Abstract

Fiber structures and pathological features, e.g., inflammation and glycosaminoglycan (GAG) deposition, are the primary determinants of aortic mechanical properties which are associated with the development of an aneurysm. This study is designed to quantify the association of tissue ultimate strength and extensibility with the structural percentage of different components, in particular, GAG, and local fiber orientation. Thoracic aortic aneurysm (TAA) tissues from eight patients were collected. Ninety-six tissue strips of thickened intima, media, and adventitia were prepared for uni-extension tests and histopathological examination. Area ratios of collagen, elastin, macrophage and GAG, and collagen fiber dispersion were quantified. Collagen, elastin, and GAG were layer-dependent and the inflammatory burden in all layers was low. The local GAG ratio was negatively associated with the collagen ratio (r2 = 0.173, p < 0.05), but positively with elastin (r2 = 0.037, p < 0.05). Higher GAG deposition resulted in larger local collagen fiber dispersion in the media and adventitia, but not in the intima. The ultimate stretch in both axial and circumferential directions was exclusively associated with elastin ratio (axial: r2 = 0.186, p = 0.04; circumferential: r2 = 0.175, p = 0.04). Multivariate analysis showed that collagen and GAG contents were both associated with ultimate strength in the circumferential direction, but not with the axial direction (collagen: slope = 27.3, GAG: slope = −18.4, r2 = 0.438, p = 0.002). GAG may play important roles in TAA material strength. Their deposition was found to be associated positively with the local collagen fiber dispersion and negatively with ultimate strength in the circumferential direction.

1 Introduction

Thoracic aortic aneurysms (TAAs) are significant contributors to deaths [1]. TAAs are mostly asymptomatic during diagnosis as lesions are commonly identified unexpectedly during an echocardiography or computed tomography, or when taking an X-ray image of the chest. Thoracic aneurysms have a mean growth rate of 0.1 cm/year, where this value is location- and patient-dependent [2]. Moreover, the rupture rate of lesions ≥6.0 cm was found to be at least 7%, which is more than 25 times higher than that of the lesions with diameters between 4.0 cm and 4.9 cm2. However, lesions with less than 6 cm diameter were also found to show a tendency to rupture or to develop into a dissection [3]. These findings strongly suggest that close radiological monitoring is critical for identifying the lesions that show rapid growth despite being under the surgical repair threshold.

The main pathological feature of TAAs is cystic medial degeneration, which is commonly accelerated with other clinical conditions such as hypertension, bicuspid aortic valve aortopathy, and genetic mutations. Degeneration of the media layer and the deposition of proteoglycans and thereafter glycosaminoglycans (GAG) often take place simultaneously in the aorta wall, which is also accompanied by fragmentation of the elastic lamellae [4]. GAG are long unbranched polysaccharides that consist of a repeating disaccharide unit. Despite the fact that GAG makes up a small portion of the artery wall (2%–5% by dry weight), they play key roles in both physiological and the pathological processes that are undergone by a vessel [5,6] and also contribute to the mechanical behavior of the tissue [7]. In its healthy state, the arterial wall undergoes mechanical loading as a result of the dynamic blood pressure and flow. Undergoing such continuous mechanical loading is likely to modulate and have adverse effects on the biological and biochemical physiology of the endothelium [8] and smooth muscle cells [9] of the arterial wall. In addition, following degeneration, the vessel wall may become more likely to rupture or dissect as a result of the dynamic mechanical loading. These factors point toward a clear need for obtaining a better understanding of the material properties and mechanical strength of TAA and any association between its mechanics and structural composition as well as microstructural features.

The major load-bearing components of the arterial extracellular matrix (ECM) are elastin and collagen fibers [10]. Their orientations and structural organizations are associated with the strength of arteries in both health and disease [11]. Existing literature on the mechanical behavior of arterial ECM has so far commonly targeted collagen and elastin, where the role of GAG has often been neglected. Despite being a minor constituent of the ECM, GAG could play a significant role in the development of dissections [12] as excessive accumulation and aggregation of GAG could lead to high levels of local stress concentrations that can cause delamination of the artery wall [13]. Moreover, a reduced GAG content is known to lead to earlier collagen and elastin fiber recruitment in porcine thoracic aortas, giving the aortas a stiffer mechanical behavior [14].

In this study, TAA tissue strips were tensile tested to quantify the ultimate stretch and material strength of the tissues, histopathological investigations were performed to quantify the structural area occupied by distinct microstructural components, and statistical analyses were performed to explore the correlations between the mechanical properties of the tissue and its detailed composition, which also includes GAG.

2 Materials and Methods

2.1 Collection of Tissues.

Eight TAA tissue pieces were collected from eight patients (one female; ages were 61.4±12.2 years three from ascending aorta, four from the aortic arch and one from the descending aorta; three were chronic dissections and five fusiform aneurysms), who underwent open aortic repair in Royal Papworth Hospital, NHS Foundation Trust, Cambridge, UK. None of the lumens showed thrombus presence that has been confirmed by the histologic examination. The local ethics committee approved this study and written patient consents were obtained.

2.2 Mechanical Testing.

Tissue pieces were cut into rings (circumferential sections) (Fig. 1(a)) or axial sections shortly after the tissue collection. The thickened intima, media, and adventitia of each arterial piece were separated (Fig. 1(b)). Tissue strips from each layer, with ∼1.5 mm width and ∼15 mm length, along both axial and circumferential directions to the blood flow were prepared and the width, thickness, and length of each tissue strip were documented (Figs. 1(c)1(e)) as detailed in the Supplemental Materials on the ASME Digital Collection. Two water-proof markers were placed on the tissue surface to trace the local displacement (Fig. 1(f)). Each tissue strip was mounted on an in-house designed microtester to perform a uniaxial tensile test. The loadcell reading was set to zero when the tissue strip was loose, and then a preload of 0.001 N was applied to achieve the starting state. Before testing, preconditioning with ∼5% stretch at a speed of 0.05 mm·s−1 was repeated five times and the strip was then stretched until failure at a speed of 0.01 mm·s−1 in a 37 °C phosphate-buffered saline bath. New tissue strips would be prepared and tested again when strips failed near one of the clamps.

Fig. 1
Representative aneurysmal ring (a) and isolated tissue strips (c)–(e) for mechanical testing and a tissue strip under stretching (f; markers were placed on the surface to trace the local displacement to compute the local stretch ratio; markers were identified as enclosed by the lines automatically; (g) a representative displacement-force curve; the gap between adjacent lines in (a)–(e) is 0.5 mm): (a) a intact ring, (b) isolated tissue strip, (c) adventitia, (d) media, (e) thickened intima, (f) markers, and (g) the curve of displacement-force
Fig. 1
Representative aneurysmal ring (a) and isolated tissue strips (c)–(e) for mechanical testing and a tissue strip under stretching (f; markers were placed on the surface to trace the local displacement to compute the local stretch ratio; markers were identified as enclosed by the lines automatically; (g) a representative displacement-force curve; the gap between adjacent lines in (a)–(e) is 0.5 mm): (a) a intact ring, (b) isolated tissue strip, (c) adventitia, (d) media, (e) thickened intima, (f) markers, and (g) the curve of displacement-force
Close modal

During stretching, the force and tissue images were recorded by a sensor and a camera, respectively, and the displacement of the clamp was also recorded (Fig. 1(g)). The microtester comprises a uniaxial stepper motor (miniature steel motorized linear stage, Newport Corporation, Irvine, CA), a custom-made linear load cell, a camera (PixeLink PL-B776 U 3.1 MP USB2 Color Camera, PixeLink, Ottawa, ON, Canada), and a custom control interface developed in labview (National Instruments, Austin, TX). The position resolution of the stepper motor was 0.1 μm and the precision of the load cell was 0.0005 N. The size of each image frame was 2048 × 1536 pixels, with an 80 × 60 mm2 field of view. The marker was in dark blue and it had a strong contrast with white tissue strips and their background. A dark blue color threshold and standard morphological image operations available in matlab (imclearborder, imerode, imfill, and imclose functions of matlab) were utilized to generate marker spot binary masks from the acquired red-green-blue color space images. The described mechanical testing protocol has been developed, optimized, and used in previous studies [15].

2.3 Calculation of Stress-Stretch.

A semi-automatic matlab platform was developed to compute the local displacement of each arterial tissue from images acquired at the end of each stepper motor displacement increment. The markers on the tissue surface were recognized automatically to determine their centroid coordinates (Fig. 1(f)). The distance between the centroids of the two markers was used to compute the local stretch, λi = li/l0, in which li is the distance between the centroid of the two markers at the ith increment and l0 is the distance at rest. Cauchy stress was computed using, σi = (Fi·λi)/A0, where Fi was the measured force at each displacement increment, A0 was the initial cross-sectional area of the arterial tissue strip. The stress–stretch of each tissue strip was shown in Fig. S2 available in the Supplemental Materials on the ASME Digital Collection. The tissue ultimate stretch and material strength were defined, respectively, as the stretch ratio and the stress before failure as indicated by the sudden drop in the force–displacement curve (Fig. 1(g)).

Fig. 2
Histopathological sections showing tissue local fiber structure and pathological features. (a) hematoxylin and eosin stain (H&E) for an overall assessment; (b) Sirius red stain showing collagen fibers; (c) elastin Van Gieson (EVG) stain showing elastin; (d) cluster of differentiation 68 (CD68) stain for macrophages; and (e) Alcian blue stain for GAG). Area ratios of the corresponding microstructural components are also provided at the bottom left of each histopathological slide image: (a) H&E, (b) Sirius red, (c) EVG, (d) CD68, and (e) Alcian blue.
Fig. 2
Histopathological sections showing tissue local fiber structure and pathological features. (a) hematoxylin and eosin stain (H&E) for an overall assessment; (b) Sirius red stain showing collagen fibers; (c) elastin Van Gieson (EVG) stain showing elastin; (d) cluster of differentiation 68 (CD68) stain for macrophages; and (e) Alcian blue stain for GAG). Area ratios of the corresponding microstructural components are also provided at the bottom left of each histopathological slide image: (a) H&E, (b) Sirius red, (c) EVG, (d) CD68, and (e) Alcian blue.
Close modal

Compared with the ultimate stretch and material strength, the stress–stretch curve at a low-stress level reflects the tissue physiologic mechanical behavior better. In this study, Young's modulus (E50) defined by the slope of a straight line approximated the stress–stretch curve at a stress level of [0, 50] kPa was calculated. This stress range was chosen as the curve in this range could be well approximated linearly.

2.4 Histology and Image Processing.

Tissue strips adjacent to those used for mechanical testing were collected for immunohistochemical examination. Following standard processing, each tissue strip was embedded in paraffin wax, cut into 4 μm thick slices, and stained with hematoxylin and eosin (H&E), sirius red, elastin van Gieson (EVG), alcian blue and cluster of differentiation 68 (CD68) to visualize collagen, elastin, GAG, calcium, and macrophages, respectively (Fig. 2). Each immunohistochemically stained slice was digitized using Nano-Zoomer (Hamamatsu, Hamamatsu City, Japan) at a magnification rate of 40×. Collagen appears red in Sirius Red, elastin dark purple in EVG, GAGs blue in alcian blue staining, and macrophages appear brown in CD68. A semi-automatic platform developed in matlab was used to compute the area percentage of each microstructural component for each histopathological slide.“
areapercentage=numberofmicrostructuralcomponentpixelstotalnumberofpixelsintheregionofinterest×100%

Collagen fiber is a mechanically important tissue constituent, which bears loading at higher strains and determines the strength of the tissue [16]. Along with the overall area percentages, the local distribution of collagen fiber orientations was computed with the developed matlab platform. Fiber dispersion was used to describe the spread of fiber orientations, where smaller dispersion implies stronger alignment. In this study, the standard deviation of fiber orientation in a region was used to characterize the local fiber dispersion. Considering the heterogeneous distribution of the microstructural components, two hundred region of interest squares with 200 × 200 pixels were randomly placed in each histology image to quantify the local association between fiber dispersion and GAG content. Details of image processing can be found in Ref. [17] and the Supplemental Materials on the ASME Digital Collection.

2.5 Statistical Analysis.

Since multiple observations of GAG ratio and fiber dispersion were taken from each strip, linear mixed-effects models were used to account for the data hierarchy. In addition, to better reflect the heterogeneity between strips, the models assumed both random intercepts and random slopes. The nonsimultaneous marginal predictions and their 95% confidence intervals (CIs) were then plotted to illustrate the population trend.

Correlations between percentages of microstructural constituents (collagen, elastin, GAG, and macrophages) and the mechanical properties were evaluated using a linear model with the consideration of random effect using least-squares fit. For each estimated coefficient, the coefficient of determination (r2) and the corresponding p-value for t-statistic was reported.

Differences in microstructural constituent percentages and mechanical properties between different tissue types were evaluated using the Wilcoxon signed-rank test. Statistical analysis was performed in matlab (MathWorks, Inc.). A significant difference was assumed if p < 0.05.

3 Results

Data from 96 tissue strips of eight aneurysm samples were prepared for testing and analyses: 48 strips for material testing and 48 matched strips for the histopathological examination. Each 48-group is composed of one strip from each of thickened intima, media, and adventitia in the axial direction and one in the circumferential direction from each sample, that is, 12 tissues from each aneurysm sample and six for the material testing, and six for the histopathological examination.

3.1 Layer-Specific Component Contents.

Collagen and elastin were found to be the two main constituents of the aortic wall, followed by GAG and macrophages (Fig. 3). Among the three layers, the thickened intima was found to have the lowest collagen content, and the adventitia was found to have the highest collagen content; in contrast, the intima was found to have more elastin than the adventitia and the elastin area ratios of intima and media were found to be comparable. Both intima and media were found to have higher GAG contents than adventitia (p < 0.001). Compared with collagen, elastin, and GAG, the macrophage area ratio in the arterial wall was significantly less (p < 0.001). Correlation analysis showed that GAG area ratio was negatively associated with the collagen area ratio (r2 = 0.173, p < 0.001) but positively with elastin (r2 = 0.037, p = 0.038).

Fig. 3
Area ratio of collagen, elastin, GAG and macrophages in different layers
Fig. 3
Area ratio of collagen, elastin, GAG and macrophages in different layers
Close modal

3.2 Layer-Specific Collagen Fiber Architectures and Their Association With Glycosaminoglycan Deposition.

The association between local fiber dispersion and GAG deposition was explored. Collagen fiber dispersion in the intima was 43.5 deg [40.8 deg, 44.4 deg] (median [interquartile range]) in the circumferential and 42.3 deg [40.2 deg, 44.6 deg] in the axial direction, with no significant difference between these two directions (p = 0.677). Fiber dispersion in the media was smaller in the circumferential than in the axial direction (39.4 deg [36.2 deg, 40.7 deg] versus 44.2 deg [42.5 deg, 46.2 deg], p = 0.001).

Across layers, fiber dispersion of the media in the circumferential direction was smaller than in the intima (p = 0.003). Local GAG content was positively associated with fiber dispersion in the media and adventitia in both circumferential and axial directions, but not in the intima in either direction (Fig. 4).

Fig. 4
Correlation analysis using linear mixed-effects models showing the local correlation between dispersion of collagen fiber orientations and GAG area ratio in different layers in both axial and circumferential directions. The nonsimultaneous marginal predictions and their confidence intervals are plotted to illustrate population-wise association: (a) intima in the axial direction, (b) intima in the circumferential direction, (c) medial in the axial direction, (d) media in the circumferential direction, (e) adventitia in the axial direction, and (f) adventitia in the circumferential direction.
Fig. 4
Correlation analysis using linear mixed-effects models showing the local correlation between dispersion of collagen fiber orientations and GAG area ratio in different layers in both axial and circumferential directions. The nonsimultaneous marginal predictions and their confidence intervals are plotted to illustrate population-wise association: (a) intima in the axial direction, (b) intima in the circumferential direction, (c) medial in the axial direction, (d) media in the circumferential direction, (e) adventitia in the axial direction, and (f) adventitia in the circumferential direction.
Close modal

3.3 Layer-Specific Mechanical Properties.

Failure tests showed that the adventitia in the circumferential direction had the highest ultimate material strength (Fig. 5(a)). The material strength of intima in the axial direction was similar to that of media in the same direction (395.9 [202.3, 567.2] versus 236.4 [213.3, 341.4], p = 0.31; unit: kPa), where the intima and media strengths were also similar in the circumferential direction (518.4 [249.5, 750.3] versus 606.5 [358.8, 708.2], p = 0.84; unit: kPa). Across layers, the intima had similar material strengths in both directions (p = 0.46), while the media and adventitia were stronger in the circumferential than the axial direction (both with p = 0.02). In general, the ultimate stretch of all layers in both directions was similar, except for the adventitia in the circumferential direction, which was more extendable than media in the same direction (1.33 [1.21, 1.54] versus 1.16 [1.14, 1.33], p = 0.02) (Fig. 5(b)).

Fig. 5
Ultimate material strength and stretch of tissue strips
Fig. 5
Ultimate material strength and stretch of tissue strips
Close modal

3.4 Relationship Between Component Contents and Tissue Mechanical Properties.

Correlation analyses showed that the ultimate stretches in both axial and circumferential directions were associated only with elastin area ratio (Axial: r2 = 0.186, p = 0.04; Circumferential: r2 = 0.175, p = 0.04). The ultimate material strength in the circumferential direction associated with collagen positively, while with GAG negatively (Collagen: slope = 32.7, r2 = 0.328, p = 0.003; GAG: slope = −25.5, r2 = 0.227, p = 0.019), but not with strength in the axial direction. Further multivariate analysis confirmed the association between the material strength in the circumferential direction with collagen and GAG contents (collagen: slope = 27.3, GAG: slope = −18.4, r2 = 0.438, p = 0.002); and this association remained when data from both axial and circumferential directions were pooled (collagen: slope = 17.0, GAG: slope = −12.6, r2 = 0.249, p = 0.002).

Univariable analyses have shown that E50 was associated with a different area ratio of different components in the axial and circumferential direction differently (Table S1 available in the Supplemental Materials on the ASME Digital Collection). When data from different directions were pooled, no significant association was found. When data from three layers in the two directions were pooled, a significantly negative correlation between E50 and the elastin area ratio was found (r = −0.29, p = 0.047), but such a significant association was not found when the multivariable regression analysis was performed.

4 Discussion

This study showed that distributions of collagen, elastin, GAGs, and macrophages were layer-dependent and GAG deposition was associated with local fiber dispersion. Mechanical tests demonstrated that collagen and GAG contents were associated with tissue ultimate stretch and material strength.

In addition to fiber type and other material contents, fiber orientation is an important determinant for tissue extensibility and strength. However, the fiber orientation is best characterized in the three-dimensional setting, such as via the use of multiphoton microscopy [18]. The two-dimensional (2D) histological slides as used in this study would provide the information of projected fiber orientations on a certain slide plane. Due to the limitation of 2D projection, the fiber dispersion rather than the concrete spatial orientations were used when the relationship between GAGs and the fiber microstructure was accessed.

The negative relationship between GAG content and ultimate strength and the positive relationship between GAG content and the ultimate stretch in the circumferential direction implies that GAGs are a weaker material than collagen. This is in agreement with their molecular structures that consist of linear chains of repeating disaccharide units [19]. GAG is highly negatively charged and thus sequester water and contribute directly to the compressive, rather than tensile stiffness of the artery [12,19]. The negatively charged GAG could alter the regional distribution of interstitial water causing a swelling pressure, which has been speculated to be similar in magnitude to blood pressure [12]. Such swelling might damage the arterial wall by separating the fiber lamellae [20,21] as evidenced by greater separation distances between elastic lamellae where the accumulation of GAG was found [22], which might lead to the development of aortic dissection.

Apart from the inherent material characteristics of GAG, which can modify the local structure and strength, the deposition of GAG pools in the arterial wall also alters the local mechanical environment. Due to the lower tensile stiffness, pooled GAG deposition could elevate the local stress concentration by 3–5 times as compared with the far-field stresses, which are not affected by the local pooling of GAGs [13]. This GAG pooling induced stress concentration and swelling pressure might cause local delamination in the aortic wall and therefore contribute to the development of aortic dissection. GAG deposition may also play an important role in atherosclerotic plaque erosion in coronary circulation. Plaque erosion involves the formation of thrombus in an area of endothelial denudation adjacent to an atherosclerotic plaque without disruption of the fibrous cap [23,24] and is responsible for ∼25% of deaths due to acute myocardial infarction [24]. Histopathological analyses have demonstrated the superficial accumulation of specific GAGs and hyaluronan at the sites of plaque erosion (Fig. 6) [23]. As shown in this study, the GAG accumulation weakens the local material, implying that the superficial accumulation of GAGs can elevate the local stretch (Figs. 6(b) and 6(c)). The elevated stretch level might promote local endothelium cell apoptosis [8] leading to endothelial denudation. It also induces specific changes in the synthesis and organization of GAG by vascular smooth muscle cells [25], e.g., large stretch increases versican, biglycan, and perlecan core proteins, and this generates a positive feedback loop further promoting the accumulation of GAG. Moreover, given the large stretch together with weaker material, the GAG accumulation might damage the local tissue integrity by forming microfissures. The endothelial denudation and microfissures alone or in combination might lead to the local development of thrombosis and fatal events [26,27]. The mechanical analysis shown in Figs. 6(b) and 6(c) have followed previous reports [28] and mechanical properties of each component were adopted from previous direct measurements [15]. Detailed numerical processes that were used to produce Fig. 6 can be found in the Supplemental Materials.

Fig. 6
A 2D finite element analysis showed that superficial GAG deposition increased local stretch levels (a) the finite element model was reconstructed based on the Movat stainingof a coronary atherosclerotic lesion from a 38-year-old woman in Fig. 5 from Farb et al. [23]; (b) mechanical analysis demonstrated that the superficial GAG deposition induced increased local tissue stretch particularly around the lumen region; (c) when no GAG deposition was assumed, the stretch level reduced by over 10%; details about the analysis can be found in the Supplemental Materials
Fig. 6
A 2D finite element analysis showed that superficial GAG deposition increased local stretch levels (a) the finite element model was reconstructed based on the Movat stainingof a coronary atherosclerotic lesion from a 38-year-old woman in Fig. 5 from Farb et al. [23]; (b) mechanical analysis demonstrated that the superficial GAG deposition induced increased local tissue stretch particularly around the lumen region; (c) when no GAG deposition was assumed, the stretch level reduced by over 10%; details about the analysis can be found in the Supplemental Materials
Close modal

Tissues from TAA seem to be less extensible and stiffer than those from an abdominal aortic aneurysm, for which quantitative comparisons are provided in the table in Supplemental Materials. This study also showed a heterogeneous transmural distribution of GAG. Both intima (area percentage, 19.7% [13.6%, 24.7%]) and media (17.4% [10.8%, 25.7%]) had a significantly higher GAG content than adventitia (4.1% [1.6%, 7.8%]) where similar distribution patterns have also been reported in the healthy arterial wall [29]. However, in healthy arterial wall GAG occupy a small percentage of the total extracellular matrix (4%), but increase dramatically in the early lesion phase, e.g., ∼50% in the atherosclerotic lesion, and as lesions become more advanced, collagen predominates with decreased GAG content (∼20%) [30]. The heterogeneous transmural distribution may be a source of residual stress [29] which can present itself as a radial cut on an arterial ring [31] or as a radial cut on an atherosclerotic lesion [32]. The interactions between GAGs and other extracellular components also play an important role in the mechanics of the arterial wall. Although removal of GAGs results in an earlier transition point of the nonlinear stress–strain curves during extension mechanical testing, the stiffness was not significantly different after GAG removal [14]. Microscopic examination showed that when GAGs were removed, the adventitial collagen fibers were straighter, and both elastin and collagen fibers were recruited at lower levels of strain, in agreement with the mechanical change [14].

Sodium magnetic resonance imaging (MRI), delayed gadolinium-enhanced MRI of cartilage and T1-rho mapping have been developed to permit the visualization of the charged GAG distribution in cartilage [33]. These imaging techniques might be modified for the detection of GAG accumulation in the aortic wall, particularly in the thoracic aorta where most dissections are located. Successful development of this aspect would be beneficial for the prediction of dissection development.

Despite obtaining interesting findings, limitations exist in this study: (1) samples were obtained from the arch region where diseased tissues were mostly located. However, tissue structure and properties vary across the thoracic aorta. Studies with a bigger sample size are, therefore, needed to confirm findings reported in this study; (2) although uniaxial tension tests were performed in both circumferential and axial directions, the anisotropic material behavior of aneurysmal tissues cannot be assessed comprehensively; (3) no controls were provided for comparison since it is unethical to obtain healthy human aortic tissues; (4) although tissues used for mechanical tests and histopathological examination were from a narrow adjacent region, the heterogeneity of aneurysmal tissues means that measurements from adjacent tissue strips might not represent the inherent mechanical and histopathological features; (5) the fiber dispersion was used to characterize the fiber architecture instead of the absolute fiber orientation as the local curvature and orientation of each tissue strip would not affect the dispersion. However, the nonuniform distortion induced by the processing procedure for histopathological stains could affect the fiber dispersion analysis. The effect of this has been reduced by the utilized small square patch dimension (0.18 × 0.18 mm2; please see the Supplemental Materials on the ASME Digital Collection for more details). Moreover, the fiber dispersion was quantified in 2D, not three-dimensional; (6) the type of collagen fibers was not differentiated and considered, and the possible effect of fiber waviness on the calculation of its dispersion was not assessed; and (7) further studies with a bigger sample size are needed to explore the association of GAG contents with patient clinical characteristics.

Acknowledgment

Dr. Tokgoz is supported by an EPSRC Doctoral Training Award at the University of Cambridge and Dr. Wang by the China Scholarship Council.

Funding Data

  • British Heart Foundation (Grant No. PG/18/14/33562; Funder ID: 10.13039/501100000274).

  • Engineering and Physical Sciences Research Council (Grant No. EP/P021654/1; Funder ID: 10.13039/501100000266).

  • NSERC (Grant No. 6799-427538-2012; Funder ID: 10.13039/501100000038).

  • NIHR Cambridge Biomedical Research Centre (Grant No. BRC-1215-20014; Funder ID: 10.13039/501100018956).

  • China Scholarship Council (Funder ID: 10.13039/501100004543).

Contributions

AT and SW performed the testing and analyzed the data; PS collected the samples and interpreted their physiologic and pathologic features; CS and YH analyzed the data; NLF performed the histopathological examinations and interpreted the pathological features; MRB, SS, and JHG provided facilities and supports for the mechanical testing and histopathological examinations and interpreted the pathologic findings; MPFS designed the study and revised the paper significantly; ZT designed and supervised the whole study and drafted the paper.

Conflict of Interest

Other authors do not have any conflict of interest related to this study to declare.

Nomenclature

Abbreviations

    Abbreviations
     
  • CD68 =

    cluster of differentiation 68

  •  
  • CI =

    confidence interval

  •  
  • ECM =

    extracellular matrix

  •  
  • EVG =

    elastin Van Gieson

  •  
  • GAG =

    glycosaminoglycan

  •  
  • H&E =

    hematoxylin and eosin

  •  
  • MRI =

    magnetic resonance imaging

  •  
  • TAA =

    thoracic aortic aneurysm

References

1.
Jae
,
L. H. S.
,
2011
,
Epidemiology and Natural History of Thoraco-Abdominal Aortic Aneurysms
,
Springer
,
Milano, Italy
.
2.
Davies
,
R. R.
,
Goldstein
,
L. J.
,
Coady
,
M. A.
,
Tittle
,
S. L.
,
Rizzo
,
J. A.
,
Kopf
,
G. S.
, and
Elefteriades
,
J. A.
,
2002
, “
Yearly Rupture or Dissection Rates for Thoracic Aortic Aneurysms: Simple Prediction Based on Size
,”
Ann. Thorac. Surg.
,
73
(
1
), pp.
17
18
.10.1016/S0003-4975(01)03236-2
3.
Elefteriades
,
J. A.
, and
Farkas
,
E. A.
,
2010
, “
Thoracic Aortic Aneurysm Clinically Pertinent Controversies and Uncertainties
,”
J. Am. Coll. Cardiol.
,
55
(
9
), pp.
841
857
.10.1016/j.jacc.2009.08.084
4.
Sariola
,
H.
,
Viljanen
,
T.
, and
Luosto
,
R.
,
1986
, “
Histological Pattern and Changes in Extracellular Matrix in Aortic Dissections
,”
J. Clin. Pathol.
,
39
(
10
), pp.
1074
1081
.10.1136/jcp.39.10.1074
5.
E. Karangelis
,
D.
,
Kanakis
,
I.
,
P. Asimakopoulou
,
A.
,
Karousou
,
E.
,
Passi
,
A.
,
D. Theocharis
,
A.
,
Triposkiadi
,
F.
,
B. Tsilimingas
,
N.
, and
K. Karamanos
,
N.
,
2012
, “
Glycosaminoglycans as Key Molecules in Atherosclerosis: The Role of Versican and Hyaluronan
,”
Curr. Med. Chem.
,
17
(
33
), pp.
4018
4026
.10.2174/092986710793205354
6.
Camejo
,
G.
,
1982
, “
The Interaction of Lipids and Lipoproteins With the Intercellular Matrix of Arterial Tissue: Its Possible Role in Atherogenesis
,”
Adv. Lipid Res.
,
19
, pp.
1
53
.10.1016/B978-0-12-024919-0.50007-2
7.
Mattson
,
J. M.
,
Wang
,
Y.
, and
Zhang
,
Y.
,
2019
, “
Contributions of Glycosaminoglycans to Collagen Fiber Recruitment in Constitutive Modeling of Arterial Mechanics
,”
J. Biomech.
,
82
, pp.
211
219
.10.1016/j.jbiomech.2018.10.031
8.
Jufri
,
N. F.
,
Mohamedali
,
A.
,
Avolio
,
A.
, and
Baker
,
M. S.
,
2015
, “
Mechanical Stretch: Physiological and Pathological Implications for Human Vascular Endothelial Cells
,”
Vasc. Cell
,
7
(
1
), p.
8
.10.1186/s13221-015-0033-z
9.
Sotoudeh
,
M.
,
Li
,
Y. S.
,
Yajima
,
N.
,
Chang
,
C. C.
,
Tsou
,
T. C.
,
Wang
,
Y.
,
Usami
,
S.
,
Ratcliffe
,
A.
,
Chien
,
S.
, and
Shyy
,
J. Y.
,
2002
, “
Induction of Apoptosis in Vascular Smooth Muscle Cells by Mechanical Stretch
,”
Am. J. Physiol. Hear Circ. Physiol.
,
282
(
5
), pp.
H1709
H1716
.10.1152/ajpheart.00744.2001
10.
Schriefl
,
A. J.
,
Zeindlinger
,
G.
,
Pierce
,
D. M.
,
Regitnig
,
P.
, and
Holzapfel
,
G. A.
,
2012
, “
Determination of the Layer-Specific Distributed Collagen Fibre Orientations in Human Thoracic and Abdominal Aortas and Common Iliac Arteries
,”
J. R. Soc. Interface
,
9
(
71
), pp.
1275
1286
.10.1098/rsif.2011.0727
11.
O'Rourke
,
M.
,
1995
, “
Mechanical Principles in Arterial Disease
,”
Hypertension
,
26
(
1
), pp.
2
9
.10.1161/01.HYP.26.1.2
12.
Humphrey
,
J. D.
,
2013
, “
Possible Mechanical Roles of Glycosaminoglycans in Thoracic Aortic Dissection and Associations With Dysregulated Transforming Growth Factor-β
,”
J. Vasc. Res.
,
50
(
1
), pp.
1
10
.10.1159/000342436
13.
Roccabianca
,
S.
,
Ateshian
,
G. A.
, and
Humphrey
,
J. D.
,
2014
, “
Biomechanical Roles of Medial Pooling of Glycosaminoglycans in Thoracic Aortic Dissection
,”
Biomech. Model Mechanobiol.
,
13
(
1
), pp.
13
25
.10.1007/s10237-013-0482-3
14.
Mattson
,
J. M.
,
Turcotte
,
R.
, and
Zhang
,
Y.
,
2017
, “
Glycosaminoglycans Contribute to Extracellular Matrix Fiber Recruitment and Arterial Wall Mechanics
,”
Biomech. Model. Mechanobiol.
,
16
(
1
), pp.
213
225
.10.1007/s10237-016-0811-4
15.
Teng
,
Z.
,
Zhang
,
Y.
,
Huang
,
Y.
,
Feng
,
J.
,
Yuan
,
J.
,
Lu
,
Q.
,
Sutcliffe
,
M. P. F.
,
Brown
,
A. J.
,
Jing
,
Z.
, and
Gillard
,
J. H.
,
2014
, “
Material Properties of Components in Human Carotid Atherosclerotic Plaques: A Uniaxial Extension Study
,”
Acta Biomater.
,
10
(
12
), pp.
5055
5063
.10.1016/j.actbio.2014.09.001
16.
Ottani
,
V.
,
Raspanti
,
M.
, and
Ruggeri
,
A.
,
2001
, “
Collagen Structure and Functional Implications
,”
Micron
,
32
(
3
), pp.
251
260
.10.1016/S0968-4328(00)00042-1
17.
Wang
,
S.
,
Tokgoz
,
A.
,
Huang
,
Y.
,
Zhang
,
Y.
,
Feng
,
J.
,
Sastry
,
P.
,
Sun
,
C.
,
Figg
,
N.
,
Lu
,
Q.
,
Sutcliffe
,
M. P. F.
,
Teng
,
Z.
, and
Gillard
,
J. H.
,
2019
, “
Bayesian Inference-Based Estimation of Normal Aortic, Aneurysmal and Atherosclerotic Tissue Mechanical Properties: From Material Testing, Modeling and Histology
,”
IEEE Trans. Biomed. Eng.
,
66
(
8
), pp.
2269
2278
.10.1109/TBME.2018.2886681
18.
Sugita
,
S.
, and
Matsumoto
,
T.
,
2017
, “
Multiphoton Microscopy Observations of 3D Elastin and Collagen Fiber Microstructure Changes During Pressurization in Aortic Media
,”
Biomech. Model. Mechanobiol.
,
16
(
3
), pp.
763
–7
73
.10.1007/s10237-016-0851-9
19.
Alberts, B., Lewis
,
J.
,
Raff
,
M.
,
Roberts
,
K.
, and
Walter
,
P. J. A.
,
2002
,
Molecular Biology of the Cell
,
USA Garland Science
,
New York
.
20.
Evanko
,
S. P.
,
Johnson
,
P. Y.
,
Braun
,
K. R.
,
Underhill
,
C. B.
,
Dudhia
,
J.
, and
Wight
,
T. N.
,
2001
, “
Platelet-Derived Growth Factor Stimulates the Formation of Versican-Hyaluronan Aggregates and Pericellular Matrix Expansion in Arterial Smooth Muscle Cells
,”
Arch. Biochem. Biophys.
,
394
(
1
), pp.
29
38
.10.1006/abbi.2001.2507
21.
Hosoda
,
Y.
, and
Iri
,
H.
,
1966
, “
Angiolathyrism. 1. A Histological and Histochemical Study on Successive Changes of the Lathyritic Rat Aorta
,”
Acta Pathol. Jpn.
,
16
(
3
), pp.
239
252
.10.1111/j.1440-1827.1966.tb01957.x
22.
Guo
,
D. C.
,
Regalado
,
E. S.
,
Minn
,
C.
,
Tran-Fadulu
,
V.
,
Coney
,
J.
,
Cao
,
J.
,
Wang
,
M.
,
Yu
,
R. K.
,
Estrera
,
A. L.
,
Safi
,
H. J.
,
Shete
,
S. S.
, and
Milewicz
,
D. M.
,
2011
, “
Familial Thoracic Aortic Aneurysms and Dissections: Identification of a Novel Locus for Stable Aneurysms With a Low Risk for Progression to Aortic Dissection
,”
Circ. Cardiovasc. Genet.
,
4
(
1
), pp.
36
42
.10.1161/CIRCGENETICS.110.958066
23.
Farb
,
A.
,
Burke
,
A. P.
,
Tang
,
A. L.
,
Liang
,
T. Y.
,
Mannan
,
P.
,
Smialek
,
J.
, and
Virmani
,
R.
,
1996
, “
Coronary Plaque Erosion Without Rupture Into a Lipid Core. A Frequent Cause of Coronary Thrombosis in Sudden Coronary Death
,”
Circulation
,
93
(
7
), pp.
1354
1363
.10.1161/01.CIR.93.7.1354
24.
Arbustini
,
E.
,
Dal Bello
,
B.
,
Morbini
,
P.
,
Burke
,
A. P.
,
Bocciarelli
,
M.
,
Specchia
,
G.
, and
Virmani
,
R.
,
1999
, “
Plaque Erosion is a Major Substrate for Coronary Thrombosis in Acute Myocardial Infarction
,”
Heart
,
82
(
3
), pp.
269
272
.10.1136/hrt.82.3.269
25.
Lee
,
R. T.
,
Yamamoto
,
C.
,
Feng
,
Y.
,
Potter-Perigo
,
S.
,
Briggs
,
W. H.
,
Landschulz
,
K. T.
,
Turi
,
T. G.
,
Thompson
,
J. F.
,
Libby
,
P.
, and
Wight
,
T. N.
,
2001
, “
Mechanical Strain Induces Specific Changes in the Synthesis and Organization of Proteoglycans by Vascular Smooth Muscle Cells
,”
J. Biol. Chem.
,
276
(
17
), pp.
13847
13851
.10.1074/jbc.M010556200
26.
Kolodgie
,
F. D.
,
Burke
,
A. P.
,
Farb
,
A.
,
Weber
,
D. K.
,
Kutys
,
R.
,
Wight
,
T. N.
, and
Virmani
,
R.
,
2002
, “
Differential Accumulation of Proteoglycans and Hyaluronan in Culprit Lesions: Insights Into Plaque Erosion
,”
Arter. Thromb. Vasc. Biol.
,
22
(
10
), pp.
1642
1648
.10.1161/01.ATV.0000034021.92658.4C
27.
Partida
,
R. A.
,
Libby
,
P.
,
Crea
,
F.
, and
Jang
,
I. K.
,
2018
, “
Plaque Erosion: A New In Vivo Diagnosis and a Potential Major Shift in the Management of Patients With Acute Coronary Syndromes
,”
Eur. Hear J.
,
39
(
22
), pp.
2070
2076
.10.1093/eurheartj/ehx786
28.
Teng
,
Z.
,
Brown
,
A. J.
,
Calvert
,
P. A.
,
Parker
,
R. A.
,
Obaid
,
D. R.
,
Huang
,
Y.
,
Hoole
,
S. P.
,
West
,
N. E. J.
,
Gillard
,
J. H.
, and
Bennett
,
M. R.
,
2014
, “
Coronary Plaque Structural Stress is Associated With Plaque Composition and Subtype and Higher in Acute Coronary Syndrome: The BEACON i (Biomechanical Evaluation of Atheromatous Coronary Arteries) Study
,”
Circ. Cardiovasc. Imaging
,
7
(
3
), pp.
461
470
.10.1161/CIRCIMAGING.113.001526
29.
Azeloglu
,
E. U.
,
Albro
,
M. B.
,
Thimmappa
,
V. A.
,
Ateshian
,
G. A.
, and
Costa
,
K. D.
,
2008
, “
Heterogeneous Transmural Proteoglycan Distribution Provides a Mechanism for Regulating Residual Stresses in the Aorta
,”
Am. J. Physiol. Hear Circ. Physiol.
,
294
(
3
), pp.
H1197
H1205
.10.1152/ajpheart.01027.2007
30.
Wight
,
T. N.
,
2018
, “
A Role for Proteoglycans in Vascular Disease
,”
Matrix Biol.
,
71–72
, pp.
396
420
.10.1016/j.matbio.2018.02.019
31.
Chuong
,
C. J.
, and
Fung
,
Y. C.
,
1986
, “
On Residual Stresses in Arteries
,”
ASME J. Biomech. Eng.
,
108
(
2
), pp.
189
192
.10.1115/1.3138600
32.
Teng
,
Z.
,
Feng
,
J.
,
Zhang
,
Y.
,
Sutcliffe
,
M. P. F.
,
Huang
,
Y.
,
Brown
,
A. J.
,
Jing
,
Z.
,
Lu
,
Q.
, and
Gillard
,
J. H.
,
2015
, “
A Uni-Extension Study on the Ultimate Material Strength and Extreme Extensibility of Atherosclerotic Tissue in Human Carotid Plaques
,”
J. Biomech.
,
48
(
14
), pp.
3859
3867
.10.1016/j.jbiomech.2015.09.037
33.
Gray
,
M. L.
,
Burstein
,
D.
,
Kim
,
Y. J.
, and
Maroudas
,
A.
,
2008
, “
Elizabeth Winston Lanier Award Winner. Magnetic Resonance Imaging of Cartilage Glycosaminoglycan: Basic Principles, Imaging Technique, and Clinical Applications
,”
J. Orthop. Res.
,
26
(
3
), pp.
281
291
.10.1002/jor.20482

Supplementary data