STENT-GRAFT COMPLIANCE MODULATES REGIONAL MECHANICAL AND STRUCTURAL REMODELING OF THE AORTA

Abstract

Endovascular aortic stent-grafts often introduce a compliance mismatch with the native aorta, potentially leading to adverse remodeling. This study assessed how stent-graft compliance affects structural, mechanical, and morphometric adaptation in a porcine model by comparing a conventional stiff polyethylene terephthalate stent-graft (CS-SG) to a compliant nanofibrillar stent-graft (NF-SG). Four months after implantation, aortas were harvested and systematically analyzed. Morphometric assessment of aortic rings, planar biaxial mechanical testing, and quantification of elastin, collagen, and smooth muscle cell (SMC) densities were performed. Remodeling with the CS-SG was primarily localized to regions under stent contact and included increased collagen accumulation (~25-30% vs ~15% in the inner media), elastin compaction, significant loss of SMCs (~10-15% vs ~20-25% in inner media), and thick neointima formation. These changes were associated with increased aortic stiffness (0.47 MPa vs. 0.37 MPa) and loss of non-linearity, especially in distal stented segments, while remote regions exhibited minimal alterations. In contrast, NF-SG implantation was associated with fewer localized structural and mechanical changes, including preserved SMC density, reduced collagen accumulation, reduced neointima formation, and maintained compliance at stented and adjacent regions. Overall, these findings suggest that compliant stent-grafts may mitigate localized adverse remodeling at stent–tissue interfaces compared to conventional stiff devices. These results highlight the potential importance of compliance-matching in stent-graft design, although broader effects across the aorta were limited within the four-month observation period. Future studies in diseased and aged models with extended follow-up are warranted.

Graphical Abstract

1. INTRODUCTION

The aorta, as the largest blood vessel in the human body, is susceptible to a range of pathologies associated with serious cardiovascular complications. Among these, aortic aneurysms and dissections in elderly individuals[1], as well as traumatic injuries such as blunt aortic trauma and aortic transection[2], which are more common in younger populations, represent acute, life-threatening conditions. These pathologies can affect both the thoracic and abdominal segments of the aorta and often require immediate surgical intervention. In recent years, endovascular aortic repair (EVAR) and thoracic endovascular aortic repair (TEVAR) have become a standard of care for treating aortic pathologies[3] in addition to surgical repair.

These minimally invasive endovascular techniques, which involve the placement of a stent-graft within the diseased segment of the aorta, are now increasingly used in both elderly and younger patients, accounting for approximately 80% of aneurysm repairs in recent years[4]. However, despite their benefits, these procedures are associated with several complications that may arise months to years after surgery[5]. The first group includes device-related complications such as graft failure and endoleaks, which often require secondary interventions[6]. The second group involves more systemic consequences, including long-term changes in aortic and cardiac function as well as adverse cardiovascular remodeling[5,7]. Previous studies have shown that during the early and chronic phases following stent-grafting, patients may develop aortic stiffening[8], increased blood pressure[9], alterations in left ventricular mass[10] and volume[11], and, in severe cases, cardiac events such as arrest. These mid-to-late term complications are believed to stem primarily from the disruption of the aorta’s key physiological role, known as the Windkessel function. This mechanism relies on the aorta’s elastic properties to buffer the high-volume pulse ejected from the left ventricle during systole by expanding, and to recoil during diastole, thereby ensuring continuous and stable blood flow to distal organs[12].

A key factor underlying these complications is the compliance mismatch introduced by stent-grafts[13], which are typically constructed from stiff materials such as polyethylene terephthalate (PET) or expanded polytetrafluoroethylene (ePTFE), in contrast to the elastic nature of the native aorta[14]. This mismatch in mechanical properties highlights the critical need to develop stent-grafts that more closely resemble native aorta, in order to mitigate long-term complications and maladaptive vascular remodeling. This need is particularly urgent as (T)EVAR is increasingly applied to younger patients, whose naturally compliant aortas are more susceptible to the effects of stiffness mismatch and who face a longer lifetime exposure to these adverse changes[15,16].

Our team recently developed a novel Elastomeric Nanofibrillar Graft (ENG) with artery-tuned nonlinear compliance. Human and porcine arteries were mechanically characterized to establish target properties, and the ENG was fabricated using electrospinning techniques that introduced residual strain into aligned nanofibers to replicate the native vessel's mechanical behavior (for detailed graft fabrication methods, see[17]). This compliant graft material was integrated with a custom-designed nitinol skeleton to produce a new generation of stent-graft with improved compliance matching (for detailed stent-graft design, see[18,19]). The resulting device was systematically evaluated for its hemodynamic performance, including changes in pulse wave velocity (PWV), pulse pressure (PP), and aortic distensibility, using both bench-top in vitro mock-loop systems[18] and in vivo porcine models[19], as well as for its impact on left ventricular remodeling in vivo [20]. Compared to commercially available stiff stent-grafts, the ENG-based stent-graft demonstrated superior hemodynamic characteristics at a 4.3-month follow-up[19].

While in vivo studies are valuable for assessing functional hemodynamic outcomes, they offer limited insight into the underlying biomechanical and structural remodeling of the aortic wall. To fully understand the long-term consequences of stent-graft implantation, it is essential to evaluate additional parameters beyond pressure-flow metrics. These include intrinsic mechanical properties of the aortic wall, changes in the content and organization of key structural components such as elastin, collagen and smooth muscle cells (SMCs), and alterations in aortic geometry and morphometry. Such parameters directly reflect the tissue-level remodeling response and provide a more complete understanding of how stent-graft compliance influences aortic biomechanics over time. These assessments require post-sacrifice analysis, as they cannot be reliably or systematically obtained through in vivo imaging or pressure-based measurements alone.

In this study, we analyzed aortic tissues harvested from a previously conducted in vivo porcine experiment to investigate the effects of stent-graft compliance on vascular remodeling. Our objective was to characterize and compare the mechanical, structural, and geometric properties of the aorta following implantation of a compliant nanofibrillar stent-graft (NF-SG) and a conventional stiff PET-based stent-graft (CS-SG). Post-sacrifice analysis included planar biaxial mechanical testing to assess intrinsic wall stiffness, histological imaging to evaluate the content and organization of key structural components, and morphometric measurements to quantify geometric remodeling. This comprehensive, multi-metric approach allowed us to systematically compare chronic aortic changes between treatment groups and evaluate the impact of stent-graft compliance on preserving native aortic biomechanics.

2. METHODOLOGY

2.1. Study design

Two different stent-grafts were used in this study. For the CS-SG group, two overlapping Gore Excluder AAA Contralateral Limb endoprosthesis (W.L. Gore & Associates, USA) were used. The NF-SG prototype consisted of an elastomeric nanofibrillar graft (ENG) integrated with a nitinol support skeleton. The ENG was fabricated by electrospinning biomedical-grade polyether urethanes (Pellethane^® 5863-82A and 2363-55DE) on a conical mandrel, producing a tapered tubular graft. The ENG fabric had a circumferential stiffness of 1.6 MPa in the physiologic stress range (75–100 kPa), which is substantially lower than conventional PET grafts (~9.5 MPa) and closer to the compliance of native porcine thoracic aorta (~0.4-0.5 MPa). Additional details on the design and fabrication of the NF-SG can be found in [18,19].

The complete animal study protocol is described in [19]. The study was approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Nebraska Medical Center. A total of 20 Yucatan mini-pigs (baseline weight: 58.3 ± 4.8 kg; age: 13.6 ± 1.9 months) were randomly assigned to three experimental groups, including a control group with no stent-graft implantation (N=7), a CS-SG group implanted with a commercially available stiff stent-graft (N=8), and an NF-SG group implanted with a compliant stent-graft (N=5). All animals underwent two surgical procedures, with the second occurring 18 weeks after the first. In the stent-grafted groups (CS-SG and NF-SG), the initial procedure consisted of endovascular stent-grafting of the descending thoracic aorta (DTA), and the second procedure was performed for long-term hemodynamic assessment and necropsy. In contrast, control pigs underwent both procedures without receiving any stent-graft, serving as a baseline to assess changes in the aortic wall associated with normal growth.

In the stent-grafted groups, all stent-grafts were deployed under fluoroscopic guidance in the proximal descending thoracic aorta, with the proximal landing zone positioned just distal to the left subclavian artery. Device sizing was based on baseline CTA with a target oversizing of approximately 10% at the proximal side. In the CS-SG group, the graft was expanded to its final diameter immediately after deployment using balloon-expandable stents placed proximally, distally, and mid-graft to reproduce the chronic loss of pulsatility typically reached after 6-12 months of remodeling in human TEVAR. The NF-SG devices were implanted using the same landing zone and oversizing strategy but without balloon-expandable stents. Further details regarding the surgical protocol and animal procedures are described in our previous report[19].

2.2. Specimen preparation

At 18 weeks post-implantation, the animals were euthanized, and the aortas were harvested for biomechanical, histological, and morphometric analysis. At the time of tissue harvest, arterial pressures were recorded to verify comparable physiological conditions across treatment groups. Pulse pressure (Control: 29.9 mmHg, NF-SG: 26.9 mmHg, CS-SG: 28.8 mmHg, p = 0.760), mean arterial pressure (Control: 96.3 mmHg, NF-SG: 86.7 mmHg, CS-SG: 90.3 mmHg, p = 0.773), and systolic arterial pressure (Control: 116.2 mmHg, NF-SG: 104.7 mmHg, CS-SG: 109.6 mmHg, p = 0.744) were similar among groups, confirming no significant hemodynamic differences at the time of aorta excision. The number and location of tissue samples varied depending on whether the animal belonged to the stent-grafted or control group. In control pigs, tissue was collected from five locations, including the middle of the ascending thoracic aorta (ATA), proximal descending thoracic aorta (pDTA), middle descending thoracic aorta (mDTA), distal descending thoracic aorta (dDTA), and abdominal aorta (ABA). These locations are illustrated in Figure 1. In the stent-grafted groups, samples were collected from seven locations. Five of these (ATA, pDTA, mDTA, dDTA and ABA) corresponded to the same anatomical regions as in the control group. Since mDTA segment was in direct contact with the stent-graft in these animals, it is denoted as mDTA-st. Additionally, two extra samples were collected from the proximal and distal stent-grafted portion of the aorta (pDTA-st and dDTA-st), enabling direct comparisons of adjacent stented and unstented tissue. Not all segments were available in all animals.

Figure 1.

Anatomical locations of tissue sample collection in porcine aortas. In the control group, samples were harvested from the ascending thoracic aorta (ATA), proximal descending thoracic aorta (pDTA), middle of descending thoracic aorta (mDTA), distal descending thoracic aorta (dDTA), and abdominal aorta (ABA). In the stent-grafted group, in addition to matching non-stented locations, samples were also collected from proximal and distal DTA locations directly contacting the stent-graft: pDTA-st and dDTA-st. Gray shading indicates the stent-grafted regions of the aorta.

From each anatomical location, several specimens were cut for mechanical, structural, and morphometric analyses: two thin (~1 mm) rings, one retained as a load-free ring, and one cut radially from the posterior side to serve as a stress-free ring (cut-ring); and a 13 mm × 13 mm square specimen was collected from the lateral wall for planar biaxial mechanical testing.

2.3. Morphometry measurements

Ring and cut-ring samples were submerged in phosphate-buffered saline (PBS) at room temperature for at least 30 minutes prior to imaging (Figure 2A). After equilibration, each specimen was photographed and analyzed using a custom-developed image-processing software[21]. Cut-ring samples were analyzed to measure the opening angle, stress-free wall thickness, and stress-free lengths of the intimal and adventitial layers. The inner and outer surfaces of each cut-ring were manually traced as arcs (red and blue arcs in Figure 2A), and circles were fitted to these contours. Stress-free wall thickness was measured as the difference between the radii of these fitted circles. The opening angle was measured as the angle between the start and end points of the traced arc. Intimal and adventitial stress-free lengths were measured as the arc lengths of the inner and outer surfaces. Load-free ring samples were used to measure aorta radius and thickness in the unloaded but intact configuration. The intimal and adventitial layers (green and brown arcs in Figure 2A) were traced, and circles were fitted to the inner and outer surfaces. The difference between the inner and outer radii was taken as the load-free wall thickness. In stent-grafted aortas, ring samples could not be obtained at locations that were in direct contact with the stent-graft, specifically the pDTA-st, mDTA-st, and dDTA-st, because the aorta had to be cut longitudinally to remove the stent-graft prior to further processing.

Figure 2.

Experimental procedures for morphometric, mechanical, and histological analysis of aortic samples. (A) Representative image of load-free and stress-free ring specimens used for morphometry analysis. Opening angle, wall thickness, radii, and arc lengths (perimeter) were quantified from traced inner and outer surfaces. (B) Planar biaxial testing setup showing a 13 × 13 mm square aortic sample mounted along longitudinal and circumferential axes. Fiducial graphite markers were applied for optical tracking. (C) Representative stained ring section divided into four medial layers (M1–M4) from adventitia to intima. (D) Component-specific RGB thresholding across defined regions.

2.4. Biaxial mechanical testing and analysis

Planar biaxial mechanical testing was performed on 13 mm × 13 mm square specimens to assess the passive mechanical behavior of the aortic wall. Samples were mounted using rakes in a BioTester (CellScale, Canada), keeping longitudinal and circumferential directions aligned with x- and y-axes (Figure 2B). Testing was performed in 0.9% PBS at 37 °C using 2.5 N load cells. Graphite powder was applied to the specimen surface as fiducial markers to track deformations. Prior to testing, all samples underwent 20 cycles of equibiaxial preconditioning to establish repeatable force-stretch responses. Mechanical characterization was then performed using a stretch-controlled multi-ratio protocol consisting of 21 loading sequences. These included stretch ratios ranging from 1:0.1 to 1:0.9 (circumferential: longitudinal) and 0.9:1 to 0.1:1, with increments of 0.1. Equibiaxial loading (1:1) was interspersed at the beginning, middle, and end of the test sequence to monitor potential tissue damage and ensure data consistency. The maximum stretch applied in each direction was tailored for each specimen to obtain the non-linear response while avoiding structural damage, with values ranging from 1.15 to 1.5 depending on tissue compliance. All tests were performed at a constant strain rate of 0.01 s⁻¹. Assuming incompressibility, Cauchy stresses averaged through thickness were calculated as follows:

$$t_{zz}^{exp}={\lambda }_z\frac{p_z}{H{L}_{\theta }}$$ Equation 1

$$t_{\theta \theta }^{exp}={\lambda }_{\theta }\frac{p_{\theta }}{H{L}_z}$$ Equation 2

Here, $p_z$ and $p_{\theta }$ are the applied loads in the longitudinal and circumferential directions, and $L_z$ and $L_{\theta }$ are the corresponding undeformed lengths, ${\lambda }_z$ and ${\lambda }_{\theta }$ are the biaxial stretches, and $H$ is the reference thickness.

To enable quantitative analysis, several mechanical indices were derived from the equibiaxial stress-stretch curves. Tissue stretches (compliance) were measured at two stress levels of 50 and 125 kPa. These values represent vessel loading conditions throughout the cardiac cycle, with the 50 kPa approximating a low-stretch regime and 125 kPa corresponding to a high-stretch regime. In addition, the tangent modulus, defined as the local slope (derivative) of the stress-stretch curve at these two stress levels (50 and 125 kPa), was calculated to estimate the intrinsic stiffness of the tissue. These were referred to as the low-range (50 kPa) and high-range (125 kPa) tangent moduli. The non-linearity index was defined as the ratio of high-range to low-range tangent moduli, characterizing the degree of J-shaped behavior. Finally, the anisotropy index, based on the formulation proposed by Bogoni et al.[22], was calculated to quantify directional mechanical differences in the aortic wall within the stress range of 50 kPa to either 125 kPa or the maximum common stress reached in both directions, whichever was lower. The index was defined as:

$$\eta (t)=\text{sgn}[\xi (t)]\frac{1-\xi (t)}{1+\mid \xi (t)\mid }$$ Equation 3

where $\xi (t)=\frac{{\lambda }_{\theta \theta }(t)}{{\lambda }_{zz}(t)}$, with ${\lambda }_{zz}$ and ${\lambda }_{\theta \theta }$ representing stretch values in the longitudinal and circumferential directions, respectively, at matched stress levels. By design, $\eta =0$ indicates isotropic behavior (equal stretch in both directions), values approaching +1 indicate higher compliance in the longitudinal direction (i.e., $\eta >0$ and ${\lambda }_{zz}>{\lambda }_{\theta \theta }$), and values approaching −1 reflect greater compliance in the circumferential direction (i.e., $\eta <0$ and ${\lambda }_{zz}<{\lambda }_{\theta \theta }$).

2.5. Histological analysis

To evaluate the wall structure and organization, we performed histological imaging on ring specimen from each aortic location. Prior to imaging, specimens were fixed in Methacarn for at least 48 hours, dehydrated in 70% ethanol, and embedded in paraffin. Sections were then stained with Movat’s Pentachrome (Movat) and Masson’s Trichrome (MTC) to visualize and quantify elastin, collagen, and SMCs. Histological images were analyzed using a semi-automated, custom-developed software that performed RGB color thresholding to quantify the area fraction of these constituents[21]. For this analysis, multiple regions of the aortic media were selected from both ring and longitudinal samples. Each region was subdivided into four layers of equal thickness across the media, labeled M1, M2, M3, and M4, progressing from the adventitia towards the intima (Figure 2C). This transmural approach is important as the composition of the vessel wall is not uniform across its thickness. Specific color thresholds were set for elastin (black, from Movat's stain), collagen (blue, from MTC), and SMCs (red, from MTC). The area fraction was calculated as the number of pixels for a given component divided by the total pixels in the analyzed region. To reduce intra-operator error, a single operator performed all analyses twice, and the results were averaged.

2.6. Statistical analysis

Statistical analyses were performed using R within the RStudio environment to assess differences in morphometrical, mechanical, and histological features of the aorta across treatment groups (control, CS-SG, and NF-SG) and anatomical locations. The Shapiro-Wilk test was used to assess normality, while Levene’s test evaluated the homogeneity of variances. Depending on the data characteristics, group comparisons were conducted using t-tests, one-way ANOVA, one-way repeated measures ANOVA, or mixed ANOVA, as appropriate. When assumptions of normality or homogeneity were violated, non-parametric alternatives to t-tests and ANOVA were applied. For repeated measures analyses, the assumption of sphericity was tested using Mauchly’s test, and Greenhouse-Geisser corrections were applied when this assumption was violated. Post-hoc pairwise comparisons were carried out with Bonferroni correction to identify groups with significant differences following ANOVA tests. A p-value less than 0.05 was considered statistically significant.

At the proximal and distal landing zones of stent-grafted aortas, two specimens were typically collected (pDTA and pDTA-st; dDTA and dDTA-st). In contrast, the control group provided only one specimen at each corresponding region (pDTA and dDTA), which served as the baseline for comparisons with both stented and non-stented segments from the stent-grafted groups. In some cases, data from the NF-SG group, particularly for mechanical testing and collagen and SMC density, were missing or limited at pDTA/-st and dDTA-st. As a result, the NF-SG group was excluded from statistical analyses at these locations when appropriate. In addition, CS-SG at pDTA location had a limited number of specimens for mechanical testing and was also excluded from comparison when appropriate.

3. RESULTS

3.1. Morphometrical analysis

Six morphometrical features were calculated in this study to evaluate geometric changes in the aorta between control and stent-grafted groups at non-stented locations, as shown in Figure 3. The opening angle remained similar across treatment groups. The outer and inner radii of the aorta were also comparable across groups and locations (ATA: p = 0.969 and 0.970; pDTA: p = 0.163 and 0.216; dDTA: p = 0.063; ABA: p = 0.547 and 0.261), except at dDTA, where the NF-SG group showed a 14% larger outer radius compared to the control group (6.4 mm vs. 5.6 mm, p = 0.022). The ratio of stress-free thickness to load-free thickness (thickness ratio) was also measured and compared between groups (Figure 3D). While this ratio remained similar between groups in proximal locations (ATA: p = 0.682; pDTA: p = 0.788), changes were observed in distal segments. Specifically, at dDTA, the CS-SG group exhibited 20% smaller value than both the control (0.8 vs. 1.0, p = 0.033) and NF-SG (0.8 vs. 1.0, p = 0.060) groups. At ABA, the CS-SG group showed a significantly smaller value than NF-SG (0.7 vs. 0.9, p = 0.020). Additionally, the ratio of outer perimeter between stress-free and load-free samples was calculated and compared (Figure E). No significant differences were found among groups at any location. The load-free wall thickness, calculated as the difference between the outer and inner radii (Figure 3F), was similar across treatment groups in the ABA and ATA regions. However, in the pDTA, the NF-SG exhibited greater thickness compared to controls (2.2 mm vs. 1.8 mm, p = 0.010), and in the dDTA, the CS-SG group showed increased thickness relative to controls (1.6 mm vs. 1.3 mm, p = 0.040). Due to the lack of load-free ring specimens in the stented locations, morphometrical features could not be measured in these regions (pDTA-st, mDTA-st and dDTA-st).

Figure 3.

Morphometric parameters of non-stented aortic segments across treatment groups. (A) Opening angle, (B) outer radius, (C) inner radius, (D) thickness ratio (stress-free to load-free), (E) outer perimeter ratio (stress-free to load-free), and (F) load-free thickness were measured at four non-stented anatomical locations (ATA, pDTA, dDTA, ABA) for control (black), CS-SG (red), and NF-SG (blue) groups. Data are shown as mean ± SD. Statistical comparisons were performed using one-way ANOVA between treatment groups within each anatomical location, followed by post-hoc pairwise comparisons where applicable.

Statistically significant morphometric, mechanical, and histological differences across treatment groups and anatomical locations are summarized in Table 1.

Table 1.

Summary of statistically significant morphometric, mechanical, and histological differences between treatment groups across anatomical locations. Only statistically significant comparisons (p < 0.05) are shown; blank cells indicate non-significant differences.

	Parameter	Location	Direction / Wall Region	Control vs CS-SG	Control vs NF-SG	CS-SG vs NF-SG
Morphometry	Outer radius (mm)	dDTA	NA		5.6 vs 6.4 (p = 0.022)
	Thickness ratio	dDTA	NA	1.0 vs 0.8 (p = 0.033)
		ABA	NA			0.7 vs 0.9 (p = 0.020)
	Thickness (mm)	pDTA	NA		1.8 vs 2.2 (p = 0.01)
		dDTA	NA	1.3 vs 1.6 (p = 0.040)
Mechanical	Stretch at 50 kPa	ABA	Long			1.27 vs 1.34 (p = 0.008)
		ABA	Circ		1.11 vs 1.20 (p = 0.003)	1.13 vs 1.20 (p = 0.016)
		dDTA-st	Long	1.34 vs 1.27 (p = 0.023)
	Stretch at 125 kPa	ABA	Circ		1.16 vs 1.24 (p = 0.016)
	Low-Range Tangent Modulus (MPa)	dDTA-st	Long	0.37 vs 0.47 (p = 0.049)	NC	NC
	High-Range Tangent Modulus (MPa)	dDTA-st	Circ	2.13 vs 0.92 (p = 0.024)	NC	NC
		mDTA-st	Circ		1.10 vs 1.76 (p = 0.022)	0.90 vs 1.76 (p = 0.003)
	Non-Linearity Index	dDTA-st	Long	2.23 vs 1.64 (p = 0.048)	NC	NC
		dDTA-st	Circ	2.68 vs 1.54 (p = 0.004)	NC	NC
	Anisotropy Index	mDTA-st	NA			0.007 vs 0.037 (p = 0.010)
		pDTA-st	NA	0.028 vs 0.014 (p = 0.041)	NC	NC
		dDTA-st	NA	0.0 56 vs 0.022 (p = 0.010)	NC	NC
Histological	Elastin Density (%)	dDTA	M1-M4	v	26.4 vs 30.5 (p = 0.034)	26.8 vs 30.5 (p = 0.050)
		pDTA-st	M3	38.2 vs 48.1 (p = 0.014)		37.3 vs 48.1 (p = 0.011)
		mDTA-st	M4	31.9 vs 39.9 (p = 0.020)
		dDTA-st	M1-M4	26.4 vs 32.3 (p = 0.007)	26.4 vs 34.2 (p = 0.002)
	Collagen Density (%)	pDTA-st	M3	15.3 vs 25.0 (p = 0.003)	NC	NC
		pDTA-st	M4	14.8 vs 23.1 (p = 0.050)	NC	NC
		mDTA-st	M4	16.0 vs 29.7 (p = 0.004)		29.7 vs 19.0 (p = 0.035)
		dDTA-st	M1	19.0 vs 12.2 (p = 0.009)	NC	NC
		dDTA-st	M2	14.6 vs 10.5 (p = 0.025)	NC	NC
	Smooth Muscle Cells (%)	pDTA	M1	19.7 vs 28.6 (p < 0.001)		28.6 vs 20.6 (p = 0.001)
		pDTA	M2	23.6 vs 30.1 (p = 0.042)
		pDTA-st	M3	24.8 vs 13.4 (p = 0.004)	NC	NC
		pDTA-st	M4	20.0 vs 9.6 (p = 0.002)	NC	NC
		mDTA-st	M1		17.9 vs 28.6 (p = 0.024)
		dDTA-st	M3	27.3 vs 19.3 (p = 0.004)	NC	NC
		dDTA-st	M4	19.8 vs 9.7 (p = 0.002)	NC	NC

NA denotes parameters for which direction or wall region was not applicable. NC denotes pairs were not compared due to limited sample size.

3.2. Mechanical analysis

The full set of equibiaxial stress–stretch curves for all non-stented and stented aortic locations, as well as individual curves for each animal, are provided in the Supplementary Material (Figures S1 and S2). Quantitative mechanical parameters derived from these curves are summarized in Figures 4 and 5.

Figure 4.

Quantified tissue compliance at two physiological stress levels in longitudinal and circumferential directions. Stretch values at 50, and 125 kPa are shown for non-stented (A, B, E, and F) and stented (C, D, G, and H) aortic segments across control (black), CS-SG (red), and NF-SG (blue) groups. Panels A, E, C, and G display longitudinal stretch, while panels B, F, D, and H present circumferential stretch. Data are shown as mean ± SD. Statistical comparisons were performed using one-way ANOVA between treatment groups within each anatomical location, followed by post-hoc pairwise comparisons where applicable.

Figure 5.

Biomechanical metrics derived from equibiaxial mechanical testing across anatomical locations and treatment groups. (A) Tangent modulus calculated in circumferential at 125 kPa, providing estimates of tissue stiffness under elevated loading conditions. (B) The anisotropy index was calculated as the ratio of longitudinal to circumferential compliance, representing directional mechanical differences in the aortic wall. (C-D) The non-linearity index, defined as the ratio of high-range to low-range modulus, quantifies the extent of strain-stiffening behavior in both directions. Box plots show the median and interquartile range for control (black), CS-SG (red), and NF-SG (blue) groups across all tested regions. Statistical comparisons were performed using one-way ANOVA between treatment groups within each anatomical location, followed by post-hoc pairwise comparisons where applicable. Data from pDTA and pDTA-st are not shown due to insufficient sample size, and were excluded from the analysis.

At 50 kPa, no significant differences in compliance were observed among treatment groups at ATA or dDTA in either loading direction. In contrast, at the ABA location, the NF-SG group exhibited significantly higher compliance compared to CS-SG in the longitudinal direction (1.34 vs. 1.27, p = 0.008) and compared to both CS-SG (1.20 vs. 1.13, p = 0.016) and control (1.20 vs. 1.11, p = 0.003) in the circumferential direction. Among stented segments, compliance was similar across groups at pDTA-st and mDTA-st; however, at dDTA-st, the CS-SG group showed significantly lower longitudinal compliance than control (1.27 vs. 1.34, p = 0.023). At 125 kPa, no significant differences in compliance were detected across treatment groups at ATA, dDTA, or stented locations. At ABA, NF-SG exhibited significantly higher circumferential compliance compared to control (1.24 vs. 1.16, p = 0.016).

Changes in the tangent modulus are shown only for the high-range and circumferential direction as a representative example in Figure 5A. In the low-range longitudinal direction, no significant differences were observed across treatment groups at non-stented or stented locations, except at dDTA-st, where the CS-SG group exhibited significantly higher stiffness than control (0.47 MPa vs. 0.37 MPa, p = 0.049). No significant differences in low-range circumferential stiffness were detected at any location. For the high-range tangent modulus, no significant differences were observed in the longitudinal direction across treatment groups. In contrast, significant differences were detected in the circumferential direction at stented locations. At dDTA-st, the CS-SG group showed significantly lower stiffness than control (0.92 MPa vs. 2.13 MPa, p = 0.024). At mDTA-st, the NF-SG group exhibited significantly higher stiffness compared to both CS-SG (1.76 MPa vs. 0.90 MPa, p = 0.003) and control (1.76 MPa vs. 1.10 MPa, p = 0.022).

The non-linearity index, defined as the ratio of high-range to low-range tangent modulus, is shown in Figure 5C-D. No significant differences were observed among treatment groups across non-stented locations in either direction. At stented locations, significant differences were detected only at dDTA-st, where the CS-SG group exhibited a significantly lower non-linearity index compared to control in both the longitudinal (1.64 vs. 2.23, p = 0.048) and circumferential (1.54 vs. 2.68, p = 0.004) directions. No other significant group differences were identified.

Changes in anisotropy indices are summarized in Figure 5B. No significant differences were observed among treatment groups at non-stented locations. In contrast, significant differences were detected in stented regions. At mDTA-st, the NF-SG group exhibited a significantly higher anisotropy index compared to the CS-SG group (0.037 vs. 0.007, p = 0.010). At pDTA-st and dDTA-st, the CS-SG group showed significantly lower anisotropy compared to control (pDTA-st: 0.014 vs. 0.028, p = 0.041; dDTA-st: 0.022 vs. 0.056, p = 0.010).

3.3. Histological analysis

Histological analysis revealed distinct differences in medial structure and composition between stent-grafted groups and controls, with changes predominantly observed in the inner medial regions adjacent to the stent-graft. Representative specimens from the mDTA region of the control group and the mDTA-st region of the stent-grafted groups, stained with Movat’s Pentachrome and MTC, are shown in Figures 6 and 7, respectively.

Figure 6.

Representative Movat’s Pentachrome-stained cross-sections of porcine descending thoracic aorta at mDTA. Panels A–C show the control group; D–F show CS-SG; G–I show NF-SG. Panels A, D, and G highlight the inner media near the intima (M3 and M4); panels C, F, and I highlight the outer media near the adventitia (M1 and M2); panels B, E, and H display the full-thickness aortic wall with magnified insets. In the control group, elastin fibers (black) were uniformly wavy and continuous across the wall. In CS-SG, elastin fibers in the inner media appeared more compact, thinner, and less undulated, with evidence of fragmentation and medial compression (red arrows). In NF-SG, similar but milder alterations were observed, with preserved fiber continuity and fewer compression artifacts. Neointimal hyperplasia was evident only in CS-SG sections, characterized by dense ECM accumulation and the presence of vasa vasorum.

Figure 7.

Representative Masson’s Trichrome (MTC)-stained cross-sections of porcine descending thoracic aorta at mDTA. Panels A–C show the control group; D–F show CS-SG; G–I show NF-SG. Panels A, D, and G display the inner media (M3 and M4); panels C, F, and I show the outer media (M1 and M2); panels B, E, and H present full-thickness views. Collagen (blue) was uniformly distributed in controls, with a balanced presence of smooth muscle cells (SMCs, red). In CS-SG, collagen fibers were denser and more prominent in the inner media, particularly adjacent to the stent-graft, while SMC content was reduced. NF-SG samples also exhibited increased collagen in the inner media, but the distribution was lighter and SMC density was relatively preserved. Red arrows highlight regions of stent-induced compression and potential remodeling.

The most pronounced structural alterations in the aortic wall were observed in regions directly exposed to the stent-graft, namely mDTA-st, pDTA-st, and dDTA-st. In the control group, elastin fibers appeared continuous, uniformly distributed, and exhibited consistent thickness and waviness throughout the media. In contrast, the stent-grafted groups, particularly CS-SG, exhibited marked disruption of elastin architecture. These changes were especially evident in the inner media of the stented regions, where elastin fibers appeared denser, thinner, less wavy, and sometimes fragmented. In the outer media, the elastin maintained more regular undulation and spacing. Collagen content was also substantially elevated in the inner media of stented segments. In CS-SG, dense and dark collagen bundles suggested more mature matrix remodeling, while in NF-SG, collagen was also elevated but appeared less dense and more diffuse, suggesting a less advanced or compact remodeling process. Similarly, SMC density was reduced in the inner media of stented areas compared to both the outer media and the control group, with the greatest depletion observed in CS-SG. NF-SG, in contrast, preserved a relatively higher SMC presence in these regions.

In the stent-contact areas, mechanical compression from the metal skeleton led to visible indentations in the luminal surface, often with sinusoidal deformation. In these zones, elastin, collagen, and SMCs appeared densely packed, potentially as a response to mechanical strain. Notably, neointimal hyperplasia was evident near the stent interface in CS-SG, particularly in MTC-stained sections, with abundant SMCs and collagen fibers, suggesting active neointimal formation, possibly driven by SMC migration from the media. In Movat-stained sections, these neointimal areas also displayed pale bluish or greenish ECM components and scattered microvessels, indicating ongoing remodeling involving proteoglycans and vasa vasorum.

These structural changes were predominantly observed in regions directly contacting the stent-graft, while adjacent non-stented or remote segments (e.g., pDTA, dDTA, ATA, ABA) exhibited either minimal or distinctly different remodeling patterns, underscoring the localized impact of stent-induced mechanical forces.

The visual observations were further supported and confirmed by quantitative analysis of component densities, enabling more precise comparisons across treatment groups and anatomical locations. Figures 8 and 9 present the results for elastin, collagen, and SMCs densities measured at each anatomical location across the media thickness, from the adventitia toward the intima (M1 to M4). These results are organized by treatment groups, allowing for comparisons both between groups and within different medial segments.

Figure 8.

Quantitative distribution of wall components in non-stented aortic segments. Densities of elastin (A, E, I, M), collagen (B, F, J, N), and SMCs (C, G, K, O) are shown across four medial depths (M1–M4) from adventitia to intima at ATA, pDTA, dDTA, and ABA. Summary box plots (D, H, L, P) present overall medial densities per location. Differences between treatment groups (Control, CS-SG, NF-SG) are observed primarily in the inner media (M3–M4), particularly for collagen and SMC content. Data are shown as mean ± SD. Statistical comparisons were performed using mixed ANOVA with treatment group as the between-subject factor and medial subsegment (M1–M4) as the within-subject factor, with post-hoc pairwise comparisons where applicable.

Figure 9.

Quantitative distribution of wall components in stented aortic segments. Densities of elastin (A, E, I), collagen (B, F, J), and SMCs (C, G, K) across media depth (M1–M4) are shown for pDTA-st, mDTA-st, and dDTA-st zones. Panels D, H, and L show aggregate densities of each wall component across the full media thickness. CS-SG shows decreased SMC density and increased collagen content, especially in the inner media, while NF-SG exhibits milder remodeling responses. Data are shown as mean ± SD. Statistical comparisons were performed using mixed ANOVA with treatment group as the between-subject factor and medial subsegment (M1–M4) as the within-subject factor, with post-hoc pairwise comparisons where applicable.

Elastin density did not differ significantly among treatment groups at non-stented ATA, pDTA, or ABA. At dDTA, the NF-SG group exhibited significantly higher elastin density compared to both CS-SG (30.5% vs. 26.8%, p = 0.050) and control (30.5% vs. 26.4%, p = 0.034) across the medial thickness (Figure 8I).In stented regions, significant differences in elastin density were observed. At pDTA-st, elastin density was significantly higher in the CS-SG group within the inner media (M3) compared to both control (48.1% vs. 38.2%, p = 0.014) and NF-SG (48.1% vs. 37.3%, p = 0.011) (Figure 9A). Similar increases in inner medial elastin density for CS-SG were observed at mDTA-st, particularly in M4 compared to control (39.9% vs. 31.9%, p = 0.020) (Figure 9E). At dDTA-st, both stent-grafted groups showed significantly higher elastin density across all medial subsections compared to control (NF-SG: 34.2% vs. 26.4%, p = 0.002; CS-SG: 32.3% vs. 26.4%, p = 0.007) (Figure 9I).

Collagen density did not differ significantly among treatment groups at non-stented locations (ATA, pDTA, dDTA, and ABA). In contrast, significant changes in collagen distribution were observed at stented segments, primarily in the CS-SG group. At pDTA-st, CS-SG exhibited significantly increased collagen density in the inner media compared to control (M3: 25.0% vs. 15.3%, p = 0.003; M4: 23.1% vs. 14.8%, p = 0.050) (Figure 9B). At mDTA-st, collagen density in CS-SG was significantly higher than control in the inner media (M4: 29.7% vs. 16.0%, p = 0.004) and higher than NF-SG (M4: 29.7% vs. 19.0%, p = 0.035) (Figure 9F). At dDTA-st, CS-SG showed significantly reduced collagen density in the outer media compared to control (M1: 12.2% vs. 19.0%, p = 0.009; M2: 10.5% vs. 14.6%, p = 0.025) (Figure 9J).

SMC density did not differ significantly among treatment groups at non-stented ATA, dDTA, or ABA. At non-stented pDTA, the CS-SG group exhibited significantly higher SMC density in the outer media compared to both control (M1: 28.6% vs. 19.7%, p < 0.001; M2: 30.1% vs. 23.6%, p = 0.042) and NF-SG (M1: 28.6% vs. 20.6%, p = 0.001) (Figure 8G). In stented regions, significant reductions in SMC density were observed in the CS-SG group. At pDTA-st, SMC density in the inner media was significantly lower in CS-SG compared to control (M3: 13.4% vs. 24.8%, p = 0.004; M4: 9.6% vs. 20.0%, p = 0.002) (Figure 9C). At dDTA-st, CS-SG also exhibited significantly reduced inner medial SMC density relative to control (M3: 19.3% vs. 27.3%, p = 0.004; M4: 9.7% vs. 19.8%, p = 0.002) (Figure 9K). At mDTA-st, the NF-SG group showed significantly higher SMC density in the outer media compared to control (M1: 28.6% vs. 17.9%, p = 0.024) (Figure 9G).

Figure 10 compares paired non-stented and stented segments from the same aorta at the proximal (pDTA vs. pDTA-st) and distal (dDTA vs. dDTA-st) landing zones in the CS-SG group. Quantitative analysis demonstrated significant structural differences between adjacent segments. At the proximal landing zone, stented segments exhibited significantly higher elastin density in the inner media (M3: 48.1% vs. 36.1%, p = 0.005), lower collagen density in the outer media (M1: 11.6% vs. 16.0%, p = 0.045), higher collagen density in the inner media (M3: 25.0% vs. 19.2%, p = 0.033), and lower SMC density across the media (p = 0.004). At the distal landing zone, stented segments showed significantly higher elastin density in M3 (37.3% vs. 26.3%, p = 0.012), lower collagen density in M1 (12.2% vs. 18.9%, p = 0.008), and reduced SMC density in the inner media (M3–M4, p ≤ 0.004). Media thickness was significantly reduced at the proximal landing zone (1.54 mm vs. 1.34 mm, p = 0.043), whereas no significant thickness change was observed distally.

Figure 10.

Localized structural remodeling at stent-graft landing zones in CS-SG group. Representative Movat (A–B) and MTC (C–D) stained images illustrate medial wall architecture at proximal (pDTA vs. pDTA-st) and distal (dDTA vs. dDTA-st) landing zones. Images focus on the inner media in paired non-stented and stented regions of the same aorta. Quantitative analysis of elastin (E, I), collagen (F, J), and SMC (G, K) densities across medial depth (M1–M4) reveals significant remodeling in stented regions. Media thickness comparisons (H, L) indicate a significant 13% reduction in pDTA-st versus pDTA (p = 0.043), while no significant difference was observed in dDTA-st versus dDTA (p = 0.094). Data are shown as mean ± SD. Statistical comparisons were performed using mixed ANOVA with treatment group as the between-subject factor and medial subsegment (M1–M4) as the within-subject factor, with post-hoc pairwise comparisons where applicable.

4. DISCUSSION

The growing use of EVAR and TEVAR in younger patient populations underscores the importance of understanding their long-term impact on aortic physiology. The full scope of stent-induced aortic remodeling remains incompletely understood. While many in vivo studies have focused on hemodynamic parameters to assess changes following stent-grafting[23-25], they are limited in their ability to capture intrinsic structural and mechanical changes in the aortic wall. These metrics are critical because hemodynamics are driven by wall mechanics, which depend on vessel structure and composition. To date, a few studies have examined these tissue-level changes using ex vivo or in vitro models[26]. However, most have been limited by short implantation durations, a narrow focus on either structural[27-30] or mechanical properties[26,31], or reliance on in vitro mock-loop setups that lack physiological fidelity[32,33]. In contrast, the present study offers a more comprehensive approach. It builds upon a previously conducted in vivo porcine study with a four-month implantation period and integrates morphometrical, mechanical, and histological analyses. Importantly, it not only evaluates the chronic effects of stent-grafting on aortic properties but also investigates whether a compliant stent-graft can attenuate adverse remodeling.

In our study, the aorta was divided into three zones relative to the stent-graft: proximal (ATA, pDTA), stented (pDTA-st, mDTA-st, dDTA-st), and distal (dDTA, ABA). In the proximal non-stented segments, no significant changes were observed in morphometrical, mechanical, or histological parameters following stent-graft implantation. The only exception was a localized increase in SMCs density in the outer media of the pDTA. Although the absence of mechanical data in some proximal samples may have limited statistical interpretation, the histological analysis, where data were complete, also showed no notable alterations in ECM composition or wall architecture. These findings are consistent with previous studies, which often report minimal remodeling at the proximal edges of stent-grafts characterized by preserved SMC content, limited intimal thickening, and reduced elastin fragmentation[29,30]. Mechanical investigations have even shown increased distensibility in proximal regions, likely reflecting altered wave propagation dynamics due to compliance mismatch[23,25]. Although this higher distensibility may initially appear favorable, it could instead represent a reactive response to elevated pulse pressure and reflected waves from the stiff stented segment, rather than truly preserved compliance. In this context, the proximal aorta may expand due to altered hemodynamic loading rather than a change in its mechanical properties. Nevertheless, the absence of ECM breakdown or abnormal wall thickening suggests adaptive remodeling. This ability to accommodate stress while preserving structural integrity reflects the resilience of the proximal aorta in juvenile models with short implantation, explaining the minimal remodeling observed in our proximal non-stented segments and supporting its inherent resistance to early pathological adaptation.

In the stented regions, substantial remodeling was observed, particularly in the CS-SG group, compared to the control group. Histologically, the inner media exhibited extensive alterations, with a significant accumulation of collagen fibers adjacent to the stent-graft, in some cases nearly doubling compared to the control. This collagen-rich inner media was accompanied by a reduction in collagen density in the outer media and a decrease in SMC content throughout the wall, especially in the inner layers. The metal stent rings left clear impressions on the intimal surface, creating localized compressed zones with sinusoidal deformations that served as focal sites for collagen deposition. The localized histological changes near the grooves suggest that stent design and material properties can create substantial local stress concentrations and stiffness gradients, even when segmental mechanics appear similar. Additionally, these contact regions developed a thick neointima, rich in collagen, SMCs, and extracellular proteins. In our study, the CS-SG and NF-SG had different stent architectures, and integration of the more compliant NF material likely reduced stiffness mismatch and strain amplification at stent–tissue interfaces, thereby attenuating microstructural damage compared with the stiffer CS-SG design. These findings are consistent with previous studies across species (porcine[27-29], canine[30], and human[32,34]) and vascular beds (thoracic[27,29,32], abdominal[28,30], and peripheral arteries[34]), all of which report intense local remodeling characterized by collagen deposition, SMC loss, neointimal formation, and inflammatory cell infiltration at stent-wall interfaces.

The apparent increase in matrix density in regions directly beneath stent struts may, at least in part, reflect mechanical compaction. Accordingly, these findings were interpreted in conjunction with wall thickness, neointima formation, and inner-outer media differences to help distinguish true remodeling from compression-related effects. These intense remodeling changes can be explained by the altered mechanical environment created by stent-graft implantation. The deployment of a stiff, oversized stent dramatically changes how loads are distributed within the vessel wall[27]. The entire stented segment is exposed to elevated mechanical stresses as the pulsatile aortic pressure acts against the rigid graft. These stresses exceed the physiological range normally experienced by the wall and are sensed at the cellular level as supraphysiological stimuli, triggering structural and cellular responses that drive remodeling[35]. Collagen fibers are the primary load-bearing components of the aortic wall, normally maintaining a balance between synthesis and degradation. Under excessive loads, similar to hypertension, wall cells upregulate collagen synthesis to reinforce the tissue, causing inner media accumulation and contributing to stiffening as a protective mechanism[36]. At the same time, the stent struts create focal points of stress concentration where they contact the wall. These high-stress sites often cause microinjuries that initiate a cascade of cellular events leading to neointima formation[29,32]. In these regions, SMCs migrate from the media into the intima and undergo a phenotypic switch from their normal contractile state to a synthetic phenotype[32,37]. This transition promotes their proliferation, migration, and production of extracellular matrix components, particularly collagen, contributing to the observed neointimal thickening[27,28].

These remodeling processes are also accompanied by significant alterations in elastin. As the media undergoes overstretching and collagen accumulates in the inner layers, elastin fibers lose their characteristic waviness due to sustained high strain[29]. Compression of the media reduces interlamellar spacing, causing elastin fibers to appear thinner and more compact[27]. Additionally, the migration of SMCs and the dominance of collagen synthesis in the remodeled regions are associated with reduced elastin production and increased fragmentation of existing fibers. Together, these changes weaken the elastin network[30], diminishing its ability to store energy and contributing to the shift toward a stiffer, collagen-dominated mechanical behavior[27]. Interestingly, our analysis showed a higher elastin density in the inner media of stented regions compared to controls in both NF-SG and CS-SG groups. This apparent increase, however, likely reflects media thinning and fiber compaction rather than true elastin preservation. The observed loss of waviness, thinning, and fragmentation confirms that elastin integrity was compromised despite the apparent density increase.

These structural changes are expected to reduce aortic elasticity, but their impact on measurable mechanics likely depends on the degree of overstretching and duration of exposure to the stiff stent[27]. Structural remodeling may precede detectable changes in wall mechanics, as the transformation from microstructural damage to macroscopic stiffening often requires time[26]. This may explain why pDTA-st and mDTA-st did not show statistically significant differences in stiffness or non-linearity compared to controls, despite a clear trend toward stiffening in their stress-stretch responses. In contrast, at dDTA-st, the CS-SG group exhibited markedly increased stiffness, particularly in the longitudinal direction at lower stress levels, and displayed more linear mechanical behavior in both directions. This pattern reflects the combined effects of elastin degradation and collagen accumulation on wall mechanics at low strain. Loss of elastin integrity and increased collagen content make the tissue stiffer and more linear as it shifts earlier to a collagen-dominated response[38]. Previous studies also reported similar inconsistencies, where some found no changes in mechanical properties[26] while others observed clear stiffening[32] particularly in the longitudinal direction[31], likely reflecting variations in injury severity, the timing of assessment, and the specific segments analyzed. In contrast to CS-SG, the NF-SG group exhibited no severe medial remodeling in the stented segments, with structural and mechanical characteristics largely resembling those of the control group across most parameters. In areas such as collagen deposition in the inner media or SMC loss, only minimal, non-significant changes were observed. Notably, SMC density in the outer media increased in the NF-SG stented regions, suggesting limited or no migration of SMCs from the media to the intima. This was further supported by the minimal presence of neointima in the NF-SG group.

In the distal segments of the stent-graft, subtle changes in aorta behavior were observed, and these alterations were more pronounced than those seen proximally. The only notable compositional difference was a ~15% increase in elastin density in the NF-SG group compared to both CS-SG and controls. This apparent elastin increase did not translate into statistically significant changes in mechanical properties at the dDTA, although stress–stretch curves suggested a trend toward greater compliance. The NF-SG group also demonstrated an approximately 14% larger outer radius at the dDTA; however, no aneurysm formation or focal pathological dilations were observed, and this enlargement likely reflects the slightly more proximal sampling location in the NF-SG group, due to the shorter stent-graft length, as well as inter-animal variability, rather than adverse distal compliance-mismatch. At the ABA location, however, the NF-SG group demonstrated significantly higher compliance at both low and high stress levels, indicating improved mechanical behavior in the distal segments. In contrast, the CS-SG group displayed a different response. While no significant stiffening was detected in the ABA segment, trends toward increased longitudinal stiffness, reduced circumferential compliance, and higher anisotropy were observed. Moreover, both dDTA and ABA segments in this group exhibited a significantly lower stress-free to load-free thickness ratio, reflecting reduced wall recoil when residual stresses were released. This finding suggests elevated residual compressive stress and decreased wall elasticity in the distal segments of CS-SG[39]. Previous studies have similarly reported more adverse remodeling in distal aortic regions compared to proximal ones. Reduced distensibility and diameter at distal sites have been reported after implantation of stiff stent-grafts[23,25]. Animal models have also shown greater neointimal formation at the distal ends of stiff stents[40,41]. Together, these findings suggest that the distal aorta is more susceptible to early mechanical changes and may have a predominant role in long-term adaptation.

The predominance of remodeling in distal regions can be explained by both mechanical and hemodynamic factors. From a mechanical standpoint, stiff stent-grafts such as CS-SG generate a greater compliance mismatch with the native aorta. Although stent-grafts are tapered, the naturally smaller diameter and higher stiffness of the distal aorta can exacerbate the overdilation, resulting in higher chronic outward force, over-constraint, and elevated residual stresses that mechanically lock the wall and promote maladaptive remodeling[42,43]. Consistent with this mechanism, the reduced thickness ratio (stress-free to load-free thickness ratio) observed distally in the CS-SG group indicates greater wall thinning upon residual stress release, reflecting altered residual stress distribution associated with distal remodeling. The high stress concentrations also increase the likelihood of micro-injuries, which can accelerate neointimal formation. Hemodynamics further amplifies this vulnerability. The distal segments lie near major branches and downstream reflection sites, where flow is complex and impedance is elevated. Stent-grafting and overdilation disturb local flow, creating areas of reduced wall shear stress (WSS), as previously reported[44]. Low WSS impairs endothelial function and stimulates SMCs proliferation in both the media and neointima[45]. Together, these mechanical and hemodynamic mechanisms make the distal regions more vulnerable to maladaptive remodeling and long-term stiffening after stent-grafting. These findings suggest that, similar to stented segments, the distal aorta, which is critical for regulating downstream hemodynamics, may benefit from a more compliant stent-graft such as NF-SG.

In this study, pronounced structural remodeling was evident, particularly in the CS-SG group, yet biaxial mechanical properties showed only modest differences between groups. This likely reflects the relatively short four-month duration. Remodeling after stent-graft implantation begins locally within the inner media and at stent–wall contact regions, where collagen accumulation, elastin degradation, and neointimal formation can precede detectable changes in bulk mechanics. Because these early alterations are highly localized and biaxial testing measures the average response of the full wall thickness, their impact on global stiffness is initially diluted. Consistent with this, the distal aorta exhibited more observable and consistent changes than the proximal segments. The distal region may be more susceptible to early remodeling due to its smaller diameter, higher baseline stiffness, greater compliance mismatch with the graft, and more complex hemodynamics, whereas the more elastin-rich proximal aorta may resist early stiffening. The relatively young age of the animals may further slow adverse remodeling compared with older or hypertensive models. Together, these factors explain why clear structural remodeling was present at four months, whereas larger changes in global mechanical behavior are likely to emerge only at later time points.

Study Limitations and Future Directions.

This study provides important insights into aortic remodeling after stent-grafting with different degrees of compliance-matching, but limitations should be considered when interpreting the findings. First, the follow-up period of four months offered a valuable window into early remodeling, but longer observation periods would be necessary to fully characterize the long-term mechanical adaptation of the aortic wall. Second, while the analysis of proximal segments provided key information, additional data from all proximal sites could further enhance understanding of their structural and mechanical response over time. Third, although stiffness was comprehensively assessed, incorporating wall strength measurements in future work could offer complementary information on rupture risk and mechanical vulnerability. Additionally, the relatively small cohort size, although consistent with prior large-animal TEVAR studies and sufficient to detect meaningful differences, may limit statistical power and introduce some influence of inter-animal variability. Finally, the swine model used here, based on young, healthy animals, provided a controlled setting to isolate the effects of stent compliance, but it does not replicate the complex conditions of diseased or aged human aortas, where structural and mechanical properties vary significantly along the vessel length[38,46]. Future studies building on these findings in more clinically relevant models and over extended follow-up periods would further clarify the long-term implications of stent-graft compliance.

5. CONCLUSION

This study provides a comprehensive evaluation of aortic remodeling following stent-grafting in a porcine model, integrating morphometrical, mechanical, and histological analyses. The results demonstrate that remodeling is highly localized to the regions with stent-graft contact, where the conventional stiff stent-graft (CS-SG) induced marked pathological alterations, including collagen accumulation, elastin compaction, smooth muscle cell depletion, and neointimal formation. These changes were accompanied by increased stiffness and loss of non-linear mechanical behavior, particularly in the distal stented segments. Adjacent non-stented regions showed minimal or adaptive responses, while distal non-stented sites exhibited signs of altered residual stress and reduced wall elasticity in the CS-SG group. In contrast, the compliant nanofibrillar stent-graft (NF-SG) was associated with fewer localized structural and mechanical changes at stent–tissue interfaces, including preserved smooth muscle cell density, reduced collagen accumulation, and limited neointimal formation at select regions. These findings highlight the role of compliance matching in modulating localized aortic remodeling following stent-grafting and suggest that more compliant devices may better preserve native biomechanics and reduce the risk of long-term complications.

statement of significance.

This study quantifies morphological, mechanical, and structural changes in the porcine aorta four months after implantation of either a stiff or a compliant stent-graft. Aortic remodeling was highly localized, occurring primarily in the stented and distal segments, with minimal changes in regions remote from the stent-graft. Our results demonstrate that compliant stent-grafts better preserve aortic structure, smooth muscle cell density, and mechanical function compared to stiff devices, which induce pronounced pathological changes in areas directly exposed to stent contact. These findings underscore the clinical importance of designing more compliant stent-grafts to mitigate adverse vascular remodeling and improve long-term outcomes after endovascular aortic repair.