Apparent growth tensor of left ventricular post myocardial infarction – In human first natural history study
Graphical abstract
Introduction
Myocardial infarction (MI), also known as heart attack, involves cardiomyocyte death following the loss of antegrade coronary blood flow. After acute MI, the left ventricle (LV) undergoes a sequence of adaptations in structure and function, regulated by mechanical, neuro-hormonal and genetic factors [7,17,47]. This myocardium growth & remodelling (G&R) may become maladaptive. The LV may experience dilation in association with wall thinning around the infarct zone, and hypertrophy or fibrosis (collagen fibre scar) in the remote zone. Adverse G&R, i.e. dila-tion, can lead to heart failure, a major source of morbidity and mortality [16], and sudden cardiac death due to malignant arrhythmias. Over the years, with improvement in the treatment of acute MI, more people are surviving post-MI, but with a subsequent rise in heart failure [42,51].
Cardiac G&R has previously been categorized into two types: (1) concentric G&R with wall thickening caused by pressure over-loading; (2) eccentric G&R with dilation of the ventricular cavity and wall stretching, both induced by pressure over-loading [3,24]. However, MI-induced LV G&R is more complex, and recently a classification, such as reverse remodelling, no remodelling, and adverse remodelling, was proposed in terms of 12% relative changes in LV end-diastolic volume (LVEDV) [11]. MI-induced progressive G&R is an important clinical determinant of heart failure [30], and arresting or prompting reverse remodelling is a long-term aim to reduce the risk of heart failure [7].
Usually, LV G&R is quantified by evaluating LV global changes, such as LVEDV, end-systolic volume (ESV), ejection fraction (EF) and wall thickness [10]. However, these indices only describe an incomplete picture of the G&R process. More detailed local patterns of G&R in MI patients, the difference in MI and remote regions can provide additional insights into identifying patients with a high risk of adverse remodelling and evaluating interventions aimed at reducing the progression of heart failure.
Quantification of local LV G&R following MI is difficult, requiring serial measurements of in vivo myocardium geometry and motion. For research purposes, measurements are often made in animals by tracking several small-sized beads embedded in the LV wall through open-chest surgery [26,27,39,49]. For example, Kass et al. [27] measured disproportionate epicardial dilation after transmural infarction of the canine LV within 7 days of acute MI. By suturing stainless steel spherical beads to the epicardial surface, they found that 1-h after infarction, the end-diastole area in the infarct region increased by 20.3% above healthy control (normal LV before infarction) compared with a 7.9% increase in the remote region. 24-hour post-MI, both the regions expanded by an additional 10%. One-week post-MI, the end-diastole area in the infarct region increased to 31.4% above the control, but the remote area became to only 8.5% above the control. Holmes et al. [26] measured scar transmural remodelling after infarction in five porcine hearts. Fitting the measured bead positions into a finite element (FE) model at 1- and 3-week post-MI, they calculated the post-MI strain against the reference configuration (at end-diastole before infarction). Their results showed that at 1-week post-MI, infarct expansion dominated the remodelling process, as characterized by the circumferential and longitudinal in-plane stretch and wall thinning. However, post-MI strain at 3-week is much more complex. Two out of five LVs showed continuous in-plane infarction expansion with wall-thinning, and the remaining three had in-plane shrinkage either with wall thinning or thickening. In a more recent study, Tsamis et al. [49] implanted three transmural columns of bead sets across the mid-lateral equatorial region of a sheep LV and recorded the bead positions using biplane cine-radiography before infarction and 7-week post-MI. They calculated the end-diastolic and end-systolic post-MI strains in a lateral-equatorial region adjacent to infarct by fitting bead positions using a FE model and found there were chronic muscle fibre lengthening ( 10%) in the longitudinal direction, 25% shrinkage in the radial or transmural directions, and 15% decrease in the ventricular volume.
Although animal experiments can provide useful data for calibrating, validating and verifying various cardiac growth modelling theories, the bead implantation method is invasive and not suitable for human subjects. With the development of cardiac magnetic resonance (CMR) [5,6,15,54] and ultrasound imaging [45,46], it is now possible to quantify some regional geometry and function post-MI in vivo. However, using in vivo CMR images to track G&R post-MI is challenging. Longitudinal 3D cine images may not be available, motion artefacts may affect in the images, and ventricular geometries at different time-points must be co-registered. O'Regan et al. [40] studied LV remodelling during the first year after reperfused acute ST-elevation MI (STEMI) in 47 patients using a three-dimensional (3D) co-registration approach. Both short-axis cine and late gadolinium enhanced (LGE) images were co-registered to an atlas template and local expansion of the heart wall was tracked by using intensity-based similarities. End-diastolic remodelling is defined as the mean change of vertex separations from the matched endocardial and epicardial surfaces. Their results showed that in-plane expansion in the infarct region is greater than in the remote myocardium (1.6%1.0 vs 0.3%0.9), and associated with wall-thinning in the infarct region but no changes in wall thickness in the remote region. Additionally, they demonstrated that MI transmurality and microvascular obstruction (MVO) may affect local remodelling considerably. To our knowledge, O'Regan et al. [40] were the first group to report patient measurements of MI-induced local myocardial remodelling in vivo. However, they only studied the local expansion characterization and did not estimate the growth tensor.
Understanding G&R of the LV post-MI is critically important to risk stratifying preventive therapy for heart failure. A key element of monitoring a diseased heart is to extract growth tensor evolution with time from non-invasive clinical images. Computational models based on continuum mechanics and clinical imaging assessment techniques have recently been developed to overcome the limitations by describing G&R based on routine clinical images [2,9,55,56].
In this work, to quantify G&R in patients with recent MI and evaluate its relation to adverse remodelling, a continuum mechanics approach is employed to evaluate the LV apparent growth tensor in 16 patients post-MI undergoing CMR at 4 distinct time-points: at 4–12 h (baseline), 3 days, 10 days, and 7 months after reperfusion. Primary percutaneous coronary intervention (PPCI) was performed within a few hours (<12 h) following the onset of acute MI. The end-diastolic (ED) LV geometries are first reconstructed based on CMR cine images at different time points. The MI regions are then delineated from corresponding short-axial LGE images. The reconstructed LV geometries are co-registered to a template LV finite element mesh, and the apparent growth tensors at end-diastole in MI and remote regions are calculated against the baseline LV geometry. All patients are further classified into three groups (Dilation, Shrinkage or No-Change) according to their end-diastolic volume changes based on the criteria proposed by Bulluck et al. [11] for adverse remodelling, reverse remodelling and no-remodelling behaviours. Finally, the invariants of the growth tensor extracted in both the MI and remote regions are investigated and their statistical significance is examined against the short-term (one week) and long-term (6 months) pump function recovery in these groups. We believe the outcomes of this study will be useful to quantitatively characterize G&R for patients post-MI, and provide much needed clinical input for mechanobiology modelling of G&R process in the human heart.
Section snippets
Patient population and CMR protocol
This study involves 16 STEMI patients with four longitudinal CMR scans, at 4–12 h, 3 days, 10 days, and 7 months after reperfusion. All patients were enrolled between May 2011 to November 2012 with acute STEMI (Clinical trial registration no, NCT 02072850). All patients provided written informed consent.
CMR imaging was performed on a Siemens MAGNETOM Avanto 1.5-T scanner (Erlangen, Germany) using an anterior phased-array body coil (12-element) and a posterior phased-array spine coil
Validations
Validation 1: the LVEDV and the MI volume ratio () from our reconstructions are compared with the measurements obtained by K.M. at 4–12 h and 7 months post-MI, as shown in Fig. 7. In general, good agreement of the LVEDV volume ratio is observed between clinical assessment and values derived from our 3D reconstructions with a mean error (clinical - reconstruction) of +14.43% at 4–12 h and +11.13% at 7 months, respectively. Excellent agreement of the MI volume ratio is also achieved with a
Results
Fig. 9 shows the normalized LVEDV changes (LVEDV/LVEDV0) at 3 days, 10 days and 7 months after reperfusion. According to the criterion of a 12% LVEDV change at 7-months post-MI [11], out of 16 patients, nine (ID: 1, 2, 3, 4, 6, 7, 8, 11, 12) (56.25%) are in the Dilation group; two (ID: 9, 10) (12.5%) are in the No-Change group, and five (ID: 5, 13, 14, 15, 16) are in the Shrinkage group. The averaged clinical indices at baseline and 7-month follow up for the three groups are listed in Table 3.
Discussion
To the best of our knowledge, this has been the first time that growth tensor is quantitatively analysed from the human studies post-MI. It is challenging to measure growth tensors in situ from animal studies [26,49], but more so from in vivo clinical studies. Imaging techniques can therefore provide an alternative source of information. For example, O'Regan et al. [40] quantified LV remodelling by tracking ventricular expansion using a 3D co-registration approach based on intensity-based
Conclusion
In this paper, we have developed an image-derived method for estimating apparent myocardial growth tensors in 16 patients post-MI. Both routinely available cine and LGE CMR images are used for personalized LV geometry reconstruction. We have demonstrated that the estimated growth tensor agrees well with published experimental data and conventional clinical observation. Our results show that the greatest shape change occurs in the Dilation group and the least in the Shrinkage group. Both and
Funding
This project is funded by the UK Engineering and Physical Sciences Research Council grants (EP/S020950, EP/S014284/1, EP/S030875, and EP/N014642).. KM and CB were also supported by grants from the British Heart Foundation (PG/11/228474; PG/14/64/31043; FS/15/54/31639; CoE/RE/18/6134217).
Data availability statement
All quantities including the growth tensor estimated for all 16 patients are made available through the Supplemental Data.
CRediT authorship contribution statement
Wenguang Li: Segmentation, Reconstruction, Software, improvement, Data curation, Result analysis, Writing - original draft. Hao Gao: Methodology, Software, Data curation, Critical reviewing. Kenneth Mangion: Data curation, Clinical analysis, Critical reviewing. Colin Berry: Data curation, Clinical analysis, Critical reviewing. Xiaoyu Luo: Conceptualization, of this study, Result analysis, Critical reviewing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
We would like to thank Prof. Ray Ogden for helpful discussions.
References (56)
- et al.
Automated cardiovascular magnetic resonance image analysis with fully convolutional networks
J. Cardiovasc. Magn. Reson.
(2018) - et al.
Quantitative assessment of myocardial fibrosis in an age-related rat model by ex vivo late gadolinium enhancement magnetic resonance imaging with histopathological correlation
Comput. Biol. Med.
(2015) Modeling of cardiac growth and remodeling of myofiber orientation
J. Biomech.
(2012)- et al.
A randomized trial of deferred stenting versus immediate stenting to prevent no-or slow-reflow in acute st-segment elevation myocardial infarction (defer-stemi)
J. Am. Coll. Cardiol.
(2014) - et al.
Contrast-enhanced magnetic resonance imaging of myocardium at risk: distinction between reversible and irreversible injury throughout infarct healing
J. Am. Coll. Cardiol.
(2000) - et al.
Mechanisms of post-infarct left ventricular remodeling
Drug Discov. Today Dis. Mech.
(2007) - et al.
A multiscale model for eccentric and concentric cardiac growth through sarcomerogenesis
J. Theor. Biol.
(2010) - et al.
Disproportionate epicardial dilation after transmural infarction of the canine left ventricle: acute and chronic differences
J. Am. Coll. Cardiol.
(1988) - et al.
Stress-dependent finite growth in soft elastic tissues
J. Biomech.
(1994) - et al.
Standardized image interpretation and post processing in cardiovascular magnetic resonance: society for cardiovascular magnetic resonance (scmr) board of trustees task force on standardized post processing
J. Cardiovasc. Magn. Reson.
(2013)
Data mining framework for identification of myocardial infarction stages in ultrasound: a hybrid feature extraction paradigm (part 2)
Comput. Biol. Med.
Kinematics of cardiac growth: in vivo characterization of growth tensors and strains
J. Mech. Behav. Biomed. Mater.
Prediction of myocardial infarction by assessing regional cardiac wall in cmr images through active mesh modelings
Comput. Biol. Med.
Growth and remodelling of living tissues: perspectives, challenges and opportunities
J. R. Soc. Interface
Computational method for identifying and quantifying shape features of human left ventricular remodeling
Ann. Biomed. Eng.
Model of anisotropic reverse cardiac growth in mechanical dyssynchronyn
Sci. Rep.
A segmentation method for myocardial ischemia/infarction applicable in heart photos
Comput. Biol. Med.
Adverse remodeling and reverse remodeling after myocardial infarction
Curr. Cardiol. Rep.
Learning distributions of shape trajectories from longitudinal datasets: a hierarchical model on a manifold of diffeomorphisms
Cardiovascular magnetic resonance in acute st-segment–elevation myocardial infarction: recent advances, controversies, and future directions
Circulation
Defining left ventricular remodeling following acute ST-segment elevation myocardial infarction using cardiovascular magnetic resonance
J. Cardiovasc. Magn. Reson.
Remodeling of cardiac fiber structure after infarction in rats quantified with diffusion tensor mri
Am. J. Physiol. Heart Circ. Physiol.
Automatic 3d bi-ventricular segmentation of cardiac images by a shape-refined multi-task deep learning approach
IEEE Trans. Med. Imag.
Left ventricular remodeling in the post-infarction heart: a review of cellular, molecular mechanisms, and therapeutic modalities
Heart Fail. Rev.
Changes and classification in myocardial contractile function in the left ventricle following acute myocardial infarction
J. R. Soc. Interface
Left ventricular strain and its pattern estimated from cine CMR and validation with DENSE
Phys. Med. Biol.
Dynamic finite-strain modelling of the human left ventricle in health and disease using an immersed boundary-finite element method
IMA J. Appl. Math.
Parameter estimation in a Holzapfel–Ogden law for healthy myocardium
J. Eng. Math.
Cited by (11)
Using LDDMM and a kinematic cardiac growth model to quantify growth and remodelling in rat hearts under PAH
2024, Computers in Biology and MedicinePhysics-informed graph neural network emulation of soft-tissue mechanics
2023, Computer Methods in Applied Mechanics and EngineeringAn updated Lagrangian constrained mixture model of pathological cardiac growth and remodelling
2023, Acta BiomaterialiaImage-based estimation of the left ventricular cavity volume using deep learning and Gaussian process with cardio-mechanical applications
2023, Computerized Medical Imaging and GraphicsElectro-fluid-mechanics of the heart
2022, Journal of Fluid MechanicsGPU accelerated digital twins of the human heart open new routes for cardiovascular research
2023, Scientific Reports