G2 - Image-based analysis of oxygen transport impairment in heart failure

Introduction: : Myocardial reactive interstitial fibrosis (RIF) is characterized by excessive deposition of extracellular matrix (ECM) proteins, such as collagen, in the interstitium[1] . This pathological process can occur in both ischemic and nonischemic heart failure (HF)[2,3], and is strongly associated with diastolic dysfunction as in HF with preserved ejection fraction (HFpEF). Recent studies suggest that structural changes associated with RIF are associated with reduced capillary density and increased oxygen diffusion distance[4,5], which in turn, can impair oxygen (O2) delivery and cause regional hypoxia. On the other hand, mismatch between O2 supply and demand may occur as fibrotic tissue may decrease the tissue O2 consumption rate due to reduced metabolic activity and the cardiomyocytes may compensate by increasing metabolic workload[4,6,7]. The precise impact of structural and metabolic changes on O2 diffusion and utilization remain poorly understood in RIF. Specifically, it is unclear how and to what extent RIF and interstitial space enlargement affect O2 delivery to the myocytes in HF. To address this issue, we developed a novel image-based computational framework that can compute O2 diffusion and consumption directly from histological images acquired at the cellular/capillary levels. Using this framework, we quantified the effects of interstitial space enlargement and RIF on O2 transport in left ventricular (LV) myocardium of healthy dogs and dogs with HF to understand the impact of this pathological process.

Materials and

Methods: : The image-based computer modeling framework is developed using the open-source FE library (FEniCS[8]), to quantify O2 concentration directly from histological images showing the spatial distribution of capillaries, myocytes and interstitial spaces.

Image data processing:
High-resolution histological images of myocardial tissue slices of both healthy and failing LV myocardium (with varying degrees of fibrosis - mild, moderate and severe) were acquired from a canine HF model. The images were segmented to identify subdomains corresponding to capillaries, myocytes and interstitial space. The entire segmented image was used to define the computational finite element (FE) mesh, where each square pixel was subdivided into two triangular elements that were encoded with the subdomain (capillary, myocyte or interstitial space) information. To ensure accurate representation of the image, interfaces between the subdomains were further refined by subdividing the elements. Elements corresponding to blood-filled capillaries were excluded from the computation to improve efficiency.

Numerical Model:
Oxygen transport from the capillaries to the cardiomyocytes was modeled based on a reaction-diffusion system with Michaelis-Menten kinetics describing the O2 consumption[9]. Figure 1 illustrates the computational framework and the governing equation that describes O2 concentration in the tissue domain. Capillary wall thickness was assumed to be uniform at 1 μm. An O2 concentration of 25 mmHg was prescribed at the inner surface of the capillary wall, while a zero-flux boundary condition was applied to the outermost surface to simulate a closed system. Timestep for simulations was chosen based on the smallest grid size and highest diffusivity to avoid numerical instability.

Results, Conclusions, and Discussions:: We use the above computational framework to analyze 5 healthy and 5 HF myocardium cases. In the framework, morphological changes reflected on the images are directly incorporated into the FE meshes, while functional changes are considered by modifying parameters associated with O2 diffusion (D) and metabolic rates (Gmax) in the HF cases. Myocardial consumption rate in viable tissue was prescribed to be increased by 50% in the HF cases based on a previous study[10]. In fibrotic regions, oxygen diffusivity was reduced by 50–90%, and consumption rates were prescribed to be 30–70% lower compared to healthy tissue. By integrating these morphological and functional changes, we evaluated the impact of RIF on myocardial O2 concentration.
Histological images from the HF cases showed that the interstitial space fraction (that increases O2 diffusion distance) is increased to ~29% (compared to ~16% in normal myocardium). When combined with functional changes, i.e., with increased metabolic load in the myocytes as well as decreased O2 diffusivity and metabolic rate in the interstitial space, a substantial ~ 25% reduction in O2 concentration in the tissue domain was observed. These findings suggest that an increase in RIF significantly impairs oxygen delivery even as viable myocardium compensates to meet the O2 demand by increasing its metabolic load.
In summary, we developed a robust computational framework that quantify O2 concentration and transport dynamics from histological images showing the myocytes, interstitial space and capillaries. This pipeline can be extended as an image-based analysis tool to quantify changes in O2 metrics from histological images acquired for other diseases and organs.

Acknowledgements and/or References (Optional):: References

1. Díez J, González A, Kovacic JC. Myocardial Interstitial Fibrosis in Nonischemic Heart Disease, Part 3/4: JACC Focus Seminar. J Am Coll Cardiol. 2020;75:2204-2218. doi:10.1016/j.jacc.2020.03.019
2. Boneva T, Groudeva V, Karamfiloff K. Myocardial interstitial fibrosis in different cardiomyopathies as a basis of heart failure. Bulgarian Cardiology. 2024;30:40-50. doi:10.3897/bgcardio.30.e128810
3. González A, Schelbert EB, Díez J, Butler J. Myocardial Interstitial Fibrosis in Heart Failure: Biological and Translational Perspectives. J Am Coll Cardiol. 2018;71:1696-1706. doi:10.1016/j.jacc.2018.02.021
4. Sabbah HN, Sharov VG, Lesch M, Goldstein S. Progression of heart failure: A role for interstitial fibrosis. Mol Cell Biochem. 1995;147:29-34. doi:10.1007/BF00944780
5. Lunde IG, Rypdal KB, Van Linthout S, Diez J, González A. Myocardial fibrosis from the perspective of the extracellular matrix: Mechanisms to clinical impact. Matrix Biology. 2024;134:1-22. doi:10.1016/j.matbio.2024.08.008
6. Camici PG, Tschöpe C, Di Carli MF, Rimoldi O, Van Linthout S. Coronary microvascular dysfunction in hypertrophy and heart failure. Cardiovasc Res. 2020;116:806-816. doi:10.1093/cvr/cvaa023
7. Campbell DJ, Francis VCM, Young GR, Woodford NWF. Association of Coronary Microvascular Rarefaction and Myocardial Fibrosis With Coronary Artery Disease. J Am Heart Assoc. 2024;13:e037332. doi:10.1161/JAHA.124.037332
8. Baratta IA, Dean JP, Dokken JS, et al. DOLFINx: The next generation FEniCS problem solving environment. doi:10.5281/ZENODO.10447666
9. Beard DA, Bassingthwaighte JB. Modeling Advection and Diffusion of Oxygen in Complex Vascular Networks. Ann Biomed Eng. 2001;29(4):298-310. doi:10.1114/1.1359450
10. Strauer BE. Left ventricular dynamics, energetics and coronary hemodynamics in hypertropic heart disease. Eur Heart J. 1983;4(Suppl. A):137-142. doi:10.1093/EURHEARTJ/4.SUPPL_A.137