Understanding the Mechanistic Links Between Mid-Wall Fibrosis and Arrhythmic Risk in Non-Ischemic DCM Using a Combined Modelling & Clinical Approach

Lead Research Organisation: King's College London
Department Name: Imaging & Biomedical Engineering

Abstract

Non-ischemic dilated cardiomyopathy (NIDCM) is a common form of structural heart disease. It is characterised by enlargement of the heart's chambers, stretching and thinning the muscle walls, causing dilation. In many cases, the structure of the heart tissue itself changes with noticeable regions of fibrous tissue developing within the centre of the muscle wall. These structural changes associated with NIDCM can be seen and measured with the latest MR imaging techniques. Patients suffering from NIDCM have a very high chance of experiencing cardiac arrhythmias and associated sudden cardiac death. More specifically, a high-impact study from our Team showed a very clear link between these fibrous regions that develop within the centre of the heart wall in some NIDCM patients and death from arrhythmias. However, the reasons underlying this clear link between these structural changes associated with NIDCM and lethal arrhythmias is extremely poorly understood. Consequently, assessing this risk in patients and planning treatments remains a significant challenge. Understanding the link between structural changes identified in clinical MR data with the key physiological processes responsible for arrhythmias in NIDCM suggests the use of an image-based model to aid mechanistic interpretation. In this study, we aim to combine clinical MRI with advanced image analysis and computational modelling to improve our physiological understanding of the mechanisms underlying the link between changes in heart structure with cardiac arrhythmias. Gaining such knowledge will help to elucidate novel clinical risk-factors associated with NIDCM and help clinicians plan therapeutic interventions. Achieving our goals will provide more precise, informed decisions regarding the optimal way to treat an particular patient's heart, as well as the individual arrhythmia risk posed by a patient's NIDCM and the need for an implantable defibrillator, improving NIDCM survival rates and reducing health-service costs on unnecessary operations.

Technical Summary

Non-ischemic dilated cardiomyopathy (NIDCM) is a common, yet complex, form of structural heart disease characterised by pathological ventricular dilation and tissue remodelling. Despite its strong association with cardiac arrhythmia and sudden death, risk stratification in this population remains a significant challenge. Recent MR imaging studies have highlighted a strong correlation between identification of regions of mid-wall fibrosis in NIDCM patients and arrhythmia incidence and related death. However, the physiological mechanisms underlying these empirical findings are, as yet, unknown, limiting further application and use of this information for early identification of those patients at elevated arrhythmia risk. We aim to use state-of-the-art image processing and analysis techniques upon both extensive existing, and prospective, MR data cohorts to define specific structural and morphological quantitative metrics linking mid-wall fibrotic tissue remodelling with arrhythmic risk. Advanced computational simulations will be performed with idealised models to gain an in-depth understanding of how the presence of a mid-wall fibrotic region relates directly to arrhythmia burden as well as a detailed understanding of the parameters that govern these effects. A series of patient-specific image-based models will then be created, validated by in-vivo EP measurements from the prospective cohort, to provide key mechanistic insight underlying these associative links, specifically aiming to understand the relative importance of structural and electrophysiological remodelling upon arrhythmia risk. Uncovering these associative links between more specific structural metrics defining mid-wall fibrosis and arrhythmia risk, along with the in-depth mechanistic insight provided by the modelling, will help advance the direct use of these identified quantitative MR metrics to inform catheter ablation therapy and as a tool for clinical risk stratification and therapy optimisation.

Planned Impact

This project will have three major outcomes: provide a series of refined links between arrhythmic risk and, not just the presence of mid-wall fibrosis, but its specific anatomical distribution; provide detailed insight regarding the physiological mechanisms driving these processes and the resulting arrhythmia dynamics to guide therapies; to understand the relative importance of structural versus electrophysiological remodelling in these scenarios.

Basic Science
> Mechanistic Findings - Experimental animal models of NIDCM do not show the same changes in electrophysiology and fibrosis as witnessed clinically. Furthermore, knowledge from the clinic and donor hearts is heavily restricted due to the relative paucity of measurements. Thus, the use of advanced computational modelling, parameterised and validated with clinical measurements and previous donor heart data, provides the opportunity, for the first time, to gain a fully 3-dimensional understanding of the arrhythmia mechanisms at the whole ventricle level in NIDCM. We thus expect that the mechanistic findings from this work will be highly novel and unlike anything presented previously, facilitating a move towards more clinically-focussed basic research in this particular area.
> Technological Advances - The project will also make a number of important technological advances. The application of dictionary machine learning approaches to relate the defined imaging metrics with arrhythmia follow-up data is only possible due to the extensive nature of our retrospective cohort (>500 patient datasets), which has not been attempted before in this area. Secondly, representation of fibrosis using the discrete finite element method discussed in the proposal has not, as yet, been attempted at the ventricle level. As such, many of the technological details regarding the construction and parameterisation of ventricular models will provide important, novel methodological guidance, the publication of which will have important impact in the modelling community.
> Model Availability - The finite element computational models will be made freely available for public download, significantly increasing the impact of the research as they will be linked to associated papers.

Clinical
> Catheter Ablation - For the first time, our findings will help to characterise the nature of reentrant circuits in NIDCM, and their relationship with mid-wall fibrosis (imaged and located with LGE MR). In combination with anatomical information from LGE images, our findings may thus help clinicians interpret their (limited) EP recordings, guiding their decision making for the optimal ablation target.
> Risk Stratification - The increased knowledge provided by this work will pave the way to future use of specific metrics defining anatomical mid-wall fibrosis in NIDCM as a clinical guide to arrhythmic risk (e.g. combined to define an 'arrhythmic risk score'). This would provide clinicians with a useful tool to stratify patients, allowing more informed decisions to be made regarding ICD implantation, significantly advancing the currently-used LVEF measure, which is known to be inadequate in the setting of NIDCM. This would ultimately improve patient quality of life for those patients currently receiving unnecessary ICDs and would save the lives of those currently incorrectly deemed at low risk to not require a device.
> Drug Target Research - Combining these findings with the latest research uncovering a genetic basis for NIDCM, will contribute significantly to earlier recognition and intervention of at-risk patients. In the longer-term, acquiring detailed knowledge of the electrophysiological and arrhythmogenic properties of fibrotic tissue in different pathologies will be imperative in order to understand, and optimise, the functioning and potential therapeutic benefit of the latest anti-fibrotic agents currently under development, using the tools developed in this work as a computational testbed.

Publications

10 25 50

publication icon
Connolly AJ (2017) Bidomain Predictions of Virtual Electrode-Induced Make and Break Excitations around Blood Vessels. in Frontiers in bioengineering and biotechnology

 
Description Brompton midwall fibrosis 
Organisation Royal Brompton Hospital
Country United Kingdom 
Sector Hospitals 
PI Contribution We provide computational modelling assistance, utilising MR data and measurements.
Collaborator Contribution Brompton researchers provide MRI data and measurements as well as clinical assistance.
Impact Joint MRC grant and associated outputs.
Start Year 2016
 
Description "Birmingham Big Bang Science Fair", STEM panellist to act as a role model and participate in 20-minute career talks, (March 13) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact "Birmingham Big Bang Science Fair", STEM panellist to act as a role model and participate in 20-minute career talks, (March 13)
Year(s) Of Engagement Activity 2019
 
Description "I am a Scientist: get me Outta Here!", online chat with high school students (March 5-16, 2018) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact "I am a Scientist: get me Outta Here!", online chat with high school students (March 5-16, 2018)
Year(s) Of Engagement Activity 2019
 
Description King's College London Christmas School Lecture 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact King's College London Christmas School Lecture. The lecture I gave was uploaded to YouTube later as part of KCL media activities.
Year(s) Of Engagement Activity 2017
 
Description Pint of Science Event 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Public/other audiences
Results and Impact Pint of Science event - local event to engage lay audience in research over a drink. Consists of ~20min talk and questions.
Year(s) Of Engagement Activity 2019
 
Description School visit (various) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? Yes
Geographic Reach National
Primary Audience Schools
Results and Impact Approximately 5 talks have been given, with audiences between 20-100 each time, speaking about Biomedical Engineering in general, getting students interested in applying to these courses for University.

Many students applied for our BEng course having said they were inspired by such talks and visits.
Year(s) Of Engagement Activity 2013