Quantification of fibrosis in infarcted swine hearts byex vivolate gadolinium-enhancement and diffusion-weighted MRI methods

Abstract

Many have speculated that MRI signal characteristics can be used to identify regions of heterogeneous infarct associated with an arrhythmogenic substrate; however, direct evidence of this relationship is limited. The aim of this study was to demonstrate the remodelling characteristics of fibrosis by means of histology and high-resolution MR imaging. For this purpose, we performed whole-mount histology in heart samples (n = 9) collected from five swine at six weeks post-infarction and compared the extent of fibrosis in the infarcted areas delineated in these histological images with that obtained ex vivo by MRI using late gadolinium-enhancement (LGE) and diffusion-weighted imaging (DWI) methods. All MR images were obtained at a submillimetre resolution (i.e., voxel size of 0.6×0.6×1.2 mm(3)). Specifically, in the histology images, we differentiated moderate fibrosis (consisting of a mixture of viable and non-viable myocytes, known as border zone, BZ) from severe fibrosis (i.e., the dense scar). Correspondingly, tissue heterogeneities in the MR images were categorized by a Gaussian mixture model into healthy, BZ and scar. Our results showed that (a) both MRI methods were capable of qualitatively distinguishing sharp edges between dense scar and healthy tissue from regions of heterogeneous BZ; (b) the BZ and dense scar areas had intermediate-to-high increased values of signal intensity in the LGE images and of apparent diffusion coefficient in the DWI, respectively. In addition, as demonstrated by the Picrosirius Red and immunohistochemistry stains, the viable bundles in the BZ were clearly separated by thin collagen strands and had reduced expression of Cx43, whereas the core scar was composed of dense fibrosis. A quantitative analysis demonstrated that the comparison between BZ/scar extent in LGE and DWI to the corresponding areas identified in histology yielded very good correlations (i.e., for the scar identified by LGE, R(2) was 0.96 compared to R(2) = 0.93 for the scar identified in ADC maps, whereas the BZ had R(2) = 0.95 for the correlation between LGE and histology compared to R(2) = 0.91 obtained for ADC). This novel study represents an intermediate step in translating such research to the in vivo stages, as well as in establishing the best and most accurate MR method to help identify arrhythmia substrate in patients with structural heart disease.

Sympathetic modulation of electrical activation in normal and infarcted myocardium: implications for arrhythmogenesis

“…It is well known that myocardial scar leads to an increase in apparent diffusivity compared with healthy tissue (37,40) as well as a loss of directionality (i.e., anisotropy). In a map of fiber orientation, this will manifest as seemingly random orientations.…”

Cartan Frame Analysis of Hearts with Infarcts

“…Thus far little work has been carried out in reconstructing models of heart wall fibers and their geometry. The image analysis tools described in [25,24,26] try to come up with a solution to the problem of modeling heart wall fibers from diffusion MR images. This latter problem is important since this solution gives information on the heart wall mechanical function [14].…”

“…All DW-MR studies were performed on a dedicated 1.5T GE Signa Excite scanner using a custom FSE pulse sequence. They used the following MR parameters [26]:…”

“…Histology of a transverse slice of an infarcted heart analyzed in [26]. The healthy muscle tissue is represented in red and the scar tissue is stained in blue by a Masson Trichrome stain.…”

“…The data we used [26] are DT-MRI of porcine hearts. Of these 8 hearts had infarcts while 2 were healthy.…”