If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Aortic dissection occurs when a weakened portion of the intima tears, and a separation of layers propagates along the aortic wall to form a false lumen filled with active blood flow or intramural thrombus. The unpredictable nature of aortic dissection formation and need for immediate intervention leaves limited serial human image data to study the formation and morphological changes that follow dissection. We used volumetric ultrasound examination, histology, and scanning electron microscopy (SEM) to examine intramural thrombi at well-defined timepoints after dissection occurs in apolipoprotein E-deficient mice infused with angiotensin II (n = 71). Stratification of red blood cell (RBC) morphologies (biconcave, intermediate biconcave, intermediate polyhedrocyte, and polyhedrocyte) in the thrombi with scanning electron microscopy (n = 5) was used to determine degree of thrombus deposition/contraction. Very few biconcave RBCs (1.2 ± 0.6%) were in the thrombi, and greater amounts of intermediate biconcave RBCs (25.8 ± 6.7%) were located in the descending thoracic portion of the dissection while more polyhedrocytes (14.6 ± 5.1%) and fibrin (42.3 ± 4.5%; P < .05) were found in the distal suprarenal aorta. Thrombus deposition likely plays some role in patient outcomes, and this multimodality technique can help investigate thrombus deposition and characteristics in experimental animal models and human tissue samples.
Aortic dissection (AD) occurs when a tear in the intima leads to separation of layers within the vessel, creating a blood-filled false lumen with complex flow patterns. Incidence of AD is roughly 4.4 cases per 100,000 people in the United States
Population-based assessment of the incidence of aortic dissection, intramural hematoma, and penetrating ulcer, and its associated mortality from 1995 to 2015.
Although AD incidence rate is low compared with other cardiovascular diseases, it is associated with high mortality rates that approach 50% within 48 hours of onset.
Owing to sudden onset and high mortality rates, there is limited human ultrasound image data from immediately before and serially after dissections take place. Animal models provide the opportunity to capture baseline and serial image data during disease development and progression.
were the first to describe the murine model of AD where chronic infusion of angiotensin II (AngII) via an osmotic pump in hyperlipidemic mice results in approximately 60% of male animals developing Ads. The apoE-deficient (or apoE−/−) AngII model is well established, widely used, and is associated with atherosclerosis. AngII-induced dissections in apoE−/− mice develop in the descending thoracic and suprarenal abdominal aorta, resembling Stanford type B or DeBakey type III dissections in humans that originate in the descending thoracic aorta and typically extend past the renal arteries. However, these mouse AD’s have a relatively low rupture risk and show separation between the medial and adventitial layers instead of within the media. Despite differences between human and murine Ads, animal models are useful to study the hemodynamic and morphological changes that lead to dissection and can evaluate therapeutic targets that prevent or stabilize dissections.
Previous animal studies have quantified the complex hemodynamic and morphological changes that take place after a dissection using in vivo and ex vivo imaging techniques.
However, the unpredictability of when dissections form, and the effort required for serial longitudinal imaging studies of experimental AD animal models means that many studies focus only on ex vivo analysis of tissue.
The lack of in vivo imaging is a limitation because researchers are unable to determine how old a dissected aneurysm is at explant and how local hemodynamics affect thrombus deposition in both acute and chronic phases. In this study, we use daily aortic ultrasound scans of mice to identify when dissections form and collect tissue samples with a better temporal understanding of AD formation. We assessed both acute and chronic changes in the aorta after dissection through the use of histology and scanning electron microscopy (SEM) to quantify thrombus composition and structure.
Although past studies have used SEM imaging of human and mouse aorta samples to qualitatively assess changes in the aortic wall structure after dissection,
to our knowledge, this report is the first the first report of quantitative analysis of thrombi in the false lumen of murine ADs using SEM. SEM has been used to quantify structural changes accompanying clot contraction in vitro
However, SEM has not been used to analyze intramural thrombus formation in murine ADs. Thrombus deposition likely plays a role in human patient outcomes, often determining if a dissection will stabilize or rupture.
We describe this multimodality imaging method as a way to investigate how local hemodynamics and dissection morphology affect thrombus deposition, structure, and composition (including thrombus contraction) at different anatomical locations during the acute and chronic phases. With these results, we are hopeful that this approach can be applied to additional aortopathies and human tissue samples, enabling the prediction of thrombus deposition and properties in future studies.
Diameter, volume, velocity, and flow in the true and false lumens of the aorta from serial ultrasound examinations (Fig 1, A and B) revealed AD formation less than 24 to 32 hours after dissection occurrence. Dissections developed 3-28 days after pump implant in 41 of the implanted mice 71 (57.7%). Animals older than 35.0 weeks developed dissections later after pump implant on average (11.1 ± 0.7 days) compared with young mice (6.4 ± 0.5 days; P < .05) (Fig 1, D-F). Despite this significant difference in dissection timing after implantation, both groups had similar increases in aortic diameter immediately after pump implantation (Fig 1, C). We did observe greater flow at baseline in the old group (Supplementary Fig 1), but this difference is likely due to larger diameters at baseline in the old group, because velocity was not significantly different between young and old mice. We used volumetric ultrasound scans to align histology and SEM sections in their anatomical position ex vivo (Fig 2). Aligning tissue sections with ultrasound improved segmentation of the true lumen, thrombosed false lumen, and patent false lumen in our ultrasound images, and allowed us to match local flow characteristics to each section (Fig 2, D-F).
Fig 1Daily ultrasound scans detect dissections less than 24 to 32 hours after they develop. Using this method, we found young mice develop dissections earlier after pump implant compared with older mice. These data can also provide serial diameter/volumetric data both before and after dissection. (A) Three-dimensional ultrasound (3DUS) scans provide volumetric data of dissecting aortic aneurysms in vivo. (B) Diameter measurements at three locations along the aorta using the 3DUS scans. (A, B) Yellow, true lumen; green, false lumen. (C) Diameter change from baseline to 3 days after pump implant showed no difference between young (9.7-21.7 weeks of age) and old (35.0-88.1 weeks of age) mice. (D-F) We observed a significant difference in the timing of dissections postimplant between young and old mice, with young mice developing dissections earlier after implant (∗∗∗∗P < .001). Scale bars, 1 mm. Ao, aorta; SMA, superior mesenteric artery; VC, vena cava.
Fig 2Daily ultrasound scans help detect when dissections occurred after pump implantation (8.0 ± 0.6 days on average) and segmentation revealed regions of thrombosed and patent false lumen. Histology and scanning electron microscopy (SEM) analysis showed high percentages of fibrin and red blood cells (RBCs) in the intramural thrombus that formed, a compressed true lumen, and collagen surrounding the adventitia. (A) Daily ultrasound scans collected hemodynamic and morphological data in vivo.(B) Short-axis schematic of abdominal aorta with surrounding organs and tissue. (C) Aorta was removed and sliced into 12 sections where every other section was saved for histology or SEM. (D-F) Ultrasound images, histology, and SEM cross-sections can be aligned anatomically. Histology and SEM sections help with the comparison of structural differences observed in ultrasound images, helping to identify composition and structure of the resulting intramural thrombus in the false lumen. Yellow, true lumen; cyan, patent false lumen; green, thrombosed false lumen. Scale bars, 500 μm. (A-C created in Biorender.com.)
Histology slides stained with hematoxylin and eosin (Fig 3, A) and Movat's pentachrome (Fig 3, B) provided compositional information about the areas that contain red blood cells (RBCs), proteoglycans, elastin, collagen, and fibrin within the AD. On average, we found significantly higher percentages of RBCs (45.3%) in the false lumen compared with all other tissue structures. There was also a large percentage of fibrin (30.8%) found in most sections (Fig 3, C). Elastin in the intima remained unfragmented around much of the circumference, and the majority of collagen was observed at the outer adventitia of the false lumen. Semiquantitative analysis by a veterinary pathologist determined that intramural hemorrhage was present in 100% of sections, with 58.6% containing intramural thrombus and relatively low amounts of tertiary lymphoid organs (24.1%) and atherosclerotic lesions (10%). Interestingly, the semiquantitative analysis revealed differences in composition based on the anatomical location of the sections along the aorta. Proximal suprarenal aorta (middle) sections had the highest degree of thrombus organization and presence of intramural thrombus in comparison with descending thoracic (proximal; P = .07, P = .14) and distal suprarenal aorta (distal; P = .75, P = .70) sections (Fig 3, E and F). Color segmentation of pentachrome images revealed that middle regions of the dissection were composed of slightly higher percentages of fibrin than proximal (P = .65) and distal sections (P = .98). Compressive force from platelets pulling on the fibrin network change RBC morphology,
so a greater percentage of fibrin could attribute to a higher degree of thrombus organization in the middle region (Fig 3, D). Despite these trends, there were no significant differences between cellular structures at different anatomical locations in our histological analyses.
Fig 3Histology showed a significant number of red blood cells (RBCs) (P < .001) and fibrin (P < .001) compared with all other structures, but no statistically significant differences between proximal, middle, and distal regions. (A, B) Histology sections stained with hematoxylin and eosin (top) and Movat’s Pentachrome (bottom). (C, D) RBCs and fibrin comprised a large portion of the intramural thrombus. Thrombus composition changes from the proximal, middle, and distal sections of the aorta. (Created in Biorender.com.) (E, F) A semiquantitative analysis identified the presence of significantly more intramural immune cells in middle and distal regions compared with proximal sections (P < .05) and slightly increased intramural thrombus and thrombus organization in middle sections. Scale bars, 50μm.
SEM imaging showed separate thrombi within the false lumen that formed at different times (Fig 4, C). Analysis of individual thrombus structures in the false lumen revealed a heterogenous mixture of RBCs and fibrin at different locations in the thrombus and at different regions along the aorta (Fig 4, A and B). On average, the intramural thrombus was composed of 61.2 ± 1.7% RBCs and 34.3 ± 1.8% fibrin (Fig 4, G). Differences in proximal and distal sections of the thrombus were observed with most intermediate biconcave RBCs being found in proximal sections (25.8 ± 6.7%) (Fig 4, D and H), whereas polyhedrocytes (14.6 ± 5.1%) were found mostly in distal portions (Fig 4, F and H). Higher percentages of fibrin were also found in distal portions of that thrombus (42.3 ± 4.3%) compared with proximal sections (17.1 ± 3.2%; P = .02). We observed a circumferential fibrin orientation in some sections, primarily in distal regions where complex, recirculating flow in the false lumen was present (Fig 4, E).
Fig 4We found that 98.3% of red blood cells (RBCs) in the intramural thrombus had deformed morphologies and regional analysis revealed 25.8% of proximal intramural thrombus was intermediate biconcave RBCs, whereas 14.6% of distal thrombus was polyhedral RBCs. Fibrin made up 29.7% of all intramural thrombus in the sections analyzed. (A) Low magnification SEM images are used to identify different thrombus structures (1-4) in the false lumen to be scanned at higher magnification. Scale bar, 500μm. (B) A grid was placed on high magnification (2000×) scanning electron microscopy (SEM) images and FIJI ImageJ cell counter was used to identify RBC types, fibrin, and other structures. Scale bar, 25 μm. (C) Black and white arrows indicate borders of different thrombi in a proximal section of thrombus. Scale bar, 500μm. (D) Intermediate biconcave RBCs were located primarily in proximal sections of thrombus. Black and white arrows point to intermediate biconcave RBCs. Scale bar, 10 μm. (E) Black and white arrows indicate circumferential fibrin and thrombi organization in a distal section. (F) Black and white arrows point to polyhedral RBCs found mostly in distal portions of the thrombus. (G, H) Each dissection has a unique composition of RBCs with varying morphology and fibrin. Composition as well as clot contraction is different in proximal and distal sections of the dissection, which can likely be attributed to local hemodynamics.
The use of high-resolution in vivo volumetric ultrasound imaging, in combination with histology and SEM, allows for the analysis and characterization of intramural thrombi at well-defined time points after dissection. Using histology and SEM ex vivo, we observed intramural thrombi that contained high percentages of deformed RBCs and fibrin. Although the RBC and fibrin percentages were heterogenous from mouse to mouse, we found consistent differences in composition at different anatomical positions along the aorta, providing further evidence for the complexity and heterogeneity of these murine ADs. Given these observations, we hypothesize that local blood flow patterns, thrombus age, and dissection morphology at different locations all play a role in thrombus composition. Although histology is useful in determining the overall composition of cellular structures in the true and false lumens after dissection occurs, SEM provides high-resolution, high-magnification images with three-dimensional information that can be used to identify all structures, including RBC morphology, in the false lumen. When thrombi form, RBCs go through a series of morphological changes from biconcave disks to polyhedrocytes as a consequence of the clot contraction process.
By stratifying RBCs into four groups (biconcave, intermediate biconcave, intermediate polyhedral, and polyhedrocytes), we estimated the degree of thrombus contraction, which affects thrombus stability and the tendency for embolization.
We observed very few biconcave RBCs in our SEM analysis, and most were located near a patent false lumen or dissection flap where there was active flow. Even in mice euthanized the same day as dissection was detected, few biconcave RBCs were present in the thrombus. This finding suggests RBCs undergo fast morphological changes accompanying clot contraction once entering the false lumen in less than 24 to 32 hours. We found greater percentages of intermediate biconcave RBCs in the proximal section compared with the distal suprarenal portion where more polyhedral RBCs and significantly more fibrin was present (Fig 4, D, F and H).The reasons for differences in fibrin content are not clear, but fibrin is necessary for clot contraction, in which platelets pull on fibrin surrounding RBCs, compressing them into polyhedrocytes.
We observed a wide range of flow patterns in these mice, with some exhibiting completely patent false lumens with recirculating flow. It is possible that greater amounts of fibrin are present in the areas directly bordering active flow, enclosing the RBCs and keeping the thrombus intact. Additionally, in some distal sections we observed circumferential alignment of thrombus structures, especially the orientation of fibrin (Fig 4, E). This outcome could be due to recirculating, complex flow patterns at those sites. In human patients, Tsai et al showed that complete thrombosis of the false lumen in Type B dissections reduces acute rupture risk, possibly owing to the decreased hemodynamic forces on the weakened vessel wall of the false lumen. However, in contrast with other studies, they showed patent false lumens have the highest survivability rate in the long term.
Generally, depending on the type of dissection, most studies have shown that patent or partially thrombosed false lumens have the worst long-term outcomes for patients.
Patent, partial, or complete thrombosis of the false lumen is influenced by maximum diameter of the false lumen, visceral branches off the false lumen, and number of re-entry tears.
In the animals we have analyzed (n = 5), the trends suggest that thrombus deposition and structure are affected primarily by local hemodynamics along the aorta. The ability to predict where and to what extent thrombus will be deposited in human patients could be invaluable to improving treatment decisions and outcomes. Future studies are needed to determine how wall shear stress, particle residence time, and pressure influence thrombus developments such that better models for prediction of thrombus deposition and structure can be created. The research methods described here could also be combined with prothrombotic therapeutic intervention studies to determine the potential feasibility and effects of acutely closing off the false lumen via medical therapy or surgical intervention.
In addition to our findings from ex vivo analyses, we observed that young mice develop dissections sooner (6.4 ± 0.5 days after AngII pump implant) than old mice (11.1 ± 0.7 days; P < .001). Although older mice developed dissections at a similar rate compared with young mice (53.8% old vs 60.0% young), it seems older mice are more resistive to dissection formation early after pump implantation. This finding is interesting, because a significant risk factor for ADs in humans is advanced age.
Our initial hypothesis focused on elucidating the morphological and hemodynamic changes that take place early on after pump implant, because the aorta remodels in response to AngII administration.
However, we found that both young and old mice experienced similar vessel diameter growth throughout the study (Fig 1, C). There were, however, significant differences (P = .05) in the baseline mean flow between young and old groups (Supplementary Fig 1, D). We also hypothesized that rapid expansion of the aorta in the initial days after pump implantation was a predictive factor determining which mice would later develop dissections. This hypothesis was not supported by the data as we found that the dissection and non-dissection groups experienced similar diameter expansion 3 days after pump implantation. Overall, these observations suggest a relationship between age and dissection timing after insult that warrants further investigation in both experimental models and patients.
There are several notable limitations of this study, especially when comparing the AngII murine model with human AD. First, the apoE−/−, AngII AD model is different than the majority of AD cases in humans that develop in the ascending thoracic aorta and have separations within the medial layer.
The AngII/apoE−/− murine dissection model typically manifests as separation between the medial and adventitial layers, potentially influencing intraluminal thrombus morphology. Second, we found that no dissections ruptured in this study, even when a patent false lumen was present. This finding is similar to previous studies that report either complete aortic rupture before a dissection occurs or a relatively slow-growing AD that rarely ruptures.
Nevertheless, the lack of rupture in this model makes it useful for studying chronic changes in thrombus structure. We also found that ultrasound detection paired with ex vivo SEM is an effective research technique for analyzing intramural thrombus formation, opening the potential for studies of thrombus formation and development in both animal models and human samples.
Finally, the intramural thrombus we observed is potentially different than intraluminal thrombus found in many patients with abdominal aortic aneurysm.
Thus, it would be interesting to apply this multimodality technique to experimental models of true or fusiform abdominal aortic aneurysms when intraluminal thrombus is observed.
Future studies could investigate chronic changes in thrombus formation with multiple, well-defined euthanasia timepoints to determine how thrombi expand and contract over time. We report blood velocity and flow changes within the true lumen of these complex aortic lesions (Supplementary Fig 1, F), but flow patterns in the false lumen may influence thrombus formation and alignment. Discovering how complex hemodynamics influence thrombus patterns observed in SEM images may require the use of computational fluid dynamics or fluid-structure interaction models with a focus on particle residence time and thrombus formation potential metrics.
Our multimodality approach for detecting AD occurrence with volumetric ultrasound imaging allows for the study of thrombi formation at acute and chronic timepoints. Daily, volumetric ultrasound scans can be performed in less than 10 minutes per animal, provide valuable information about the timing of dissection, and help to determine morphological changes before and after dissection. SEM data expands upon histological information by detailing thrombus structure, including RBC morphology, degree of thrombus deposition, and identification of individual thrombi formed at different times within the false lumen. We found that volumetric ultrasound scans can be aligned with SEM and histological sections, and flow characteristics and morphology can be matched to thrombus deposition and structural observations. Moreover, differences in thrombus deposition and structure were significantly influenced by the anatomical location along the aorta. Future work is needed to study how factors such as local flow patterns, wall shear stress, and particle residence time affect the trends observed here. These findings could have wider implications on our ability to predict thrombus deposition in human patients at early timepoints after dissection, potentially improving treatment decisions and patient outcomes.
Acknowledgments
The authors acknowledge Evan Phillips, Hannah Cebull, and Alycia Berman for their contribution collecting ultrasound images. We also acknowledge MacKenzie Macintosh for performing histology preparation and scanning. We acknowledge grant support (J.W.W.) NIH/NHLBI R01 HL148227, P01-HL40387, and R01 HL135254. Scanning electron microscopy was supported by NIH Shared Instrumentation Grant S10-OD018041 (J.W.W.). L.S. acknowledges the National Science Foundation for support under the Graduate Research Fellowship Program (GRFP) under grant number DGE-1842166. The authors acknowledge (C.J.G.) the Leslie A. Geddes Endowment for support with animal ultrasound imaging and histology.
Supplementary Methods
Sex differences and inclusion statement
Angiotensin II (AngII)-induced dissections form at a rate twice as high in male mice compared with female mice.
Therefore, to use fewer animals and study a significant number of dissections we focus on male mice in this study. We believe there are slight differences in hemodynamics and morphology between sexes, but the methods described could be applied to either sex in the future.
Methods
AngII apoE–/– aortic dissection model
All procedures were approved by Purdue University IACUC (#2002002016). Osmotic pumps were implanted subcutaneously on the right dorsal flank in male, C57Bl6/J apoE–/–, mice (n = 71), aged 9.7 to 21.7 weeks in the young group (n = 45) and 35.0 to 88.1 weeks in the older group (n = 26; Jackson Laboratories, Bar Harbor, ME). We loaded osmotic pumps with 300 μL of an AngII salt solution, released subcutaneously by the osmotic pump at a rate of 1000 ng/kg/min as previously described (Model 2004, ALZET, Cupertino, CA).
Of the 71 mice, 41 (57.7%) developed aortic dissections, 7 (9.9%) experienced rupture of the ascending thoracic aorta or were moribund, and 23 (32.4%) developed no aneurysm or dissection over 28 days of AngII administration.
In vivo ultrasound acquisition
Ultrasound images were acquired on a Vevo3100 ultrasound scanner using the MX550D transducer with a 25- to 55-MHz bandwidth, 40-MHz center frequency, and 40-μm axial resolution (FUJIFILM VisualSonics, Toronto, Ontario, Canada). We acquired images at baseline before pump implantation, daily from day 3 to days 14 to 17 after pump implantation, and a final checkpoint was acquired at day 28 after pump implant for mice that did not develop dissections within the first 14 to 17 days after pump implant (Fig 2, A). Fig 2, B, shows a diagram of anatomical structures observed with short axis (SAX) ultrasound examination. We acquired SAX brightness mode (B-mode), SAX three-dimensional (3DUS) volumetric scans, long axis (LAX) B-mode, LAX motion mode (M-mode), LAX color Doppler, and LAX pulsed-wave Doppler velocity images at baseline and once aortic dissection was detected. SAX B-mode, SAX 3DUS, and LAX B-mode images were collected during daily dissection checks. Dissection checks were performed 24 to 32 hours apart from one another enabling us to detect dissections within 24 to 32 hours of occurrence.
Ultrasound diameter and flow rate measurements
An ultrasound data analysis was performed in VevoLAB (v5.7.0; FUJIFILM VisualSonics). We calculated effective diameter from circumference measurements at three locations along the suprarenal aorta using 3DUS volumetric images (Fig 1, A, B). We used pulsed-wave Doppler velocity traces to calculate mean velocity in the aorta and paired velocity measurements with area calculations taken at the same location to determine mean flow in the suprarenal aorta.
We euthanized mice with dissections either (1) the same day, (2) 3 days after, or (3) 2 to 3 weeks after dissection was detected with ultrasound. Mice were euthanized with CO2 asphyxiation, the heart and vasculature were perfused with saline, and the aorta was removed from the ascending arch down to the renal arteries. The aorta was divided into 12 equal sections starting immediately proximal to the renal arteries and ending in the descending thoracic aorta (Fig 2, C). All sections were washed with 50 mmol/L sodium cacodylate buffer diluted in 150 mmol/L sodium chloride and fixed overnight in 2% glutaraldehyde diluted in 50 mmol/L sodium cacodylate and 150 mmol/L sodium chloride buffer, as described previously.
Computer modeling of fibrin polymerization kinetics correlated with electron microscope and turbidity observations: clot structure and assembly are kinetically controlled.
Aortic samples with varying hemodynamics and dissection morphology were selected (n = 4) for histology and scanning electron microscopy (SEM) analysis. Odd numbered sections 1 to 11 from each mouse were prepared for histology. Aortic sections were embedded in paraffin, sectioned, and stained with both hematoxylin and eosin and Movat’s pentachrome stains. Quantitative histology measurements were performed in FIJI ImageJ using the color segmentation plug-in (n = 4).
The percentage area of red blood cells (RBCs), fibrin, collagen, elastin, and proteoglycans were quantified in each section (Supplementary Fig 7, A, B). Both the aortic wall and false lumen were included in this quantitative assessment, but the surrounding tissue was cropped, and the background was removed (Fig 3, A). Sections were also analyzed and scored by a board-certified veterinary pathologist. Semiquantitative inflammatory scores were provided by the veterinary pathologist using the Supplementary Table.
SEM and cell counter analysis
The even numbered sections 2 to 12 were prepared for SEM imaging from each of the selected mice. After fixation, sections were dehydrated using a graded series of ethanol solutions and hexamethyldisilazane as previously described.
Computer modeling of fibrin polymerization kinetics correlated with electron microscope and turbidity observations: clot structure and assembly are kinetically controlled.
Sections were mounted on SEM stubs, sputter coated with gold palladium, and high- and low-magnification images were acquired on either a FEI Quanta 250FEG (FEI Company, Hillsboro, OR) or TESCAN Vega 3 (TESCAN, Brno, Czech Republic) scanning electron microscope. High magnification SEM images (×2000) were collected at each section of the selected aortas (n = 5). We identified one to four imaging locations per section based on the individual clot structures observed (Fig 4, A). A scale and grid (15 μm2 per square) was applied to each image using the scale bar in FIJI ImageJ (Supplementary Fig 7, C). Grid size was set so that individual squares were roughly the same size as one RBC (Supplementary Fig 7, D). The cell counter plug-in was used to count the number of biconcave RBCs, intermediate biconcave RBCs, intermediate polyhedral RBCs, polyhedral RBCs, fibrin, and other structures (Supplementary Fig 7, E, F).
Cell counts were used to calculate percentages of the total space occupied by each structure.
Statistical analysis
Statistical analysis was performed in GraphPad Prism v9.4.1. The data were checked for a normal distribution using a Shapiro-Wilk test and QQ-plot normality check. A repeated measures analysis of variance with a Tukey or Holm-Sidak correction test for multiple comparisons was used when there were multiple comparisons, and a two-tailed t-test was performed on datasets with less than three comparison groups. Statistical significance was set at a P value of less than .05.
Supplementary Fig 1Color Doppler imaging displayed the presence or lack of active flow in the false lumen and pulsed-wave Doppler measured velocity measurements showed increases in true lumen velocity after intramural thrombus constricts/applies force to the true lumen. (A) LAX color Doppler shows blood flow in the true lumen (red) traveling from proximal to distal while we observed recirculating flow in the false lumen (blue). (B) SAX color Doppler shows similar flow in the true and false lumens. Yellow, true lumen; cyan, patent false lumen; green, thrombosed false lumen. (C) We observed increases in true lumen flow and velocity after dissections occurred owing to constriction of the true lumen by the false lumen. Scale bars, 1 mm. (D) Old mice had significantly higher mean flow rates at baseline compared with young mice. (E) There were no significant differences in the mean flow at baseline between the dissection and no dissection group. (F) There were significant increases in mean velocity once a dissection occurred as the true lumen is compressed. ∗ ; ∗∗.
Supplementary Fig 2After pump implantation, all groups underwent a significant increase in aortic diameter from baseline to maximum diameter before dissection. There were no significant differences in diameter growth percentage between the young and old or dissection and no dissection groups. (A, B) All mice experienced aortic diameter growth from baseline measurements to 3 days after pump implantation with the young mice and ones that developed dissections increasing slightly more than old and nondissecting mice, respectively. (C, D) Mice experienced significant aortic dilation from baseline to maximum diameter, but significant diameter growth was not deterministic in the development of dissections. ∗∗∗ ; ∗∗∗∗.
Supplementary Fig 3Hematoxylin and eosin (H&E)-stained aortic slices display extracellular matrix, red blood cells (RBCs), and nuclei present in the true and false lumen of the aortic dissections (ADs). H&E samples were examined for intramural thrombus presence and organization, immune cell infiltration, intramural hemorrhage, atherosclerosis, and tertiary lymphoid organs and differences in proximal, middle, and distal sections were observed. (A-D) Four ApoE–/– mice displayed in columns A-D, where the top sections were taken from the descending thoracic aorta down to the bottom section taken just proximal to the renal arteries. Scale bars, 500 μm.
Supplementary Fig 4Movat’s pentachrome stained slides identified RBCs (purple), fibrin (red-pink), collagen (yellow-brown), elastin (black), and proteoglycans (blue). Fiji ImageJ color segmentation revealed large numbers of red blood cells (RBCs) followed closely by fibrin in the false lumen. Each section is heterogenous with some containing more fibrin than RBCs and varying amounts of collagen, elastin, and proteoglycans. Collagen consistently remained localized to the outer adventitia. (A-D) Four ApoE–/– mice displayed in columns A-D with top to bottom being proximal to distal. Scale bars, 500 μm.
Supplementary Fig 5Low magnification scanning electron microscopy (SEM) images were used to identify different thrombi structures in the false lumen of the AD. (A-D) Four ApoE–/– mice displayed in columns A-D with top to bottom being proximal to distal. Scale bars, 500 μm.
Supplementary Fig 6Structures identified in thrombi from aortic dissection (AD) murine model. (A) Biconcave red blood cells (RBCs), black arrows. (B) Intermediate biconcave RBCs, black arrows. (C) Polyhedral shaped RBCs, black arrows. (D) Intermediate polyhedral RBCs, black arrows. (E) Fibrin, black arrows. Other structures include: (F) collagen, (G) neutrophils, (H) fibroblasts in black arrows. Scale bar, 10 μm.
Supplementary Fig 7We used FIJI ImageJ color segmentation and cell counter plugins to quantify cellular structures in histology and scanning electron microscopy (SEM) images. Histology provides compositional information for the entire aorta while SEM allows for the stratification of red blood cell (RBC) morphology and the presence of fibrin strands in the intramural thrombus. (A) Movat’s pentachrome stains for RBCs (dark red/maroon), fibrin (light red/magenta), proteoglycans (blue), elastin (black), and collagen (yellow/brown). Scale bar, 500 μm. (B) Color segmentation plugin in ImageJ was used to sample colors for each cellular structure and background (white). The hidden Markov model algorithm was applied to calculate the percentages of each color present. (C) After loading a high magnification (2000×) SEM image, set the scale of the image in pixels. (D) Apply a grid size that roughly encompasses one RBC per square. Grid size = 15 μm2. (E) Rename each counter to correspond with the RBC type, fibrin, or other/space. (F) Designate each square as a RBC type, fibrin, or other/space depending on what structure the majority of the square is occupied by.
Supplementary TableSemiquantitative scoring by a board-certified veterinary pathologist identified the presence or absence of intramural hemorrhage, thrombus, and immune cells as well as thrombus organization, atherosclerosis, perivascular immune cells, and tertiary lymphoid organs. Below is a scoring rubric used to analyze H&E-stained sections
Intramural hemorrhage
0 = absent 1 = present
Intramural thrombus
0 = absent 1 = present
Thrombus organization
0 = No thrombus 1 = fibrin separate from wall 2 = incorporation into wall 3 = immune cell infiltration into thrombus 4 = organized mesenchymal cell ingrowth into thrombus
Intramural immune cells (vasculitis)
0 = absent 1 = mild, few leukocytes scattered in the wall 2 = moderate, <50% of wall circumference is occupied by leukocytes 3 = marked, >50% of wall circumference is occupied by leukocytes
Perivascular immune cells
0 = absent 1 = mild, few leukocytes scattered outside the wall 2 = moderate, <50% of outside wall circumference is occupied by leukocytes 3 = marked, >50% of outside wall circumference is occupied by leukocytes
Population-based assessment of the incidence of aortic dissection, intramural hematoma, and penetrating ulcer, and its associated mortality from 1995 to 2015.
Computer modeling of fibrin polymerization kinetics correlated with electron microscope and turbidity observations: clot structure and assembly are kinetically controlled.
The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS-Vascular Science policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest.
00009-3&id=gr1.jpg){kind=link}
00009-3&id=gr2.jpg){kind=link}
00009-3&id=gr3.jpg){kind=link}
00009-3&id=gr4.jpg){kind=link}
00009-3&id=fx1.jpg){kind=link}
00009-3&id=fx2.jpg){kind=link}
00009-3&id=fx3.jpg){kind=link}
00009-3&id=fx4.jpg){kind=link}
00009-3&id=fx5.jpg){kind=link}
00009-3&id=fx6.jpg){kind=link}
00009-3&id=fx7.jpg){kind=link}