Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jul 1.
Published in final edited form as: J Mech Behav Biomed Mater. 2021 Apr 16;119:104527. doi: 10.1016/j.jmbbm.2021.104527

Evaluation of the Probe Burst Test as a Measure of Strength for a Biologically-engineered Vascular Graft

Zeeshan H Syedain 1, Abrielle Prunty 1, Jirong Li 2, Robert T Tranquillo 1,3
PMCID: PMC8562868  NIHMSID: NIHMS1698285  PMID: 33930654

Abstract

Biologically-engineered vascular grafts have the potential to provide a viable alternative to donor vessels and synthetic grafts. In congenital heart defect patients, the need is even more dire since neither has the capacity to provide somatic growth. To ensure clinically-used grafts perform to accepted standards, mechanical strength is a crucial consideration, with burst testing being considered as one key metric. While ISO 7198 standards for prosthetic vascular grafts provide multiple choices for burst testing, most studies with tissue-engineered grafts have been performed with only pressure burst testing. Here, we compare the performance of a decellularized tube of collagenous matrix grown from dermal fibroblasts, possessing circumferential fiber alignment and anisotropic tensile properties, as determined from pressure and probe burst testing. The two burst tests showed a strong correlation with each other and with tensile strength. Further, relatively weak and strong batches of grafts showed commensurate differences in pressure and probe burst values. Both probe burst and tensile strength measurements in the central and edge regions of the grafts were similar in value, consistent with homogenous collagen content and microstructure throughout the grafts as indicated by histology, in contrast to ovine femoral and carotid arteries similarly tested. Finite element analysis of the probe burst test pre-failure for a homogeneous, isotropic approximation of the matrix constitutive behavior indicated dependence of the (inferred) effective failure stress achievable on probe diameter. The results indicate a probe burst test in a sampled edge region of this biologically-engineered graft provides a representative measure of burst strength of the entire graft.

1. Introduction

Biologically-engineered vascular grafts are currently being evaluated in clinical trials for the treatment of vascular diseases (Gutowski et al., 2020; Lawson et al., 2016; Sugiura et al., 2018). Several approaches have been developed for engineering a vascular graft (Dahl et al., 2011; L’Heureux et al., 2007; Ratcliffe, 2000; Roh et al., 2010; Syedain et al., 2011). Regardless of the approach and final composition of an engineered graft, regulatory approval requires a vascular graft to demonstrate mechanical strength among other measures of safety. For mechanical strength, burst pressure testing as described in International Organization of Standardization (ISO) standard 7198: Cardiovascular Implant and Extracorporeal Prosthesis, is a commonly used. While the standard describes multiple choices to measure burst strength, to date, most tissue-engineered vascular grafts have been evaluated using the pressure burst test where the lumen of the graft is inflated with a pressurized fluid to failure (Dahl et al., 2011; Konig et al., 2009; Stekelenburg et al., 2009; Syedain et al., 2017b; Syedain et al., 2011).

The pressure burst test is a practical and relatively simple method for small diameter (<6mm) vascular grafts. One end of the graft is cannulated and the other end is mounted to a fluid line from which test liquid is injected at a prescribed rate while pressure is recorded. However, it is not as feasible for shorter length grafts with large diameters. For pediatric applications, graft diameters of 10–16mm are desirable with a length requirement of 4–6cm. While this length is sufficient to perform a burst pressure test, it will require the entire graft to perform the pressure burst test in order to mount the graft, reducing ability to perform multiple mechanical and biochemical safety characterizations on the same graft used for implantation. Hence, a test that requires a smaller sample is desirable to measure mechanical strength of individual grafts, specifically, a small segment from the end of the graft. This can be potentially achieved with the probe burst test. In this test, a sample is mounted on a plate over a circular hole, and a probe is traversed into the sample at the center of the hole at a prescribed rate as the force is monitored. The probe burst apparatus has a direct effect on measured values, hence the dimensions and shape of probe are well defined in ISO 7198 with a diameter of 9.5mm and a hemispherical tip and center opening of the sample holder of 11.3mm diameter.

The minimum sample dimensions required for sample mounting to conduct a probe test makes it impractical to use it for small diameter grafts (<6mm) as their circumferential length (after slitting the graft axially to create a sheet) is insufficient to secure the test sample in the mounting. This is likely the reason for a dearth of published data on probe burst testing of tissue-engineered vascular grafts, which are aimed at small diameter applications (coronary and peripheral artery bypass). Recently, Geelhoed et al evaluated the use of probe burst in comparison to pressure burst for silicone rubber grafts of varied properties in comparison to bovine carotid arteries and jugular veins (Geelhoed et al., 2019). While they reported a strong correlation of probe burst and pressure burst with the silicone rubber grafts, the native vessels lacked any correlation. The authors concluded that for a probe burst test to be used as a measure of pressure burst, homogenous graft composition and structure are essential, with validation required for each type of tissue-engineered graft.

Using a sacrificial fibrin-based tissue engineering approach, we have reported pressure burst testing of 4mm and 6mm diameter grafts (Meier et al., 2014; Syedain et al., 2017b; Syedain et al., 2011; Syedain et al., 2014). The culture conditions can strongly influence the final mechanical properties of the engineered grafts. Therefore, we utilized here four sets of grafts with different pressure burst properties to evaluate any correlation between pressure burst and probe burst measurements. We built a custom probe test apparatus with the specifications for probe diameter and testing regiment defined in ISO standards. Additionally, tensile strength was measured to evaluate its correlation with burst strength properties. In addition, we measured strength in both the mid and end regions of the grafts to assess any regional difference given the aspiration of using a probe test from the end of a graft to be implanted as a reliable measure of graft strength. Finite element analysis (FEA) was conducted to elucidate the importance of testing parameters and provide a theoretical prediction of burst strength for these biologically-engineered grafts. This study informs comparison of burst properties for large diameter grafts that are being evaluated as pediatric conduits (Syedain et al., 2016) and arterio-venous grafts (Syedain et al., 2017a) and used to construct tubular heart valves (Reimer et al., 2017; Reimer et al., 2015; Syedain et al., 2015).

2. Materials and Methods

2.1. Biologically-engineered Vascular Grafts:

Grafts were produced as previously reported: Dermal fibroblast-seeded fibrin gel was formed by adding bovine thrombin (Sigma) and calcium chloride in 20 mM HEPES-buffered saline to a suspension of ovine fibroblast in bovine fibrinogen (Sigma). The final component concentrations of the cell suspension were as follows: 4mg/ml fibrinogen, 0.8U/ml thrombin, 5.0mM Ca++, and 1 million cells/ml. Cell suspensions were mixed and injected into a tubular mold containing a glass mandrel pre-treated with a 5% Pluronic F-127 solution. Subsequently, grafts were cultured in DMEM supplemented with 10% fetal bovine serum (FBS, HyClone), 100U/ml penicillin, 100μg/ml streptomycin (Gibco), 2μg/ml insulin (Sigma), 50μg/ml ascorbic acid (Sigma). Grafts were cultured for 7–9 weeks on the mandrel (Syedain et al., 2017a) or removed and mounted in a pulsed flow-stretch bioreactor (Syedain et al., 2011) with medium changed three times per week. The final dimensions of each graft were 6 mm inner diameter, 12–15 cm in length for batch 1 (n=5), 6–8 cm in length for batch 2 (n=2), 6–10 cm in length for batch 3 (n=3), and 6–8 cm for batch 4 (n=3), with all grafts being ~300 μm in thickness. The grafts were rinsed in PBS and incubated on an orbital shaker for 6 hr with 1% sodium dodecyl sulfate (SDS, Sigma) in distilled water to start the decellularization process. The SDS solution was changed three times. The grafts were rinsed in PBS and incubated with 1% Triton X-100 (Sigma) in distilled water for 30 min, extensively washed with PBS for 7 days, and then incubated in deoxyribonuclease enzyme (Worthington Biochemical) in DMEM supplemented with 10% FBS overnight. Decellularized grafts in this study were stored at 4°C in PBS until testing.

2.2. Native Arteries:

Ovine carotid and femoral arteries obtained from 9 adult sheep were received from the University of Minnesota’s Experimental Surgical Services Lab. Samples were prepared by washing with PBS solution and removing the excess tissue with forceps and dissecting scissors to expose the artery. The artery samples were then tested for collaterals branches by clamping one end with a surgical mosquito clamp, inserting a 20mL syringe in the other end, and injecting PBS solution to evaluate the location of any branches or holes in the sample. A 30mm-50mm section with no holes or branches was selected for pressure burst testing. This section was cut and placed in a labeled conical tube filled with PBS. Two 10mm sections also without holes or branches were cut from the remaining sample. These sections were cut longitudinally to create a rectangular slab sample for probe burst testing, placed into labeled conical tubes, and stored at 4°C until testing.

2.3. Pressure Burst Testing:

The test sample is cannulated on both ends with one end capped and the other end connected via a fluid line to a syringe and an in-line pressure transducer (PX613, Omega). Using a custom LabView™ (National Instrument) code with DAQ input, pressure is recorded at 10ms increments during injection of phosphate-buffered saline (PBS) into the graft lumen via syringe. The graft was immersed in PBS during the test. The pressure at failure is recorded as the burst pressure.

2.4. Probe Burst Testing:

A testing apparatus was custom made to be compatible with the Instron 8848 MicroTester mounted with a ±500N load cell (Fig. 1a,b). The probe is inserted into the cylindrical fitting on the load cell. The test apparatus is secured to the Instron Table with screws, lining up the center sample hole with the probe. A tissue sample with dimensions 20mm x 20mm prepared for probe burst pressure testing is then placed on the sandpaper side of the front removable plate over the center opening of the sample holder (Fig. 1c,d). The front plate is then pressed onto the test apparatus and the thumb screws are inserted and tightened at the four corners. A custom test program was created using WaveMatrix software to traverse the probe through the sample at a constant rate until failure. Immediately prior to testing, all samples are kept in PBS solution and warmed to room temperature. For testing, samples removed from PBS are mounted as described above without drying excess liquid on the sample. The total duration of testing for each sample from mounting to bursting was less than 2 min with no drying apparent. The probe diameter, the rate of traverse, and the failure load for each sample are recorded as stated in ISO 7198. Per ISO 7198, a hemispherical tip with diameter of 9.5mm and center opening of 11.3mm diameter were used, with a traverse rate of 50 mm/min. An equivalent burst pressure was estimated by dividing the failure load by the probe surface area and converting between N/mm2 and mmHg.

Figure 1:

Figure 1:

Open in a new tab

Images of custom probe testing apparatus in a. oblique and b. side views. Graft sample c. placed inside probe testing holder, and d. after probe testing showing rupture.

2.5. Tensile Testing:

Tissue strips cut from the middle and edge region of the vascular grafts ~3 mm wide and 18 mm long were tested for tensile properties in both the circumferential and radial directions using compression grips and an Instron system as previously described (Syedain et al., 2019). Briefly, tissue strips were clamped into custom compression grips attached to the actuator arms of an Instron testing system (Instron 8848 Mirco-Tester). The tissue was straightened with a 0.05N load. One arm with a 5N load cell was then pulled at a constant rate of 50 mm/min until failure. The force and displacement were recorded during the test in the Instron WaveMatrix® v. 1.8 software. Similarly, ~3mm wide circumferential rings were cut from ovine artery and cut into strips for tensile testing. No axial testing was performed for native arteries since only circumferential properties were compared. The maximum (membrane) tension was defined as failure force/width of tissue strip and used as a relevant measure of mechanical strength in complex layered tissues like arteries and to allow for direct comparison with graft strength (Billiar and Sacks, 2000). Modulus was defined as the linear slope of the stress-strain curve taken in the linear region during the stretch to failure. Samples from the middle and edge of grafts from the two n>2 batches were averaged for analysis.

2.6. Histology:

Cross-sections of vascular grafts and native arteries were fixed in 4% paraformaldehyde, embedded in OCT (Tissue-Tek), and frozen in liquid N2. Cross sections of 9-μm thickness were stained with Lillie’s trichrome stain. Images were taken at 10x magnification using an Olympus IX70 microscope and Metamorph v. 17.8 software.

2.7. Statistics:

Pressure and probe burst values were evaluated with linear regression and the F-test was used to calculate a P value for null hypothesis that the overall slope is zero in GraphPad Prism. The 95% confidence intervals and R2 value are also reported. Average pressure burst values and tensile properties were compared between batches using the t-test with p<0.05 to conclude significant difference. Probe burst values for middle and edge regions were taken as a ratio for n=6 grafts, and a one sample t-test was performed with a theoretical mean of 1.

2.8. Finite Element Model and Analysis

A finite element model of the probe burst test was created in ANSYS Workbench: Static Structural. To simplify, an “isotropic” stress-strain curve, based on the average of the curves for tensile testing of the biologically-engineered matrix (batch 2) in the aligned and orthogonal directions, was fit to the (hyperelastic) Mooney-Rivlin constitutive equation to generate the required strain energy density function.

The simulated geometry is shown in Suppl. Fig. 1. The tip of the probe is hemispherical and centered on the circular sample testing region of 11.3mm diameter. Two probe sizes were simulated (9.5mm and 3mm diameter) for each of three matrix thickness values (0.35mm, 0.5mm, and 0.65mm). Axisymmetry was assumed.

The probe model and matrix models were discretized by 8-node quadrilateral elements, with the numbers of elements for the different cases shown in Table 1. An augmented Lagrange-based contact algorithm was utilized to prevent penetration between the probe and matrix. A boundary condition of zero displacement of the matrix at the outer edge was applied. Two boundary conditions were separately considered in the probe-matrix contact region, frictionless, and frictional based on Coulomb’s law of friction, where in the contact zone umatrix = uprobe wherever τmatrix,tangential < μτmatrix,normal + b (sticking; identical displacement of matrix and probe) τmatrix,tangential = μτmatrix,normal + b otherwise (slipping, with frictional force τmatrix,tangential + b) Here, μ is the coefficient of friction, with increasing value representing stronger frictional coupling between the contacting surfaces, and b is the cohesion sliding resistance, the frictional stress in the absence of a normal stress; b=0 and μ = 0.1 were used for comparison to the everywhere frictionless case. The simulations were made by impinging the probe axially into the matrix up to 10mm after contact in 0.01mm steps, solving the equation of mechanical force equilibrium at each step.

Table 1.

Number of 8-node quadrilateral elements used in the FE model

Probe Diameter Matrix Thickness
0.35mm 0.5mm 0.65mm
9.5 mm 2314 2626 8342
3 mm 1521 1833 2471
Open in a new tab

3. Results

Harvested ovine arteries (femoral and carotid) and biologically-engineered grafts were cut into segments and tested for burst, probe burst, and tensile strain-to-failure (circumferential) as shown in Fig. 2.

Figure 2:

Figure 2:

Open in a new tab

Experimental design of testing performed for each group of samples, ovine arteries and biologically-engineered grafts.

3.1. Ovine Artery Testing

Measured pressure burst, probe burst, and tensile strength values were plotted against each other. Linear regression showed no correlation or non-zero slope for any of the three comparisons when data for the femoral and carotid arteries were pooled (Fig. 3a–c) although some correlations were revealed for the carotid arteries when segregated (Fig 3d–f).

Figure 3:

Figure 3:

Open in a new tab

Ovine artery testing comparison for a. probe burst vs. pressure burst b. maximum tension vs. pressure burst, and c. maximum tension vs. probe burst. Dashed lines show 95% confidence interval.

3.2. Biologically-engineered Graft Testing

Four batches of 6mm diameter grafts (n=5,3,2,3; 13 total grafts) cultured with different combinations of cell lots/fibrinogen lots/culture duration in order to generate samples of varied collagen content / material properties were evaluated and pooled for regression analysis correlation between pressure and probe burst. Failure in the probe burst test initiated near the center of the sample, not near the peripheral mounting, based on visual assessment. Failure in the pressure burst test occurred as an axial rupture in the middle region of the sample, also based on visual assessment. The grafts exhibited a linear correlation with R2 = 0.89 and a slope different from zero (Fig. 4a). Further, the increases in pressure burst and probe burst between the three batches of grafts with n>2 were of similar magnitude (Fig. 4b). One sample from each of the batches 1–3 and 3 samples from batch 4 was also evaluated for probe burst in both the central and end regions of the graft, with comparable values measured in both locations (Fig. 4c). The ratio of probe burst valve for mid region to edge region was calculated and a one-way t-test was performed against theoretical mean of 1. A resulting P-value of 0.93 indicates no difference from 1 for the ratio of mid/edge probe burst values, that is, no difference between probe burst values in the mid and edge regions of the grafts.

Figure 4:

Figure 4:

Open in a new tab

Biologically-engineered graft testing comparison for a. probe burst vs. pressure burst for grafts from all four batches, b. average pressure and probe burst values for the batches 1,2 and 4. c. values of probe burst in central and edge regions of one graft from each of batches 1–3, and n=3 from batch 4, d. Average tensile strength for central and edge region of grafts for batches 1&3.

Strips cut in the circumferential and axial directions from the grafts of the batches 1–3were also evaluated for tensile properties. They exhibited non-linear stiffening and anisotropy as we have previously observed (Syedain et al., 2011), with the ratio of the tangent modulus in the circumferential and axial directions being 1.96 and 1.45 for the two batches (Suppl. Fig. 2). Thickness for the samples in two batches along with their UTS and failure strain are shown in Table 2. The correlations between both circumferential and axial maximum tension with pressure burst and probe burst also exhibited correlation (Fig 5a–d). For batches 1&3, strips were cut from central and edge regions of the graft and compared. As shown in Fig. 4d, there was no difference in the tensile strength in the central and edge regions among grafts within a batch (919±317 N/m (n=5) vs. 1060±165 N/m (n=5), and 1953±184 N/m (n=3) vs. 1805±287 N/m (n=3)), consistent with the probe burst testing (Fig. 4c).

Table 2.

Strain-to-failure results in the circumferential direction

Batch Thickness (mm) UTS (MPa) Failure Strain
1 0.31±0.14 3.7±2.2 0.69±0.12
2 0.33±0.03 5.9±1.1 0.67±0.05
Open in a new tab

Figure 5:

Figure 5:

Open in a new tab

Biologically-engineered graft correlation of tensile strength in circumferential and radial directions to a,c. probe burst and b,d. pressure burst. Dashed lines show 95% confidence interval. Samples are color coded by batch as in Fig. 4. Only one sample from batch 2 was available.

For the batch of grafts with an average probe burst value of 46±5 N, the burst pressure was 3370±154 mmHg. If the probe burst value (force) is converted to a pressure burst value (pressure) based on the projected probe surface area, the burst value is 4839±553 mmHg. However, if the entire (hemispherical) probe surface area is used, the value is 2419±267 mmHg.

3.3. Histology Comparison between Biologically-engineered Vascular Graft and Native Artery

Trichrome staining shows a homogenous distribution of collagen across the entire thickness of the biologically-engineered grafts with measured thickness of 0.32±0.1 mm (n=13). In comparison, native arteries exhibit a well-known layered structure (Fig. 6a), with variable thickness of muscle fibers in the media (middle layer) and poorly-organized collagen in the adventitia (abluminal layer). The carotid artery thickness ranged 0.51–0.60 mm (0.57±0.04 mm), and the femoral artery thickness ranged 0.35–0.48 mm (0.43±0.05 mm).

Figure 6:

Figure 6:

Open in a new tab

Histological comparison using trichrome stained section of ovine a. carotid and b. femoral arteries with c. biologically-engineered grafts. * denotes lumen side. Scale bar = 200 μm

3.4. Probe Test Simulation

The reaction force and contact area both increased monotonically with impingement distance for both probe diameters and with matrix thickness (Fig. 7). Higher reaction forces and contact areas were generated with the larger probe at every distance, but 2–3 fold higher effective stresses occurred with the smaller probe. Maximum equivalent stress occurred at the outer edge where the no-displacement boundary condition was enforced (Fig.3). Otherwise, it occurred at or near the center of the sample, which is considered relevant (see Discussion), and this is the value plotted in Fig. 7. There was negligible dependence of contact area on matrix thickness, but effective stress increased with matrix thickness at every distance, with a more pronounced dependence for the smaller probe. There was modest difference between the frictional and frictionless values for the friction parameter value used (μ=0.1), the most notable difference being slightly higher effective stress when including friction (~5.5% for both probe sizes when the maximum equivalent strain attains a value of 0.7, the approximate failure strain from Table 2).

Figure 7: Probe test simulation using FEA.

Figure 7:

Figure 7:

Open in a new tab

a. Heat map of equivalent stress across the sample at a probe impingement distance of 5mm for (a-c) 3 mm probe, and (d-f) 9.5 mm probe and matrix thickness of 0.35mm (a,d), 0.50mm (b,e), and 0.65mm (c,f) for the case of frictionless contact. Insets of center and (otherwise excluded) outer edge regions are shown. b. Results for smaller (left) and larger (right) probe sizes are plotted for frictional (solid, μ=0.1) and frictionless (dashed, μ=0) probe contact conditions and 0.35mm (red), 0.5mm (blue), and 0.65mm (green) matrix thickness as a function of the probe impingement distance. Discontinuities in the effective stress curves for the smaller probe reflect the relative significance of fractions of elements defining the smaller contact zone.

4. Discussion

To date, probe burst strength is more commonly reported for synthetic vascular grafts made from knitted synthetic fibers that possess relatively porous structure (Morota and Takamoto, 2013; Xie et al., 2016b; Yang et al., 2015). Due to their porous nature, pressure burst testing would not be feasible for these grafts. Likewise, burst properties using a probe test are also reported for vascular patch materials (Fallon et al., 2012). However, comparison of specific testing protocols reveals consistent use of standard equipment is not applied in these previous studies, even though most published work references ISO7198 as the basis for the probe burst testing.

In ISO7198:2016(E), section ‘A.5.2.6: Probe Burst Strength’, it is specifically stated in A.5.2.6.2 that the traversing probe must have a diameter of 9.5mm (3/8 inch) and a hemispherical radius. While, Fallon et al tested a patch made from extracellular matrix using probe diameter and shape as defined in the ISO7198 (Fallon et al., 2012), others have published measurements using probe diameters ranging from 1.5mm-6mm with no description of the probe having a hemispherical end (Morota and Takamoto, 2013; Xie et al., 2016b; Yang et al., 2015). This thwarts comparison of probe burst values for different vascular grafts and assessment of relative strength. We agree with the recommendation of Geelhoed et al that for a specific vascular graft, both pressure and probe burst must be performed in order to make any correlation or comparison (Geelhoed et al., 2019). Further, as required by the ISO7198 standards, probe diameter, traversing rate, and failure force (unit of Newton) should be reported. Only then can probe burst be compared between different graft types or among batches of a specific graft.

In current testing on a custom probe burst setup, probe values exhibited strong correlation with both pressure burst and tensile properties for the biologically-engineered grafts. More importantly, when average values of three batches of grafts are compared, the percentage increase in average pressure burst was statistically the same as the increase in average probe burst. Grafts (n=6) were evaluated for the probe burst in the central and edge regions of the grafts. No difference between probe burst values was measured. Similarly, no difference occurred in the tensile strength between the central and edge regions of grafts from the two batches evaluated. This is an important characterization, as any inference of graft properties from short test regions requires validation of homogeneity of the graft along its length.

Further analysis showed that the probe burst value does not readily convert to a failure stress as measured by the pressure burst test value (i.e. burst pressure). The converted probe burst values would either under- or over-report the pressure burst value based on the total probe area and projected probe area, respectively. This was also reported by Geelhoed et al for silicone tubes and native arteries (Geelhoed et al., 2019). During the probe burst test, we observed the rupture occurs somewhere between the probe touching the graft surface and the graft completely enveloping the hemispheric surface of the probe, consistent with the pressure burst falling in the range predicted by the converted probe burst values. However, without modeling and experimental validation, the exact probe area for the conversion is unknown. The probe burst values should thus only be stated as an extrinsic force value related to the burst strength. The current testing was done with probe diameter, sample holder diameter and hemispheric surface as defined in the ISO 7198 standards, which makes probe setup specifications part of the probe burst testing and requires standard equipment to ensure measurements of probe burst can be compared from different testing labs.

Not surprisingly, given our material is not well described as a Hookean solid and its failure occurs at large strains, the accuracy is poor for predictions of failure when stress exceeds the UTS based on deformation of an isotropic linear elastic membrane, particularly for the probe burst (Begley and Mackin, 2004). Thus, FEA of the probe test was conducted to illuminate the contact area at failure and dependence of the probe test on three relevant parameters: the probe diameter, the thickness of the matrix, and the degree of probe-matrix friction. Since a pre-failure model based on tensile behavior of the matrix was used, failure can only be inferred when the computed strain exceeds the measured failure strain of the matrix. Consistent with observed failure of the matrix in the central region, the maximum strain/stress, that typically correlates with failure, is predicted to be highest in the central region. This excludes elevated values at the fixed boundary where some slippage of the sample from between the mounting plates during loading was observed to occur, suggesting these elevated values at the boundary are over-predictions. Similar simulation was reported previously but not interpreted in the context of failure (Selvadurai, 2006).

The measured failure strain was ~0.7, with corresponding UTS of ~5 MPa in the aligned (circumferential) direction for samples with thickness of ~0.3 mm tested in strain-to-failure in the circumferential direction. A maximum equivalent strain of 0.7 is predicted at an impingement distance of ~6.5mm for the 9.5mm probe used in this study, corresponding to a contact area of ~110mm2 (compared to the probe surface area of 142mm2), effective stress of ~0.1 MPa (=750 mmHg), and maximum equivalent stress of ~1.5 MPa. The latter is lower than the measured UTS, and the effective stress of 750 mmHg is lower than the measured pressure burst value of 1,500–3,500 mmHg. Potential explanations for these under-predictions include: use of failure strain from uniaxial testing in predictions of probe and pressure burst, which involve multi-axial strains, and the related importance of matrix mechanical anisotropy, which was not included in the FE model; syneresis of water from the sample during impingement (appropriately modeled as poroelasticity); and, the effects of lubrication from a thin layer of water between the probe and sample that exists for the hydrated matrix, even without syneresis. Inclusion of friction led to lower effective pressure, at least for the case of b=0 and μ=0.1 examined, and so higher levels of friction that might actually exist for this probe-matrix system would likely not explain the under-predictions. The substantial differences in the development of reaction force, effective stress, and maximum equivalent stress between the 9.5mm probe used in this study and a smaller 3.5mm probe, as used in a similar study of electrospun polymer vascular grafts (Xie et al., 2016a), based on FEA illustrate the complications in the interpretation of the results when a non-standard probe size is used, more so for testing fundamentally different graft materials. For the matrix studied here, the effective stress at failure (predicted when maximum equivalent strain attains a value of 0.7) is almost 3-fold higher for the 3mm probe versus the 9.5mm probe (0.32MPa vs. 0.11MPa for 0.35mm thickness, frictionless interaction). More generally, the (inferred) effective failure stress achievable is smaller for a larger probe diameter (Fig. 7b).

Measurements performed with harvested ovine carotid and femoral artery samples yielded pressure burst values and ranges consistent with previously published studies (Geelhoed et al., 2019; Syedain et al., 2014). For carotid arteries, pressure burst correlated with probe burst and with circumferential tensile strength. This was not observed for femoral artery samples. Based on the trichrome staining of histological sections, it is apparent that the carotid arteries had a thicker media (muscular) layer, which may have led to more consistent mechanical properties along the length of each carotid artery harvested. Generally, the carotid arteries also possessed fewer collateral branches compared with the femoral arteries, which may also have led to more axial heterogeneity in the femoral arteries, abolishing any such correlation. While Geelhoed et al reported no such correlations for bovine carotid arteries, they did not measure tensile properties along the length of their arteries to demonstrate axial homogeneity, which might explain the discrepancy with our findings.

5. Conclusions

This study demonstrates that our biologically-engineered vascular graft exhibits a strong correlation between pressure burst, probe burst and tensile strength even with its anisotropic tensile properties associated with circumferential alignment of the collagenous matrix fibers. It also confirms that for probe burst to provide a valid strength measurement, uniform material properties along the length of the graft must be established and standardized test conditions used.

Supplementary Material

1

Suppl. Fig. 1: Geometry for probe test simulation using FEA. Mounting plates are not to scale. Axisymmetry was assumed.

Suppl. Fig. 2: Biologically-engineered vascular graft tangent modulus in the circumferential and radial directions.

Acknowledgements

The authors acknowledge funding from NIH R01 HL107572 to R.T.T.

Footnotes

Declarations

ZS, RT: financial interest in Vascudyne, Inc., licensee of related University of Minnesota patents.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References:

  1. Begley MR, Mackin TJ, 2004. Spherical indentation of freestanding circular thin films in the membrane regime. J Mech Phys Solids 52, 2005–2023. [Google Scholar]
  2. Billiar KL, Sacks MS, 2000. Biaxial mechanical properties of the native and glutaraldehyde-treated aortic valve cusp: Part II--A structural constitutive model. J. Biomech. Eng 122, 327–335. [DOI] [PubMed] [Google Scholar]
  3. Dahl SL, Kypson AP, Lawson JH, Blum JL, Strader JT, Li Y, Manson RJ, Tente WE, DiBernardo L, Hensley MT, Carter R, Williams TP, Prichard HL, Dey MS, Begelman KG, Niklason LE, 2011. Readily available tissue-engineered vascular grafts. Sci Transl Med 3, 68ra69. [DOI] [PubMed] [Google Scholar]
  4. Fallon A, Goodchild T, Wang R, Matheny RG, 2012. Remodeling of extracellular matrix patch used for carotid artery repair. J Surg Res 175, e25–34. [DOI] [PubMed] [Google Scholar]
  5. Geelhoed WJ, Lalai RA, Sinnige JH, Jongeleen PJ, Storm C, Rotmans JI, 2019. Indirect Burst Pressure Measurements for the Mechanical Assessment of Biological Vessels. Tissue Eng Part C Methods 25, 472–478. [DOI] [PubMed] [Google Scholar]
  6. Gutowski P, Gage SM, Guziewicz M, Ilzecki M, Kazimierczak A, Kirkton RD, Niklason LE, Pilgrim A, Prichard HL, Przywara S, Samad R, Tente B, Turek J, Witkiewicz W, Zapotoczny N, Zubilewicz T, Lawson JH, 2020. Arterial reconstruction with human bioengineered acellular blood vessels in patients with peripheral arterial disease. J. Vasc. Surg [DOI] [PubMed] [Google Scholar]
  7. Konig G, McAllister TN, Dusserre N, Garrido SA, Iyican C, Marini A, Fiorillo A, Avila H, Wystrychowski W, Zagalski K, Maruszewski M, Jones AL, Cierpka L, de la Fuente LM, L’Heureux N, 2009. Mechanical properties of completely autologous human tissue engineered blood vessels compared to human saphenous vein and mammary artery. Biomaterials 30, 1542–1550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. L’Heureux N, Dusserre N, Marini A, Garrido S, de la Fuente L, McAllister T, 2007. Technology insight: the evolution of tissue-engineered vascular grafts--from research to clinical practice. Nat Clin Pract Cardiovasc Med 4, 389–395. [DOI] [PubMed] [Google Scholar]
  9. Lawson JH, Glickman MH, Ilzecki M, Jakimowicz T, Jaroszynski A, Peden EK, Pilgrim AJ, Prichard HL, Guziewicz M, Przywara S, Szmidt J, Turek J, Witkiewicz W, Zapotoczny N, Zubilewicz T, Niklason LE, 2016. Bioengineered human acellular vessels for dialysis access in patients with end-stage renal disease: two phase 2 single-arm trials. Lancet 387, 2026–2034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Meier LA, Syedain ZH, Lahti MT, Johnson SS, Chen MH, Hebbel RP, Tranquillo RT, 2014. Blood outgrowth endothelial cells alter remodeling of completely biological engineered grafts implanted into the sheep femoral artery. J Cardiovasc Transl Res 7, 242–249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Morota T, Takamoto S, 2013. Development and physical characteristics of novel zero-porosity vascular graft “triplex((R))”. Ann Vasc Dis 6, 67–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ratcliffe A, 2000. Tissue engineering of vascular grafts. Matrix Biol 19, 353–357. [DOI] [PubMed] [Google Scholar]
  13. Reimer J, Syedain Z, Haynie B, Lahti M, Berry J, Tranquillo R, 2017. Implantation of a Tissue-Engineered Tubular Heart Valve in Growing Lambs. Ann Biomed Eng 45, 439–451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Reimer JM, Syedain ZH, Haynie BH, Tranquillo RT, 2015. Pediatric tubular pulmonary heart valve from decellularized engineered tissue tubes. Biomaterials 62, 88–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Roh JD, Sawh-Martinez R, Brennan MP, Jay SM, Devine L, Rao DA, Yi T, Mirensky TL, Nalbandian A, Udelsman B, Hibino N, Shinoka T, Saltzman WM, Snyder E, Kyriakides TR, Pober JS, Breuer CK, 2010. Tissue-engineered vascular grafts transform into mature blood vessels via an inflammation-mediated process of vascular remodeling. Proc Natl Acad Sci U S A 107, 4669–4674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Selvadurai APS, 2006. Deflections of a rubber membrane. J. Mech. Phys. Solids 54, 1093–1119. [Google Scholar]
  17. Stekelenburg M, Rutten MC, Snoeckx LH, Baaijens FP, 2009. Dynamic straining combined with fibrin gel cell seeding improves strength of tissue-engineered small-diameter vascular grafts. Tissue Eng Part A 15, 1081–1089. [DOI] [PubMed] [Google Scholar]
  18. Sugiura T, Matsumura G, Miyamoto S, Miyachi H, Breuer CK, Shinoka T, 2018. Tissue-engineered Vascular Grafts in Children With Congenital Heart Disease: Intermediate Term Follow-up. Semin Thorac Cardiovasc Surg 30, 175–179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Syedain Z, Reimer J, Lahti M, Berry J, Johnson S, Tranquillo RT, 2016. Tissue engineering of acellular vascular grafts capable of somatic growth in young lambs. Nat Commun 7, 12951. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Syedain Z, Reimer J, Schmidt J, Lahti M, Berry J, Bianco R, Tranquillo RT, 2015. 6-month aortic valve implantation of an off-the-shelf tissue-engineered valve in sheep. Biomaterials 73, 175–184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Syedain ZH, Graham ML, Dunn TB, O’Brien T, Johnson SL, Schumacher RJ, Tranquillo RT, 2017a. A completely biological “off-the-shelf” arteriovenous graft that recellularizes in baboons. Science translational medicine 9. [DOI] [PubMed] [Google Scholar]
  22. Syedain ZH, Graham ML, Dunn TB, O’Brien T, Johnson SL, Schumacher RJ, Tranquillo RT, 2017b. A completely biological “off-the-shelf” arteriovenous graft that recellularizes in baboons. Sci Transl Med 9. [DOI] [PubMed] [Google Scholar]
  23. Syedain ZH, Jenson AC, Patel PS, Feagler C, Bahmer L, Faizer R, Tranquillo RT, 2019. Tissue-engineered transcatheter vein valve. Biomaterials 216, 119229. [DOI] [PubMed] [Google Scholar]
  24. Syedain ZH, Meier LA, Bjork JW, Lee A, Tranquillo RT, 2011. Implantable arterial grafts from human fibroblasts and fibrin using a multi-graft pulsed flow-stretch bioreactor with noninvasive strength monitoring. Biomaterials 32, 714–722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Syedain ZH, Meier LA, Lahti MT, Johnson SL, Tranquillo RT, 2014. Implantation of completely biological engineered grafts following decellularization into the sheep femoral artery. Tissue Eng Part A 20, 1726–1734. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Xie Y, Guan Y, Kim SH, King MW, 2016a. The mechanical performance of weft-knitted/electrospun bilayer small diameter vascular prostheses. J Mech Behav Biomed Mater 61, 410–418. [DOI] [PubMed] [Google Scholar]
  27. Xie Y, Guan Y, Kim SH, King MW, 2016b. The mechanical performance of weft-knitted/electrospun bilayer small diameter vascular prostheses. J Mech Behav Biomed Mater 61, 410–418. [DOI] [PubMed] [Google Scholar]
  28. Yang X, Wang L, Guan G, Zhang H, Shen G, Guan Y, Peng L, Li Y, King MW, 2015. Mechanical and biocompatibility performance of bicomponent polyester/silk fibroin small-diameter arterial prostheses. J Appl Biomater Funct Mater 13, e201–209. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1

Suppl. Fig. 1: Geometry for probe test simulation using FEA. Mounting plates are not to scale. Axisymmetry was assumed.

Suppl. Fig. 2: Biologically-engineered vascular graft tangent modulus in the circumferential and radial directions.

NIHMS1698285-supplement-1.pptx (12.6MB, pptx)

RESOURCES