolism using the rate constant K2, which represents the rate of transfer of radioactivity from tissue to blood from which MVO2 is derived.22 All participants underwent a PET scan, after overnight fasting, on a Gemini TF‐64 PET/CT scanner (Philips Healthcare, Best, The Netherlands). Data of the control group were acquired as described previously.23 Representative CMR and PET images are depicted in Figure 1. Reproducibility of CMR/PET analysis was high, with low intraobserver and interobserver variability.18 For additional image acquisitions on CMR and PET, see Data S1.

Figure 1

Cardiac images of control, genotype‐positive/left ventricular hypertrophy–negative (G+/LVH−), and genotype‐positive/left ventricular hypertrophy–positive (G+/LVH+) individuals.

Examples of a cardiac magnetic resonance (CMR) 4‐chamber view and parametric images of [11C]‐acetate positron emission tomography–derived images and the corresponding polar maps are shown for control, G+/LVH−, and G+/LVH+. k2 indicates average [11C]‐acetate clearance rate constant.

Post processing

The CMR cine images were analyzed off‐line using MASS analysis software, version 2.1 (Medis Medical Imaging Systems, Leiden, The Netherlands). End‐diastolic and end‐systolic volumes of the LV and LV ejection fraction were obtained by application of the endocardial contours. Addition of epicardial contours resulted in LV mass (LVM). End‐diastolic wall thickness at the septum was derived from the mean of 4 septal segments (anteroseptal and inferoseptal) at the basal and midventricular level. Tissue tagging images were analyzed by inTag (https://www.creat​is.insa-lyon.fr/osirix-dev/Cardi​acToo​ls.html) software (CREATIS, Lyon, France) to quantify myocardial deformation using the SinMod technique and estimate regional peak circumferential strain components (systolic circumferential strain [SCS] and diastolic circumferential strain rate). Myocardial strain was measured in the midmyocardial layer, which has been reported to be the most reproducible.24 The software runs as a plug‐in for OsiriX, version 6.5 (Pixmeo, Switzerland).25 Analysis of the LV was calculated according to the 17‐segment American Heart Association model.26 Late gadolinium enhancement (LGE) was assessed by applying the full width at half maximum method on the LGE cine short axis images and is expressed as percentages of the LV mass.27 The aQuant software package was used for analysis of dynamic PET data.28 The product of stroke volume and mean arterial pressure, which yield EW, and PET‐derived MVO2 allows assessment of MEE, according to the following equation:

HR represents the heart rate, and the caloric equivalent of 1 mm Hg·mL EW is 1.33·10−4, whereas 1 mL O2 corresponds to 20 J.29

As hypertrophy in HCM is asymmetric, affecting the septum of the heart, changes in myocardial function and efficiency may differ between LV regions. Efficiency was therefore determined in the septum and lateral wall of the LV as the ratio between regional SCS and the corresponding MVO2(beat) according to the 17‐segment model of the American Heart Association.26 Less negative values indicate reduced efficiency.19

Statistical Analysis

Statistical analysis was performed using SPSS software, version 22.0 (SPSS, Chicago, IL). Normality of data was inspected visually by means of QQ‐plots. The χ2 test was used for categorical demographic variables. Means of continuous variables were compared between groups using ANOVA tests after normality was verified or a Mann‐Whitney U test if data were not normally distributed. In case of a significant overall ANOVA test, post hoc tests were performed with a Bonferroni correction to account for multiple comparisons. An (overall) 2‐sided significance level of 5% was used for all statistical tests.

Results

Recruitment and Characteristics of Controls, G+/LVH−, and G+/LVH+

Participants were recruited from clinical centers in The Netherlands, including Amsterdam University Medical Center, Leiden University Medical Center, and Erasmus Medical Center. In total, 85 individuals were identified with a pathogenic TNNT2 gene variant (Figure 2). Four (5%) eligible G+/LVH− individuals and 6 (7%) eligible G+/LVH+ were included. The remaining subjects were excluded because of the following reasons: 22 (26%) were aged >65 years, 18 (21%) could not be reached, 12 (14%) had an implantable cardioverter‐defibrillator, 10 (12%) had cardiac phenotypes other than HCM, 5 (6%) refused participation for different reasons, and the remaining group consisting of 8 patients (9%) had died in the past years, had atrial fibrillation, had type 2 diabetes mellitus, carried a gene variant of unknown significance, and/or had a LV wall thickness measuring between 10 and 15 mm. In both G+/LVH− and G+/LVH+ groups, 2 subjects were included via their family member. The G+/LVH− group includes one identical twin. Because of the difficulty to include eligible G+/LVH−, this group consisted only of women. Analysis of MVO2 and MEE in our control group did not show sex differences (Table S1), indicating that sex does not introduce a bias. TNNT2 variants of G+/LVH− and G+/LVH+ are listed in Table 1 and shown in the schematic figure of the thin filament in Figure 3.30 G+/LVH− and controls did not use medication. G+/LVH+ had a lower heart rate compared with controls (Table 1, Figure 4), most likely explained by the use of β blockers in this group. Systolic and diastolic blood pressures were comparable between all groups. Individuals in the G+/LVH+ were predominantly men and showed elevated levels of NT‐proBNP (Table 1).

Figure 2

Distribution of (troponin T gene) variant‐positive individuals.

In total, 85 individuals were identified with a pathogenic gene variant. Among these, 5% eligible genotype positive/left ventricular hypertrophy negative (G+/LVH−) and 7% eligible genotype positive/left ventricular hypertrophy positive (G+/LVH+) were included. The remaining individuals were excluded because of the reasons indicated in each slice. ARVC indicates arrhythmogenic right ventricular cardiomyopathy; DCM, dilated cardiomyopathy; ICD, implantable cardioverter‐defibrillator; LQT, long‐QT syndrome; and NCCM, noncompaction cardiomyopathy.

Table 1

Baseline Characteristics

Characteristic	Controls (n=14)	G+/LVH− (n=6)	G+/LVH+ (n=8)
Genotype
c.277G>A; p.Glu93Lys	…	1	1
c.304C>T, p.Arg102Trp	…	3	1
c.403C>T, p.Arg144Trp	…	1	0
c.832C>T p.Arg278Cys	…	0	1
c.835C>T; p.Gln279a	…	0	1
c.853C>T, p.Arg285Cys	…	0	1
c.856C>T, p.Arg286Cys	No gene variant	1	3
Age, y	48±11	43±15	46±16
Sex (men)	9	0a , b	7b
Body surface area, m2	2.0±0.2	1.8±0.2	2.1±0.3b
Heart rate, beats/min	69±10	62±6	57±7a
Systolic blood pressure, mm Hg	123±13	114±14	123±15
Diastolic blood pressure, mm Hg	71±8	73±12	77±10
Mean arterial pressure, mm Hg	88±8	86±12	92±11
Medical treatment, n (%)
β Blockers	0 (0)	0 (0)	2 (25)
Calcium antagonist	0 (0)	0 (0)	2 (25)
ACE inhibitors	0 (0)	0 (0)	2 (25)
Metabolic parameters
k2	0.08±0.02	0.11±0.02a	0.08±0.01b
Hemoglobin, mmol·L−1	8.3±0.4	8.8±0.5	9.4±0.9a
NT‐proBNP, ng·L−1	63±55	95±65	294±178a , b
Free fatty acids, mmol·L−1	0.55±0.26	0.67±0.20	0.49±0.27
Lactate, mmol·L−1	1.4±0.6	1.3±0.6	1.1±0.3
Glucose, mmol·L−1	5.5±0.8	5.9±1.5	5.4±0.5
Echocardiographic parameters
Septal e’, cm·s−1	10.1±2.3	8.1±4.4	7.2±0.9
E/A ratio	1.4±0.3	1.5±1.0	1.3±0.2
Mean LVOT gradient, mm Hg	NA	NA	3±1
CMR parameters
Maximal LV septal wall thickness, mm	7 (6–8)	9 (8–10)a	16 (15–16)a , b
LV lateral wall thickness, mm	6.0±0.9	6.8±0.8	9.4±1.4a , b
LV end‐diastolic volume, mL·m−2	93±15	73±12a	90±14
LV end‐systolic volume, mL·m−2	36±10	27±5	28±10
Forward stroke volume, mL	100±23	71±12a	96±28
LV ejection fraction, %	62±5	65±6	73±13a
LV mass, g·m−2	49±6	39±5	67±13a , b
Indexed LV septal wall thickness, mm·m−2	3.5±0.4	5.1±0.6a	7.7±1.1a , b
Septal/lateral wall ratio	1.2±0.1	1.3±0.1	1.8±0.3a , b
Late gadolinium enhancement, n (%)	0 (0)	0 (0)	7 (11±10)

Data are presented as number, mean±SD, or median (interquartile range). ACE indicates angiotensin‐converting enzyme; CMR, cardiac magnetic resonance; E/A, early to late ventricular filling velocities; e’, early diastolic mitral annular velocity; G+/LVH+, genotype positive/left ventricular hypertrophy positive; G+/LVH−, genotype positive/left ventricular hypertrophy negative; k2, average [11C]‐acetate clearance rate constant; LV, left ventricular; LVOT, LV outflow tract; NA, not applicable; and NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide.

P<0.05 vs controls.

P<0.05 vs G+/LVH−.

Figure 3

Schematic figure of locations of (troponin T gene) variants.

Schematic of the thin filament composed of actin, tropomyosin, and the troponin complex, and myosin head of the thick filament (depicted in green). Seven actin monomers (gray) spanned by 1 tropomyosin dimer (red) and 1 troponin complex: cardiac troponin C (pink), cardiac troponin I (blue), and cardiac troponin T (orange). Dark‐blue tropomyosin depicts near‐neighbor tropomyosin dimer interaction. The gene variants detected in 6 genotype‐positive/left ventricular hypertrophy–negative (G+/LVH−) and 8 genotype‐positive/left ventricular hypertrophy–positive (G+/LVH+) study subjects are depicted relative to their location in the troponin T protein. Based on the figure shown with permission from Sequeira et al.30 Copyright ©2015, Springer Nature. C indicates C‐terminal protein end; N, N‐terminal protein end. Troponin T can be divided in three sub‐regions: the N‐terminal hypervariable region, TNT1 and the C‐terminal TNT2.

Figure 4

Reduced myocardial external efficiency at preclinical and hypertrophic cardiomyopathy (HCM) disease stage.

Scatterplots depict external work (EW), heart rate (HR), total myocardial oxygen consumption (MVO 2), left ventricular mass (LVM), and myocardial external efficiency (MEE) in controls, genotype positive/left ventricular hypertrophy negative (G+/LVH−), and genotype positive/left ventricular hypertrophy positive (G+/LVH+). EW was similar in G+/LVH+ and controls, but tended to be lower in G+/LVH−. Total MVO 2 was higher in G+/LVH+ compared with controls and G+/LVH−, which is explained by the higher LVM in G+/LVH+. Calculation of MEE revealed significantly lower values in G+/LVH− and G+/LVH+ compared with controls, indicative for reduced cardiac efficiency at early preclinical disease stage. MEE was similar at preclinical (G+/LVH−) and HCM (G+/LVH+) disease stage. Data are presented as mean with SD.

G+/LVH+ had a significantly higher maximal septal and lateral wall thickness than G+/LVH− and controls (Table 1), although no obstruction was evident from low LV outflow tract gradients (Table 1). In accordance with a higher wall thickness in G+/LVH+, LV mass in G+/LVH+ was higher compared with G+/LVH− and controls (Table 1, Figure 4). LV end‐diastolic volume and stroke volume were lower in G+/LVH− compared with controls, which is explained by female predominance in the G+/LVH− group (Table 1).31 G+/LVH+ and controls had comparable LV end‐diastolic volume and stroke volume (Table 1). None of the G+/LVH− subjects and controls had contrast enhancement on LGE images. Of 8 G+/LVH+ subjects, 7 had contrast enhancement on LGE imaging, which indicates fibrosis, with a median estimated percentage of 7%, mainly located at the right ventricle insertion points. LV ejection fraction was significantly higher in G+/LVH+ compared with controls, whereas global peak circumferential strain in G+/LVH+ was significantly lower compared with G+/LVH− and controls (Tables 1 and 2). LV ejection fraction and global peak circumferential strain in G+/LVH− were similar to controls (Tables 1 and 2).

Table 2

Regional Contractile Function and Efficiency

Variable	Controls (n=14)	G+/LVH− (n=6)	G+/LVH+ (n=8)
Peak systolic circumferential strain, %
Global	−17.5±1.4	−18.8±2.5	−15.5±2.2a , b
Septal	−16.2±2.2	−16.8±2.5	−13.3±2.7a , b
Lateral	−19.0±2.0	−20.8±2.7	−18.5±2.3
Peak diastolic circumferential strain rate, %·s−1
Global	38.3±7.0	32.0±5.2	28.7±6.8
Septal	38.2±8.7	30.9±7.5	24.6±7.2a
Lateral	40.8±7.5	34.6±5.7	37.0±6.7
MVO2 per beat, mL/beat per g·10−3
Global	1.4±0.3	2.2±0.4a	1.7±0.2a , b
Septal	1.4±0.3	2.3±0.4a	1.7±0.2b
Lateral	1.4±0.3	2.3±0.5a	1.9±0.2a
Efficiency (systolic circumferential strain/MVO2 per beat)
Global	−12 009±1797	−8458±2048a	−8047±1379a
Septal	−11 413±1430	−7666±1968a	−7744±1856a
Lateral	−13 625±2752	−9633±2464a	−9841±1617a

Data are presented as mean±SD. G+/LVH+ indicates genotype positive/left ventricular hypertrophy positive; G+/LVH−, genotype positive/left ventricular hypertrophy negative; and MVO2, myocardial oxygen consumption.

P<0.05 vs controls.

P<0.05 vs G+/LVH− individuals.

With respect to regional function, septal peak SCS in G+/LVH+ was significantly lower than G+/LVH− and controls, whereas lateral peak SCS was similar as in G+/LVH− and controls (Table 2). Septal and lateral peak SCS in G+/LVH− were similar as in controls (Table 2). Septal peak diastolic circumferential strain rate was significantly lower in G+/LVH+ compared with controls, whereas G+/LVH− only showed a tendency toward lower septal peak diastolic circumferential strain rate compared with controls (Table 2). Lateral peak diastolic circumferential strain rate in G+/LVH+ and G+/LVH− was comparable to controls (Table 2).

Overall, the anatomical and functional characteristics show no differences between G+/LVH− compared with controls, whereas the G+/LVH+ group shows cardiac changes that are characteristic for HCM, such as increased LV mass, higher LV ejection fraction, lower septal peak SCS, and fibrosis.

Reduced Myocardial Efficiency at Preclinical and HCM Disease Stage in TNNT2 Gene Variant Carriers

Figure 1 shows representative images of CMR and PET in the 3 groups. EW was similar in G+/LVH+ and controls, whereas it tended to be lower in G+/LVH− (Figure 4). Total MVO2 was significantly higher in G+/LVH+ compared with controls and G+/LVH−, which is explained by the higher LV mass in G+/LVH+ (Figure 4). Calculation of MEE revealed significantly lower values in G+/LVH− and G+/LVH+ compared with controls (Figure 4), indicative for reduced cardiac efficiency at preclinical disease stage.

Septal Versus Lateral LV Wall Changes in Efficiency of the HCM Heart

As hypertrophy specifically develops in the septum of the LV, we assessed regional efficiency to establish if reduced MEE is more severe in the septal than in the lateral wall of the LV. Figure 5 shows that septal and lateral myocardial efficiency, the ratio between peak SCS/MVO2(beat), were significantly lower (ie, less negative values) in G+/LVH− and G+/LVH+ compared with controls. The reduction in efficiency was similar in the septal and lateral wall of the LV both at preclinical and HCM disease stage. The reduction in efficiency at preclinical disease stage is explained by a significantly higher regional MVO2 as no significant difference is present in regional peak SCS. A similar pattern is observed in the lateral wall of G+/LVH+ individuals with a significantly higher oxygen consumption and unaltered systolic strain compared with controls. The reduction in septal efficiency in G+/LVH+ compared with controls is explained by significantly lower systolic strain and the hypertrophy (LV mass)–related increase in oxygen consumption. Notably, septal oxygen consumption in G+/LVH+ did not differ from controls, and was significantly lower compared with G+/LVH−, indicative for a remodeling‐related change in the hypertrophied septum of the G+/LVH+ group.

Figure 5

Cardiac efficiency, regional contractility and oxygen consumption.

Septal and lateral myocardial efficiency were significantly lower in genotype positive/left ventricular hypertrophy negative (G+/LVH−) and genotype positive/left ventricular hypertrophy positive (G+/LVH+) compared with controls. The scatterplots depicting systolic circumferential strain (SCS) and myocardial oxygen consumption (MVO 2) illustrate which regional changes cause the reduction in septal and lateral efficiency. Data are presented as mean with SD.

Discussion

Using advanced cardiac PET and CMR imaging, we show reduced cardiac efficiency at preclinical and HCM disease stage in individuals carrying a TNNT2 gene variant. The lower MEE in G+/LVH− is explained by higher global and regional oxygen consumption compared with healthy controls. Regional analysis showed significantly higher MVO2 in the septal and lateral LV wall of G+/LVH− and the lateral wall of G+/LVH+ compared with controls, indicating that the presence of a TNNT2 gene variant increases local oxygen consumption and reduces efficiency of cardiac contraction. The reduced septal efficiency in the HCM group is explained by the increase in LV mass and concomitant higher oxygen consumption and reduced systolic strain. Septal oxygen consumption was significantly lower in G+/LVH+ compared with G+/LVH−, suggesting that disease mechanisms other than the gene variant alter oxygen consumption and/or delivery in the hypertrophied myocardium. No significant changes in regional contractile parameters, both systolic and diastolic, were observed at preclinical disease stage. Although LV ejection fraction was significantly higher in patients with HCM, both systolic and diastolic strain parameters were lower compared with controls, which was most evident in the hypertrophied septal wall of the LV. Our data show that the increase in MVO2 in TNNT2 gene variant carriers precedes changes in global and regional myocardial contractility, indicating that the energetic status rather than contractile parameters reflects the initial variant‐induced pathomechanism that can be used for early diagnosis and preventive therapy. The combination of reduced strain, increased NT‐proBNP levels, and reduced peak diastolic strain rate indicates progression of disease in G+/LVH+ patients compared with G+/LVH−. In this respect, the reduction (pseudonormalization) of MVO2 observed in this study is in line with energy depletion also observed in patients with HCM and patients with heart failure.

Reduced Myocardial Efficiency Caused by Different Variants in TNNT2

Cardiac troponin T is part of the thin filament of the sarcomere, which is composed of actin, tropomyosin, and the troponin complex (Figure 3). Together with cardiac troponin I and cardiac troponin C, cardiac troponin T forms the Ca2+‐regulatory troponin complex of the thin filament. Activation of cardiomyocytes induces an increase in cytosolic [Ca2+] and Ca2+ binding to cardiac troponin C, which changes the conformation of the troponin‐tropomyosin complex, and exposes myosin‐binding sites on actin. Increased binding of myosin heads to the actin thin filament (formation of cross‐bridges) subsequently generates force. The troponin complex is thus a central player in cardiomyocyte force development during the systolic and diastolic phase of the cardiac cycle. On the basis of its central role in myofilament Ca2+ signaling and contractility, variants in TNNT2 are likely to alter myofilament properties. Several myofilament changes caused by TNNT2 variants have been reported that may underlie increased energy (oxygen) consumption and the reduction in myocardial efficiency. A common feature of myofilaments harboring TNNT2 variants is increased myofilament Ca2+ sensitivity, which coincides with increased ATPase activity.18, 32

Ca2+‐activated human myofilaments harboring a TNNT2 variant showed higher cross‐bridge kinetics, increased tension cost (ie, higher ATP use for force development), and a blunted length‐dependent activation compared with nonfailing control tissue.11, 33 The higher myofilament Ca2+ sensitivity and increased cross‐bridge kinetics may underlie reduced MEE in preclinical G+/LVH−. In addition, the blunted length‐dependent activation of myofilaments with a TNNT2 gene variant represents a highly inefficient myofilament mechanism. Studies in HCM mouse models showed large differences in the variant‐dependent increase in myofilament Ca2+ sensitivity, with the I79N variant causing the largest increase, and the R92Q variant having no significant effect.34 Our study population carried variants in different parts of the TNNT2 gene. Cardiac troponin T can be divided in 3 subregions: the N‐terminal hypervariable region (residues 1–79), TNT1 (residues 80–180), and the C‐terminal TNT2 (residues 181–288; Figure 3).35, 36 Gene variants located in the TNT1 region result in a reduced affinity of cardiac troponin T for tropomyosin, whereas C‐terminus gene variants do not alter this interaction, suggesting variant location‐dependent pathomechanisms.37, 38 Most of the TNNT2 variants in our G+/LVH+ group are located in the TNT2 region, whereas in 5 of 6 G+/LVH−, the variant is located in the TNT1 region (Figure 3). Although the location of the variant may underlie its pathomechanism and pathogenicity, we did not observe a difference in MEE between variants located in TNT1 and TNT2 as MEE is similarly reduced in individuals with TNT1 and TNT2 gene variants (Figure 4). Prospective imaging studies are warranted in young (aged 20–45 years) male and female TNT1 and TNT2 variant carriers to establish if the variant‐mediated reduction in MEE has prognostic relevance.

Limitations and Clinical Implications

Sex differences

We used the current European Society of Cardiology guidelines to classify the gene variant carriers into a G+/LVH− group and a G+/LVH+ group. A significant sex difference in the distribution of women and men in the 2 groups is present, which may be partly explained by using a cutoff value for septal thickness that is uncorrected for body surface area (BSA). In recent studies, we highlighted the importance of correcting septal thickness by BSA as women are in general smaller than men.31, 39 When we do take into account differences in BSA in the present study, indexed LV septal wall thickness is significantly higher in the female G+/LVH− group (5.1±0.6; Table 1) compared with female controls (3.5±0.4; Table S1). The increase in indexed septal wall thickness in female G+/LVH− compared with female controls is 45%, whereas the increase in indexed septal wall thickness in male patients with HCM (7.7±1.1; Table 1) compared with male controls (3.5±0.5; Table S1) is 120%. The latter supports the more advanced remodeling in the G+/LVH+ group, although it also emphasizes the need to correct septal thickness and LV mass for BSA, and necessitates a reappraisal of the definition of hypertrophy in HCM. BSA‐corrected values for MEE and septal oxygen consumption are shown in Figure S1, and illustrate a similar pattern as the uncorrected values (Figures 4 and 5). Overall, our study indicates that future studies on cardiac remodeling should take into account differences in BSA as they may explain sex‐specific differences in disease cardiac remodeling and disease progression.

Remodeling‐related changes in cardiac efficiency

In addition to the variant‐induced functional changes in the heart, additional disease mechanisms may underlie the reduction in MEE, such as capillary rarefaction,40 microvascular (endothelial) dysfunction,41 mitochondrial dysfunction, and oxidative stress.42 Coronary flow reserve may underlie changes in MEE, and has been proposed as disease marker for HCM independent of the severity of LV hypertrophy.43 The blunted vasodilator reserve in the absence of a coronary stenosis in HCM is the result of microvascular dysfunction.41 Pharmacologically induced coronary vasodilation was significantly impaired in the hypertrophied septal and nonhypertrophied free wall of patients with HCM,44 showing that vascular (endothelial) dysfunction is independent of cardiac hypertrophy. In a previous study, we did not find impaired vascular reserve in MYBPC3 G+/LVH− and reduced MEE.23 Studies to define regional coronary flow reserve at different stages of HCM with different genotypes are warranted to define the role of impaired coronary vasodilation in the pathogenesis of HCM with respect to changes in cardiac efficiency and metabolism. In addition to microvascular changes, increased interstitial fibrosis, a marker of cardiac ischemia, may contribute to inefficient cardiac function. Because the current study participants did not undergo cardiac catheterization to exclude coronary artery disease before enrollment, myocardial ischemia attributable to coronary artery disease could potentially influence the current results. However, this study has a limited sample size, participants were asymptomatic and at low risk to experience coronary artery disease, and electrocardiography did not reveal abnormalities attributable to ischemic cardiac disease. Myocardial metabolism was derived by the clearance rate of carbon‐11 acetate, which can only be measured in viable myocardium. Analysis of extracellular volume using CMR with LGE images (Table 1) showed no LGE in G+/LVH−, indicating that a change in regional extracellular volume does not influence their energetic phenotype. Because G+/LVH+ individuals presented with limited scar tissue on CMR (Table 1), and thus have limited nonviable myocardium, we did not correct MEE for scar tissue in our analyses. The effect of extracellular volume on regional contractile function was shown to be limited,45 and therefore extracellular volume measurement was not performed in this study. However, the interaction between in vivo measured myocardial structure and regional function still warrants further study.

Treatment options

Currently, there are no preventive or curative therapies for HCM. Metabolic modulators have been proposed to correct energy deficiency.46 In animal and human studies, metabolic modulators, such as trimetazidine and perhexiline, have a proven positive effect on energy efficiency and have improved diastolic function and exercise capacity.47, 48 Perhexiline improved exercise capacity in patients with obstructive HCM,46 whereas a recent study in patients with HCM did not show a beneficial effect of trimetazidine.49 Effectiveness of therapies may depend on the affected gene, as has been shown by Ho et al, who showed a positive effect of diltiazem only in preclinical MYBPC3 gene variant carriers.50 Thus far, clinical trials in patients with HCM with metabolic modulators have been performed in mixed groups of patients with HCM with and without (known) gene variants. Moreover, our study indicates that the effectiveness of metabolic therapy may depend on disease stage as oxygen consumption is not increased in the hypertrophied septal region of the LV in HCM. Alternative attractive strategies to lower oxygen consumption are therapies that aim to lower contractility of the heart muscle.51 A recent proof‐of‐concept study performed in symptomatic patients with HCM revealed significant reduction of LV outflow tract gradient and improvement of exercise capacity with a 12‐week treatment with mavacamten, which is suggested to attenuate hypercontractile myofilaments, and therefore should be investigated further as a potential treatment in patients with HCM.52 To come to disease stage‐specific and even gene‐specific treatment strategies, more knowledge is needed about the pathomechanisms underlying reduced MEE. Follow‐up studies are warranted to investigate the mechanisms (ie, metabolism, mitochondrial function, and vascular responsiveness) underlying the changes in oxygen consumption and delivery during the transition from preclinical nonhypertrophied disease stage to manifest obstructive HCM.

In conclusion, our study shows that preclinical gene variant carriers have an initial increase in oxygen consumption preceding cardiac hypertrophy and contractile dysfunction, suggesting that high oxygen consumption and reduced MEE characterize the early disease mechanisms that may be used for early diagnosis and development of preventive treatments.

Sources of Funding

This work was supported by The Netherlands Heart Foundation (CVON‐Dosis 2014–40) and Netherlands Organization for Sciences‐ZonMW (VICI 91818602).

Disclosures

None.

Data S1 Table S1 Figure S1 References 22, 28, 53 and 54

Click here for additional data file.