2. Phases of the Vascular Graft Response
If patency loss occurs within the first days or weeks, it is usually induced by thrombosis, that might be caused mainly by the biomaterial properties (i.e., thrombogenic protein adsorption due to negatively charged surfaces, hemolytic surfaces, and immunogenic surfaces)
[1] or due to an infection caused by intraoperative bacterial contamination or the spread of an infection from a nearby area
[2][10].
If the failure occurs anytime between the first months and the first two years after implantation, it has been shown to be mostly related to a compensation mechanism. In the first case, it might lead to hyperplasia intima, in which the smooth muscle cells and myofibroblasts overproliferate, or due to aneurysm generation, in which the mechanical properties of the vascular wall are inadequate to support the blood flow pressures
[11][12]. At a longer term, atherosclerosis or vascular wall calcification might occur, in which a chronic inflammatory process will modify the local calcium metabolism, stenosing and hardening the vascular wall
[13][14].
Figure 2 shows the possible failure causes in TEVGs. In most of the cases, the ultimate result is the alteration of the vessel lumen, compromising the blood flow, which can also be life-threatening. Even if there are interventions such as catheter-assisted thrombolysis, the placement of a stent, or the surgical revision of the graft to recover the blood flow, VG tend to fail over time
[1][2].
Figure 2. Commonly reported causes for TEVGs’ failure. (
A) An early failure has been reported to be associated with the inherent thrombogenicity of the biomaterial due to protein adsorption and platelet activation
[15]. (
B) An early failure can also be associated with infection, in which a formed biofilm impairs treatment and induces inflammation
[16]. Due to the regulatory mechanisms of endothelium, Mid-Late failure can be mostly related to an incomplete endothelialization. (
C) an incomplete endothelium and a mismatch in the mechanical properties with a flow disturbance will induce the activation of a synthetic phenotype in SMCs, that will overproliferate and decrease the vessel lumen
[16]. (
D) The lack in the mechanical properties of the TEVG required for specific hemodynamic conditions or a faster biodegradability rate than the vascular wall regeneration is expected to cause aneurysms
[17][18]. (
E) A slower biodegradability rate than the vascular wall regeneration is likely to cause a higher influx of pro-inflammatory cells, that might lead to a dysregulated balance between the synthesis and the degradation of the novo extracellular matrix, and if there is a chronic inflammatory stimulus, atherogenesis can occur with calcium and phosphate deposition in the disorganized tissue
[19]. While thrombogenesis and hyperplasia intima have been reported mainly on small-diameter TEVGs, and bacterial infection and aneurysms are typical in larger TEVGs, atherosclerosis has been shown in both types of TEVGs
[16]. (Created with BioRender.com, accessed on 7 November 2021).
Here, we studied the failure causes considering the in vivo studies (in different animal models) reported by Skovrind et al.
[18] and including some more recent studies in rabbits, pigs, and dogs. Then, we classified them according to the moment at which graft fails after implantation and the pathophysiology that leads to patency loss in the graft in light of the TEVG triad and the impact of such reduction in perfusion. Considering the general stages of integration of a vascular graft and the expected remodeling process, we have divided the integration process of a TEVG in three phases, defined arbitrarily by our research group: Phase (I): Peri-implantation period, in which success is defined by the capacity of the graft to overcome the surgery. Every graft that overcomes this stage is expected to begin a reendothelialization process. Phase (II) Period after the implantation and the inflammatory response, in which success is defined by the absence of indications of chronic inflammation and the reendothelialization of the graft. If the graft overcomes this phase and forms a proendothelial lining, it will expectedly begin remodeling the vascular wall. Phase (III) Vascular wall remodeling and device maturation, in which success is defined by the graft acting as an adequate substrate for repopulating and modulating smooth muscle cells, the endothelial lining, and the external fibrous layer in a controlled manner. Then, we described possible failure causes that could occur within the time periods limited by the phases defined previously.
During the first 24 h in phase I, a foreign body response and inflammatory and hemostatic processes dominate the interactions between the TEVG, the cells, and the blood components at the implantation site. Phase II lies between the first weeks to the first months, and in this case, changes and the adaptation to the hemodynamic behavior at the implantation site determine the TEVG integration of the adjacent tissue. Lastly, phase III occurs when the VG starts to function for its intended application and remodels tissue in response to the governing hemodynamics and cell turnover.
Following the Skovrind et al.
[18] study, we analyzed five types of models (caprine, swine, dog, rabbit, and rat), which were contained in a total of 119 studies. (
Table 1).
Table 1. Vascular graft features of the analyzed pre-clinical animal models.
|
Caprine |
Swine |
Dog |
Rabbit |
Rat |
Av. Diameter (mm) |
5 |
4.3 |
4.4 |
2.8 |
1.7 |
Av. Length (mm) |
47 |
50 |
45 |
21 |
8710 |
Subjects |
2 to 20 |
3 to 10 |
1 to 20 |
3 to 18 |
4 to 24 |
Following time |
1 to 9 months |
7 days to 5 months |
1 to 12 months |
1 month to 3 years |
1 to 12 months |
Number of studies |
19 |
18 |
36 |
23 |
23 |
Table 2 shows the preclinical animal studies used for the survival curve of VG and the identifications of the phases in the regeneration of TEVGs as well as their failure causes. The data regarding controls and commercial VGs was also used. PBMC (peripheral blood mononuclear cells), EC (endothelial cells), SMCs (Smooth muscle cells), BMMNCs (bone marrow mononucleated Cells), MVEC (mammary vascular endothelial cells), HIAECs (primary human iliac artery endothelial cells), HOB (human osteoblasts), PLA (polylacticacid), PCL (polycaprolactone), PGA (polyglycolicacid), PU (polyurethane), PVA (polyvinyl acetate), PS (polystyrene). The asterisk next to the year (*) in the author column indicates the additional studies included apart from the ones reported by Skovrind et al
[18].
Table 2. Summary of pre-clinical animal models analyzed for this study.
|
Author, Year |
Ref. |
Cells Seeded |
Luminal Cell Type |
External Cell Type |
Length (mm) |
Diameter (mm) |
Scaffold Type |
Localization |
Sheep |
Tillman, 2012 |
[20] |
Yes |
PB-EC |
None |
60 |
5.0 |
Decellularized Artery |
Carotid Artery AV Shunt |
Kaushal, 2001 |
[21] |
Yes |
PB-EC |
None |
45 |
4.0 |
Decellularized Artery |
Carotid Artery |
Cho, 2005 |
[22] |
Yes |
BMMNC-EC |
BMNC-SMC |
40 |
3.0 |
Decellularized Artery |
Carotid Artery |
Koenneker, 2010 |
[23] |
Yes |
PB-EC |
None |
75 |
5.0 |
Decellularized Artery |
Carotid Artery AV Shunt |
Neff, 2011 |
[24] |
Yes |
PB-EC |
Vessel SMC |
60 |
5.0 |
Decellularized Artery |
Carotid Artery |
Koch, 2010 |
[25] |
Yes |
None |
Vessel SMC |
45 |
5.0 |
Hybrid (Firbin-PLA) |
Carotid Artery |
Ju, 2017 |
[26] |
Yes |
PB-EC |
Vessel SMC |
50 |
4.8 |
Hybrid (Collagen-PCL) |
Carotid Artery |
Fukunishi, 2017 |
[27] |
No |
None |
None |
15 |
12.0 |
Synthetic (PGA-PCL) |
Inferior vena cava |
Row, 2015 |
[28] |
Yes |
EC |
SMC |
45 |
5.0 |
Natural (SIS- Fibrin) |
Carotid Artery |
Aper, 2016 |
[29] |
Yes |
PB-EC |
PB-SMC |
90 |
5.6 |
Natural (Fibrin + Coagulation Factor XIII) |
Carotid Artery |
Swartz, 2005 |
[30] |
Yes |
Vessel-EC |
Vessel-SMC |
12 |
5.0 |
Natural (Firin-Thrombin) |
Yugular Vein |
Meier, 2014 |
[31] |
Yes |
PB-EC |
None |
25 |
4.0 |
Natural (Fibrin gel) |
Femoral Artery |
Ramesh, 2013 |
[32] |
No |
None |
None |
130 |
4.0 |
Decellularized Vein |
Carotid Artery |
Aussel, 2017 |
[33] |
No |
None |
None |
35 |
6.0 |
Natural (Chitosan) |
Carotid Artery |
Weber, 2018 |
[34] |
No |
None |
None |
10 |
4.5 |
Natural (Nanocellulose) |
Carotid Artery |
Goat |
Turner, 2006 |
[35] |
Yes |
Vessel-EC |
None |
45 |
4.5 |
Hybrid (PU + cells) |
Carotid Artery |
Kim, 2007 |
[36] |
No |
None |
None |
40 |
4.5 |
Decellularized Artery |
Carotid Artery |
Rabbit |
Zheng, 2012 * |
[37] |
No |
None |
None |
15 |
2.2 |
Synthetic (PCL) |
Carotid Artery |
Cutiongco, 2016 * |
[38] |
No |
None |
None |
Not reported |
0.9–1 |
Synthetic (PVA) |
Femoral Artery |
Zhang, 2017 * |
[39] |
No |
None |
None |
30 |
4.0 |
Synthetic (ePTFE) |
Abdominal Aorta |
Wang, 2016 * |
[40] |
No |
None |
None |
10 |
2.2 |
Synthetic (PCL) |
Carotid Artery |
Wise, 2011 * |
[41] |
No |
None |
None |
20 |
2.8 |
Hybrid (PCL—Elastin) |
Carotid Artery |
Tillman, 2009 * |
[42] |
No |
None |
None |
40 |
5.0 |
Hybrid (PCL—collagen) |
Aortoiliac Bypass |
Evans, 2015 * |
[43] |
No |
None |
None |
30–40 |
2.0 |
Autologous vein |
Yugular Vein |
Zhu, 2012 * |
[44] |
Yes |
HOB |
None |
5 |
3 |
Synthetic (PCL) |
Femoral Artery |
Ishii, 2008 * |
[45] |
No |
None |
None |
24 |
3.6 |
Hybrid (PU—collagen—Hyaluronic acid) |
Abdominal Aorta |
Kajbafzadeh, 2019 * |
[46] |
No |
None |
None |
Not reported |
Not reported |
Decellularized Artery |
Femoral Arteries |
Zhao, 2019 * |
[47] |
No |
None |
None |
20 |
2 |
Synthetic (PLCL) |
Carotid Artery |
Bai, 2019 * |
[48] |
No |
None |
None |
20 |
2 |
Synthetic (PS) |
Carotid Artery |
Bai, 2018 * |
[48] |
No |
None |
None |
20 |
2 |
Synthetic (PS) |
Carotid Artery |
Pig |
Koens, 2015 |
[49] |
No |
None |
None |
35 |
4 |
Natural (elastin-collagen) |
Iliac Artery |
Rotmans, 2005 |
[50] |
No |
None |
None |
70 |
5 |
Synthetic (PTFE+ CD34 coating) |
Carotid Artery AV Shunt |
Rothuizen, 2016 |
[51] |
No |
None |
None |
40 |
4.2 |
Fibrotic capsule tube + PCL |
Carotid Artery |
Mahara, 2015 |
[52] |
No |
None |
None |
250 |
3 |
Decellularized Artery |
Femoral Artery |
Sánchez-Palencia, 2018 * |
[53] |
No |
None |
None |
17 |
4.5 |
Decellularized SIS |
Carotid Artery |
Dahan, 2017 |
[54] |
No |
None |
None |
45 |
4.0 |
Decellularized Artery |
Carotid Artery |
Valencia Rivero, 2017 |
[55] |
No |
None |
None |
150 |
4.0 |
Decellularized SIS |
Carotid Artery AV Shunt |
Zavan, 2008 |
[56] |
No |
None |
None |
50 |
4.0 |
Natural (Hyaluronan) |
Carotid Artery |
Wippermann, 2009 |
[57] |
No |
None |
None |
10 |
3.4 |
Natural (Cellulose) |
Carotid Artery |
Hinds, 2006 |
[58] |
No |
None |
None |
40 |
4.3 |
Natural (Ellastin) |
Carotid Artery |
Pellegata, 2015 |
[59] |
Yes |
hIAECS |
None |
50 |
6.0 |
Decellularized Artery |
Iliac Artery |
Dog |
Arts, 2002 |
[60] |
Yes |
MVEC |
None |
50 |
4.0 |
Synthetic (ePTFE) |
Carotid Artery |
Xie, 2010 |
[61] |
No |
None |
None |
50 |
6.0 |
Poly (carbonate urethane) (PCU) filaments |
Infra-renal aorta |
Yokota, 2008 |
[62] |
No |
None |
None |
30 |
4.0 |
Collagen microsponge with a biodegradable woven polyglycolic acid (core) and poly-L-lactic acid (sheath) fibers. |
Carotid Artery |
Zhou, 2009 * |
[63] |
No |
None |
None |
45 |
3.0 |
Decellularized/hybrid |
Carotid Artery |
Zhou, 2012 |
[64] |
Yes |
PBMC-EC |
None |
45 |
3.0 |
Decellularized Artery |
Carotid Artery |
Zhou, 2014 * |
[65] |
No |
None |
None |
45 |
3.0 |
Hybrid chitosan/poly(e-caprolactone) (CS/PCL) nanofibers |
Carotid Artery |
Considering that all specimens evaluated for caprine model lost patency within the peri-implantation period, this model was disregarded for further analyses. The results suggest that most of the failure cases are related to the transition from phase II to phase III, with an average 14% of grafts failing in early stages and 35% of them reporting loss of patency or the need for removal before the beginning of the remodeling process. This strongly indicates that most of the failure cases are related to inflammatory responses. Because an important number of studies ended suddenly before failure (mainly due to the impairment of the proper blood flow), it was impossible to determine the remodeling performance of such grafts. In these cases, we classified the graft within the defined phases and considered the explanting conditions in terms of level of vascular wall remodeling, reendothelialization, presence of thrombus, and general integrity. We calculated the percentage of successful VGs for each of the phases to build the survival curve shown in Figure 3.
Figure 3. Survival curve of tissue-engineered vascular grafts in preclinical studies. The survival curve presented starts with all the specimens participating in preclinical studies for each of the species evaluated. When a given number of individuals does not overcome a given phase of the integration process, it is subtracted from the total number of specimens participating, reducing the percentage of animals surviving the study.
Furthermore, the review showed that the proposed critical phases and types of failure are observed in every animal model. However, studies involving smaller animal models are characterized by shorter evaluation times, and therefore they arrive at later phases of integration faster. Additionally, while preseeding cells over grafts before implantation represent an advantage in the initial phases of integration, preventing an instant occlusion and stabilizing the immune response, when the regeneration process reaches the phase of remodeling of the vascular wall, there is no difference on long-term patency between preseeded and non-preseeded grafts. (Figure 4).
Figure 4. (a) One-way ANOVA analysis of different success percentages of preclinical studies in different species: Success is defined by the capacity of the graft to overcome the stabilization of the immune response and start the remodeling process of the vascular wall. From all the specimens participating in pre-clinical trials, the number of individuals participating in the preclinical studies and the percentage of the total of individuals from each species that is successful are compared to the others. (b) 95% confidence interval Tukey test for the number of vascular grafts that reached phase III between animal models. The quantitative data are expressed as the mean ± SD, where * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001 when compared between models.
According to our analyses, for phase, I failure is generally associated with thrombus formation and foreign body reaction. In contrast, for phase II, failure after the peri-implantation period can be attributed to the interactions of the graft with immune cells and the absence of hemodynamic stabilization. Finally, after overcoming this phase, success in phase III appears related to the ability of TEVG for supporting cell migration and the induction of the remodeling processes in the vascular wall.
To explain the critical points that might unchain events related to failure, we gave each of the phases different possible outcomes. These outcomes were established according to the failure causes reported on in vivo studies for vascular grafts and based on our research group experience. The first two or three outcomes aim to explain the metabolic processes that would lead to a non-regenerative response and/or patency loss due to the incompatibility of the TEVG with the native tissue in hemodynamic conditions. On the other hand, the last outcome would describe the ideal conditions to overcome the phase (i.e., the biochemical and mechanical success signals occurring in a biocompatible implanted vascular graft). The mechanistic explanation of the interactions in the TEVG triad that could elucidate success or failure outcomes are described below. For this review, we will focus on the first phase, in which the first inflammatory responses drive the most critical step, determining the beginning of the vascular wall regeneration.