Collagen Fibril Orientation in Tissue Specimens From Atherosclerotic Plaque Explored Using Small Angle X-Ray Scattering

Abstract

Atherosclerotic plaques can gradually develop in certain arteries. Disruption of fibrous tissue in plaques can result in plaque rupture and thromboembolism, leading to heart attacks and strokes. Collagen fibrils are important tissue building blocks and tissue strength depends on how fibrils are oriented. Fibril orientation in plaque tissue may potentially influence vulnerability to disruption. While X-ray scattering has previously been used to characterize fibril orientations in soft tissues and bones, it has never been used for characterization of human atherosclerotic plaque tissue. This study served to explore fibril orientation in specimens from human plaques using small angle X-ray scattering (SAXS). Plaque tissue was extracted from human femoral and carotid arteries, and each tissue specimen contained a region of calcified material. Three-dimensional (3D) collagen fibril orientation was determined along scan lines that started away from and then extended toward a given calcification. Fibrils were found to be oriented mainly in the circumferential direction of the plaque tissue at the majority of locations away from calcifications. However, in a number of cases, the dominant fibril direction differed near a calcification, changing from circumferential to longitudinal or thickness (radial) directions. Further study is needed to elucidate how these fibril orientations may influence plaque tissue stress–strain behavior and vulnerability to rupture.

1 Introduction

Every year an estimated 17 million people worldwide die of cardiovascular disease, making it responsible for one-third of all deaths [1]. Atherosclerosis is the most common form of that disease. It is a complex biological process in which plaque gradually develops at susceptible locations within certain arteries (e.g., carotid, coronary, and femoral). Plaque development starts with a disruption of the intima (inner layer of an artery) by low-density lipoprotein particles, leading to an accumulation of macrophage foam cells and lipids [2,3]. Further progression results in the formation of fibrous tissue between the bloodstream and the foam cells and lipids. Biomineralization (calcified regions) of a plaque may also occur [4]. Rupture of plaque tissue can lead to blood clots that frequently produce cardiovascular complications such as heart attacks, strokes, and acute limb ischemia [5,6]. The strength of the tissue is likely to contribute to its vulnerability to rupture.

Collagen fibrils are known to be key structural building blocks in a variety of tissues including those in arteries [7–9] and are largely responsible for providing tissue strength. Strength depends upon the orientation of the fibrils [10], with the greatest tissue strength in the longitudinal direction of fibrils. Collagen fibrils are prevalent in plaque tissue and thought to play a significant role in resistance of the tissue to rupture [11–13]. Information on collagen fibril orientation should be useful in gaining a better understanding of the vulnerability to rupture and further development of computational models for assessing that vulnerability [13–17]. Not only can fibril orientation influence tissue strength but the anisotropic material behavior associated with fibril orientation can alter computed stresses in plaque tissue relative to assumed isotropic material behavior, in some cases increasing the stresses substantially [13].

Calcified regions may be present at certain locations in the carotid and other arteries. The influence of such calcifications on resistance to plaque rupture is unclear. Some medical studies (e.g., Refs. [18–21]) suggest that calcified plaques are less likely to be symptomatic and may be more stable. Other studies (e.g., Refs. [22] and [23]) report opposite trends. Calcifications in plaque tissue can cause stress concentrations that amplify stresses near the calcifications [24–27] and thus increase the likelihood of rupture. But little appears to be known about fibril orientation in regions near calcifications. In addition to exploring the use of X-ray scattering for studying plaque tissue, we thought it would be of interest to explore collagen fibril orientation near calcifications, since fibril orientation could potentially influence local stresses and tissue strength.

Small angle X-ray scattering (SAXS) can provide information on the microstructural characteristics of materials with a repeating structure. Collagen molecules consist of three strands of polypeptide twisted into a helix approximately 3000 Å long [28]. The molecules self-assemble into collagen fibrils with a repeating longitudinal feature D caused by staggering of collagen molecules within a fibril as depicted in Fig. 1(a). The length of D is known as the “D-period” and is approximately 680 Å. When an incident X-ray beam passes through tissue containing fibrils, X-rays scatter and undergo constructive and destructive interference, forming cones of transmitted intensity, as in Fig. 1(b). The cones are created by scattering associated with the D-period of the fibrils, according to the Bragg relation [29]

$n\lambda =2D\hspace{0.17em}\hspace{0.17em}\text{sin}\hspace{0.17em}\hspace{0.17em}\theta$ (1)

Fig. 1.

(a) Schematic of a collagen fibril, (b) diagram of X-ray scattering setup and scattering pattern observed for collagen fibrils, and (c) intensity peaks from collagen fibrils

where λ = X-ray wavelength, and n is an integral order. The cones intersect an area detector, which monitors the number of X-ray photons reaching a given location, creating rings of varying intensity. Scattering from collagen fibrils can be plotted as intensity of peaks versus the magnitude of the scattering vector Q = (4π sin θ)/λ, where that vector is the difference between the X-ray vectors entering and exiting a specimen. The parameter Q varies with radial position on a detector. The peaks in Fig. 1(c) correspond to different orders of n. Using SAXS, it is possible to acquire qualitative and quantitative information about the orientation of collagen fibrils (e.g., Refs. [30] and [31]). When X-rays are scattered from collagen fibrils with a preferred alignment, arcs of greater intensity will appear on a detector, as in Fig. 1(b). Orientation is obtained from the variation in intensity around a ring as a function of angle χ in that figure. SAXS has been used to characterize collagen fibril orientation in a variety of soft tissue specimens (e.g., Refs. [32–45]) with two-dimensional (2D) fibril orientation evaluated from an intensity pattern resulting from the projection of fibrils on a detector. To characterize the three-dimensional (3D) orientation of mineralized collagen fibrils in bone, a number of studies [31,46–53] have analyzed the patterns resulting from different orientations of a specimen relative to an incident X-ray beam.

This study explores for the first time the use of SAXS to provide information about collagen fibril orientation in samples of atherosclerotic plaque tissue. The results reported here are from experiments and analyses performed between 2015 and 2019.

2 Materials and Methods

2.1 Anatomy and Pathophysiology.

Tissue specimens were obtained from human femoral and carotid arteries, with consent of patients and approval by the Stanford University Institutional Review Board. The femoral artery is located in the thigh and extends parallel to the femur [54]. It then splits into deep and superficial femoral arteries. Accumulation of atherosclerotic plaque in femoral arteries is a common disease in humans [55] and can greatly impair daily activities such as walking by reducing blood flow to the lower extremities [56,57].

Carotid arteries branch off the aortic arch and travel to the brain, sitting underneath the sternocleidomastoid muscles in the neck, and supplying oxygenated blood to the neck, face, and brain. Each carotid artery bifurcates into two major branches, the external and internal carotid arteries [54]. Rupture of carotid plaque tissue within the internal carotid artery can result in strokes [6].

2.2 Specimen Details.

Specimens designated here as A to D and E-I, E-II, and E-III were extracted from plaque tissue obtained from patients who were undergoing vascular surgery and consented to using their tissue for further study. Specimens A and B came from plaque within femoral arteries, while C, D, E-I, E-II, and E-III were from plaque within carotid arteries. The specimens for a given set of experiments depended on the shape and size of tissue that became available from surgeries just prior to the start of a set of experiments. The specimens of plaque tissue had visible, heavily calcified areas (“macroscopic” calcifications mm and larger) that also appeared in X-ray transmission maps of specimens described later.

2.3 Experimental Approach.

Small angle X-ray scattering experiments were performed using the setup shown schematically in Fig. 1(b), with angle of tilt denoted by ω. A photo of the setup is in Fig. 2. Beamline 1-5 of the Stanford Synchrotron Radiation Lightsource, with a photon flux of 5 × 10¹⁰ ph/s, was used for the experiments. A Rayonix 165 CCD image plate detector registered scattering patterns. The beamline was equipped to obtain intensity patterns at a series of locations along a given scan line and included a goniometer to tilt specimens relative to the incident beam. A typical specimen and its holder are shown in Fig. 2. Scan lines corresponded to the circumferential direction of plaque tissue, which also corresponded to the circumferential directions of arteries from which tissue was obtained. Given that all the plaque tissue specimens had significant regions of macroscopic calcification, we thought it would be useful to explore if and how collagen fibril orientation might change near calcifications, since orientation could potentially influence local tissue strength [10]. Scan lines were selected that started away from calcified regions and ended just within them.

Fig. 2.

Representative specimen enclosed within Kapton film in a holder with a circular window of 15 mm diameter and SAXS experimental setup showing the specimen holder attached to a goniometer used for tilting a specimen relative to an incident X-ray beam

To characterize the 3D orientation of the collagen fibrils at each location along a scan line, an intensity pattern was obtained with the beam normal to a specimen (0 deg tilt) and at the different tilt angles shown in Table 1, which also provides the beam size, X-ray energy for each specimen, and the duration of X-ray exposure per pattern obtained. Using 10 deg increments of tilt applied by a goniometer enabled reconstruction of fibril orientation. Calibration of the experimental setup was performed with silver behenate (AgB) powder, an effective calibrant for SAXS experiments [58].

Table 1.

Specimen and SAXS scanning information

Specimen	Arterial source of tissue	Average thickness (mm) in scanned region	Exposure time per pattern (s)	X-ray energy (keV)	Beam sizea (μm)	Tilt anglesb (deg)	No. of scan lines	No. of scattering patterns obtained
A	Femoral	0.4	120	12	300 × 300	0 deg to 50 deg	2	86
						10 deg increments
B	Femoral	0.5	80	12	300 × 300	0 deg to 50 deg	2	288
						10 deg increments
C	Carotid	0.5	120	12.7	500 × 400	−10 deg to 40 deg	1	42
						10 deg increments
D	Carotid	0.4	120	12.7	500 × 400	−10 deg to 40 deg	2	90
						10 deg increments
E-I	Carotid	0.5	110	15.5	250 × 250	0 deg to 50 deg	1	78
						10 deg increments
E-II	Carotid	0.6	110	15.5	250 × 250	0 deg to 50 deg	1	120
						10 deg increments
E-III	Carotid	0.6	90	15.5	250 × 250	0 deg to 25 deg	2	168
						5 deg increments

^aBeam sizes measured at the closest set of slits to a specimen.

^bThe variation in tilt angles resulted from the need to conduct experiments at different beam-times over several years and changes in beamline equipment configurations from one beam-time to another beyond our control. The number of tilt angles was kept constant throughout the experiments.

Possible effects of X-ray radiation on collagen fibril orientation in soft tissue do not appear to have been reported in the literature. Since radiation dosage could potentially affect our results, doses in this study were estimated using a relation given by Deymier-Black et al. [59] plus attenuation data for soft tissue [60]. The maximum dose at a given location on any of the specimens was computed to be 196 kGy. To check whether data collected here were being affected by radiation, an experiment was performed in which plaque tissue was subjected to a dose of nearly 2000 kGy. Although the magnitude of intensity values on a plot of intensity versus χ declined by approximately 30% following that dose, there was no noticeable change in location of χ peaks on the plot, indicating that if ionizing radiation was damaging a specimen, it was not doing so in a way that compromised information needed to deduce fibril orientation.

Finally, it should be noted that for required biosafety reasons, specimens used in this study were preserved in formalin, as has been done in other studies reporting collagen fibril orientation (e.g., Refs. [61–63]). While formalin is known to induce cross-linking of collagen molecules [64], studies of connective tissue [65] and skin [66] have found no significant effect of formalin on dominant fibril directions.

2.4 Data Analysis.

The fifth-order collagen fibril scattering ring observed on an area detector was used to characterize the orientation of collagen fibrils. That ring was selected because it was the one with the greatest intensity in the data collected, as shown in Fig. 1(c). Data collected by the area detector were integrated around a given ring in the azimuthal (χ) direction for a desired Q range to obtain the variation of intensity I(χ). Q range refers here to the difference between inner and outer radii of an annulus that enclosed a ring.

Next, the diffuse background intensity recorded by the area detector associated with organic material surrounding the collagen fibrils and the Kapton film used to enclose a specimen was taken into account. The azimuthal variation in intensity around an annulus like that shown by the dashed lines in Fig. 3(a) represents a background signal. That profile was subtracted from the fibril scattering data profile to obtain a corrected I(χ) profile, as done in Ref. [67]. The intensity spikes shown in Fig. 3(a) are from scattering at slits [68] used in beamlines to shape an incident X-ray beam. The spikes were avoided by use of the fifth-order collagen fibril ring for analysis. The variation of intensity I(χ) with angle χ for each tilt angle was then fitted with a sinusoidal function, as done in Ref. [43]. A sample of a fit is shown in Fig. 3(c). The intensity data in this study did not exhibit the relatively strong scattering peaks often seen, for example, in studies of bone crystals [46–51]. The intensity data at different tilts were normalized using a method [48] that accounted for differences in volumes of material undergoing scattering with tilting.

Fig. 3.

(a) Image from detector showing intensity spikes from slit scattering and an annulus (solid red lines) surrounding fifth-order collagen fibril scattering data and an annulus (dashed lines) used for background correction, (b) orientation vector $\widehat{\mathbf{r}}$, and (c) example of a sinusoidal fit to intensity versus angle χ data at a given location for different tilt angles. Intensity is expressed as X-ray counts recorded at a detector.

The I(χ) data from different tilt angles provided the input to quantify the orientation of the fibrils at each location along a scan line. To characterize the orientation of the fibrils, a SAXS approach developed by Georgiadis et al. [51] was used, but with the mathematical relations modified to account for tilting rather than rotation. Referring to Fig. 3(b), the orientation is represented by a vector $\widehat{\mathbf{r}}$ in spherical coordinates with azimuthal and polar angles φ₀ and θ₀, respectively. For a tilt ω, the vector describing the orientation is

$\widehat{\mathbf{r}}=\left[\widehat{\mathbf{r}}\right]\{\hspace{0.17em}\text{sin}(\omega +{\varphi }_0\left)\text{sin}\right({\theta }_0)\widehat{\mathit{x}}+\text{cos}(\omega +{\varphi }_0\left)\text{sin}\right({\theta }_0)\widehat{\mathit{y}}+\text{cos}({\theta }_0\left)\widehat{\mathit{z}}\right\}$ (2)

The vector $\widehat{\mathbf{r}}$ has a 2D projection ${\widehat{\mathbf{r}}}_{y-z}$ on the y–z plane, and the orientation polar angle θ₀ has a 2D projection angle χ on the y–z plane, leading to Eq. (3). The y–z plane is important, as that is where the detector was located and scattering data were recorded

$$\text{cot}(\chi )=\frac{\text{cot}\left({\theta }_0\right)}{\hspace{0.17em}\text{cos}(\omega +{\varphi }_0)}$$ (3)

Using the intensity distribution I(χ) from collagen fibril scattering, the χ value at the maximum peak intensity for each tilt angle ω was computed. The resulting pairs of (ω, χ) were least squares fitted to Eq. (3) to find the azimuthal and polar angles φ₀ and θ₀, respectively, providing the 3D orientation for the collagen fibrils.

3 Results

Femoral specimens were studied first because they happened to be the first ones available from surgeries. Specimen A is shown in Fig. 4(a), along with a corresponding X-ray transmission map in Fig. 4(b), where blue and red denote high and low absorption, respectively, and other colors such as yellow signify intermediate values. The dashed lines represent scan lines. The dark blue region corresponds to a macroscopic calcification also visible in Fig. 4(a). Mean orientation directions at locations along the two scan lines are shown in Figs. 4(c) and 4(d) by small blue rods. Front and top views in Fig. 4(c) refer to the isometric box shown and are intended to aid visualization of orientation. (Figures 5 and 6 omit isometric boxes and use only top and front views as visual aids.) In addition to the views, Cartesian components in longitudinal, circumferential, and thickness (radial) directions were also determined. The ratios of longitudinal-to-circumferential components (L/C) and thickness-to-circumferential components (T/C) were computed for each location along a scan line. All locations along scan lines 1 and 2 of specimen A had (L/C) values < 0.5 (most < 0.3) and similarly for (T/C) values, indicating a preferred fibril orientation toward the circumferential direction, also seen by the views in Fig. 4.

Fig. 4.

(a) Specimen of plaque tissue from a femoral artery of patient A, (b) transmission map, (c) isometric view of fibril orientations with front and top views along scan line 1, and (d) along scan line 2. Long (y), circ (z), and thk (x) designate longitudinal, circumferential, and thickness (radial) directions of the plaque tissue. The black dots represent heavily calcified regions without useable data on fibril orientation.

Fig. 5.

(a) Collagen fibril directions for a specimen of plaque tissue from a femoral artery of patient B and from carotid arteries of (b) patient C and (c) patient D. The black dots signify heavily calcified regions. (The spacing between locations along a given scan line may vary from that along other scan lines.)

Fig. 6.

Plaque tissue from a carotid artery of patient E, with (a) an X-ray transmission map of specimen E-I extracted from the base of one branch plus orientation directions along its scan line, (b) transmission map of a specimen E-II extracted from just below the bifurcation plus orientation directions along its scan line, and (c) transmission map from specimen E-III extracted from the base of the other branch plus orientation vectors along scan lines 1 and 2. Long (y), circ (z), and thk (x) designate longitudinal, circumferential, and thickness directions of the plaque tissue. The black dots signify calcified regions.

For specimens B to D, X-ray transmission maps and fibril orientations views are given in Fig. 5. Specimen B had a centrally located, narrow region of heavy calcification, as seen in its transmission map (Fig. 5(a)). Starting at the left-hand side of scan line 1, fibrils were primarily in the circumferential direction, with (L/C) and (T/C) values < 0.8, except for one location where (T/C) was 1.5. See Fig. 5(a). Approaching the calcified region, either (L/C) or (T/C) was between 1.0 and 1.5 at four of six locations. To the right of the calcified region, (T/C) values at all locations except the very end of the scan line were between 1.1 and 2.9 with an average of 1.5, indicating fibril orientation mainly in the thickness direction, as also seen in Fig. 5(a). For scan line 2 of specimen B, fibrils were oriented primarily in the circumferential direction with (L/C) < 0.5 and (T/C) < 0.8. An exception was three of five locations near the right-hand side of the calcified region, where (T/C) was between 1.1 and 2.0, averaging 1.5.

The transmission map for specimen C (Fig. 5(b)) reveals a centrally located region of calcification. Half of the locations along the scan line had (L/C) values of approximately 0.8 but were much lower elsewhere. At the location nearest the calcified region, (T/C) was 1.2.

Overall, fibrils were somewhat more aligned with the circumferential than the other two directions, as also suggested in the views in Fig. 5(b).

Specimen D had a large region of relatively light calcification, but also a small, heavily calcified region, as shown by the corresponding light and dark blue regions in its transmission map (Fig. 5(c)). Line 1 locations had (L/C) values ranging from 1.0 to 9.5, averaging 2.5. In this case, the longitudinal direction dominated, with substantial components in the thickness at several locations as well, as also seen in the views in Fig. 5(c). For line 2, three of five locations had (T/C) values between 1.0 and 2.0, with an average of 1.4, with considerably smaller (L/C) values. Fibrils tended toward the circumferential direction at some locations and the thickness direction at others.

As shown in Fig. 6, three specimens were obtained from a volume of carotid plaque tissue that had both branches intact. Specimen E-I was extracted from a region near the base of the left branch. As seen in the transmission map in Fig. 6(a), a scan line started in an area without significant calcification and extended to one with noticeable calcification. Locations along most of the line had (L/C) and (T/C) values < 0.7. Fibrils were thus oriented predominantly in the circumferential direction, but with increasing (L/C) and (T/C) values approaching the calcified area. (L/C) reached 2.6 and (T/C) > 11 near the calcification. The location closest to that area was an exception to that trend, with both (L/C) and (T/C) < 0.3. Specimen E-II was taken from an area just below the bifurcation. Its scan line extended across a region between two areas of calcification and ended in a region away from those areas, as indicated by the transmission map in Fig. 6(b). Locations between the two calcifications had (L/C) values ranging from 1.7 to 9.5, averaging 3.3, while (T/C) was between 1.4 and 7.1, averaging 2.9. Hence, fibrils were oriented strongly in the longitudinal direction with significant components in the thickness direction. Similar values of (L/C) and (T/C) existed for the next seven locations to the right of the calcifications, then changing toward the circumferential direction farther away. The change in orientations is also seen in Fig. 6(b).

Specimen E-III was removed from a region near the base of the right branch of plaque tissue. Scans were made along two lines that started within an areas calcification, as shown in the transmission map of Fig. 6(c) and ended away from that area. For the first four locations along scan line 1 to the right of the calcified area, (L/C) varied between 1.4 and 7.1, averaging 4.5, and (T/C) between 1.7 and 11.4, averaging 5.5. Fibrils were oriented with strong components in the longitudinal and thickness directions near the calcified region, becoming predominantly circumferential moving away from the region. Scan line 2 was located just 0.25 mm beneath line 1 to see if nearby fibril orientations would be similar or differ. Orientations were very similar.

Several observations can be made from the experimental results. First, fibril orientation differed substantially from specimen-to-specimen and even within a specimen, perhaps not surprisingly given the heterogeneous nature of plaque tissue. The component of fibril direction in the circumferential direction did tend to be larger than in longitudinal or thickness directions for the majority of scanned locations away from calcifications. As an example, at more than an mm away from calcifications, the circumferential direction was dominant for 72% of locations. The preferred fibril direction was also circumferential at locations near calcifications in certain cases, such as for scan lines 1 and 2 of specimen A (Fig. 4) and the left-hand side of line 2 of specimen B (Fig. 5(a)). On the other hand, fibrils were oriented strongly in the longitudinal or thickness directions near calcifications for cases such as scan line 1 of specimen D (Fig. 5(c)) and the scan lines of specimens E-II and E-III (Fig. 6). For other scan lines, fibril orientation near calcifications was mixed, in some instances fluctuating between thickness and circumferential directions, as seen, for example, in the right-hand side of scan 2 of specimen B (Fig. 5(a)).

4 Discussion

The primary goal of this study was to explore the ability of SAXS to find collagen fibril orientation in specimens of human plaque tissue, a heterogeneous material with a spatially varying mixture of soft tissue and mineralized components. The results obtained here demonstrate that SAXS can be successfully applied to reveal fibril orientation in that type of tissue. As this is the first known study to explore the application of SAXS to plaque tissue, there are limitations and room for improvement. The flux of the X-ray beam available for this study was 5 × 10¹⁰ ph/s, necessitating an exposure time of 80–120 s for each scattering pattern. Exposure time could potentially be reduced by approximately 2 orders of magnitude if a beamline with much higher flux (e.g., 5 × 10¹² ph/s) could be used, provided that it could be equipped with a suitable goniometer for titling and rotating a specimen. Such a capability would enable data to be obtained over larger areas of a specimen within a given time period. The potential also exists for finding collagen fibril orientation with finer spatial resolution than possible in this study. For example, experiments at the Swiss Light Source synchrotron [52,53] have been able to characterize collagen fibril orientation throughout the volume of a small trabecular bone from a vertebra and a small specimen extracted from a tooth. The volume of the bone specimen (approximately 1 × 1 × 2.5 mm) was represented by a 3D grid of 25 × 25 × 25 μm voxels, the 3D analog of pixels. X-ray scattering patterns were obtained by scanning the specimen using numerous combinations of tilt and rotation at each scan location, resulting in acquisition of a total of 1.6 × 10⁶ patterns. The data acquisition required approximately 35 h, which could be reduced by higher flux, faster X-ray detectors, and optimized equipment for motorized specimen positioning, rotation, and tilting. (Data for the tooth specimen were obtained by a similar scanning procedure.) This type of 3D scanning approach could potentially be applied to explore fibril orientation in regions of interest in plaque tissue like that in Fig. 6, separated into two halves for scanning. Regions near calcifications may be of particular interest. Details of the 3D geometry of calcifications could be obtained by microcomputed X-ray tomography.

Other imaging methods could potentially be used to investigate collagen fibril orientation near larger calcified regions like those in this study or smaller ones such as microcalcifications [24–27] in coronary or other arteries. For instance, optical coherence tomography (OCT) has recently been used to find fibril orientation in atria and ligament tissue specimens [69]. In earlier studies, polarization sensitive OCT has provided birefringence and optical axis data (indicators of material anisotropy from fibril organization) for specimens of coronary and aorta arteries containing plaque [70–72]. More recently, polarization sensitive OCT has been used to visualize the fibril orientation in bovine carotid artery walls without plaque [73]. Diffusion tensor MRI may also be used find orientation in plaque tissue [74] along with second harmonic generation microscopy [61]. Small angle light scattering has been used to find orientation in specimens of bovine pericardium and porcine aortic valve tissue [75] as well as arterial tissue without plaque [76].

A second purpose of this study was to explore fibril orientation along scan lines that started away from and ended just within a macroscopic calcification. The orientation of collagen fibrils near calcifications may be another factor to consider in assessing plaque vulnerability to rupture, owing to its potential influence on tissue strength [10] and stresses [13]. For instance, consider the longitudinal orientation of fibrils in Figs. 6(b) and 6(c). Would that orientation or similar ones make the region more prone to rupture if tensile stresses act in a circumferential direction, normal to the strengthening effect of fibrils in the longitudinal direction? We hypothesize that fibril orientation near calcifications may be a significant factor in plaque vulnerability.

To assess the influence of the observed change in fibril orientation near certain calcifications on the vulnerability to rupture, strength tests of the tissue near the calcifications are needed. As suggested by Barrett et al. [27], current micromechanical testing capabilities may make such experimentation a possibility. For instance, tension testing techniques utilizing soft tissue specimens on the order of mm in size have been developed [77,78]. Stress–strain testing of microscale specimens formed in the transition region of tendon-to-bone tissue has also been demonstrated [79]. It should be possible to design a test setup that would enable SAXS imaging of specimens of plaque tissue with a calcification(s), using a combination of different tilts and rotations of a specimen, to map 3D fibril orientation in a region near the calcification. The X-ray radiation doses in such experiments would need to be carefully limited to minimize effects on tissue mechanical properties, since large X-ray doses can significantly degrade strength [80]. Other methods for finding fibril orientation such as OCT could be used instead. Microcomputed tomography could also be used to provide a 3D description of the calcification. Then, miniature tensile specimens could be created from tissue near the calcification and scanned to see if orientation in the tensile specimens had changed relative to that in the tissue from which they were created. The miniature specimens would then be tension tested to determine strength.

Predicting conditions likely to make plaque tissue vulnerable to rupture is a challenge because of the many factors that can influence vulnerability and the variability in tissue characteristics from patient-to-patient. Collagen fibril orientation in regions near calcifications may be another factor to consider in assessing vulnerability. Knowledge of the 3D orientation of collagen fibrils in plaque tissue specimens with different locations, shapes, and sizes of calcifications should improve the fundamental understanding of the biomechanical behavior of that tissue and enable future computational models to better assess if a plaque is likely to rupture. In that regard, a finite element modeling approach that incorporates the anisotropic stress–strain behavior associated with fibril orientations in different regions of coronary plaque tissue has been developed [13] and should be applicable in modeling of carotid and other arteries if sufficient information on fibril orientation is available. Obtaining additional information on fibril orientation in plaque tissue seems a worthwhile topic for researchers to pursue.

Conclusions

Small angle X-ray scattering combined with tilting was found to be useful in characterizing the 3D collagen fibril orientation in specimens of atherosclerotic plaque tissue from human carotid and femoral arteries.

At the majority of locations away from macroscopic calcifications in the plaque specimens studied, fibrils were observed to be oriented primarily toward the circumferential direction of tissue, which corresponded to the circumferential direction of arteries from which specimens were removed.

In some cases, fibril orientation in plaque tissue did not appear to change appreciably in regions close to macroscopic calcifications, while in other cases, a pronounced change from being mainly circumferential to predominantly in the longitudinal or thickness directions was observed. The differences in collagen fibril orientation near calcifications may warrant further studies because of their potential influence on plaque tissue vulnerability to rupture.

Funding Data

• We are grateful for a grant from the Stanford Cardiovascular Institute and a Bio-X fellowship (Funder ID: 10.13039/100011098) that supported the first author.