Mechanical failure analysis of patch materials used in aortic arch reconstruction: implications for clinical practice

Abstract

OBJECTIVES

Thick-patch pulmonary homograft, autologous pericardium and CardioCel Neo are common patch materials for aortic arch reconstruction. Insufficient data exist on sutured patch strength and limits of use. We evaluated failure strength of these materials to develop a failure prediction model for clinical guidance.

METHODS

Patch failure strength was evaluated via sutured uniaxial and burst pressure testing. In sutured uniaxial testing, patches were sutured to aortic or Dacron tabs and pulled to failure. In burst pressure testing, patches were sewn into porcine aortas or Dacron grafts and pressurized to failure. Failure membrane tension was calculated. A prediction model of membrane tension versus vessel diameter was generated to guide clinical patch selection.

RESULTS

Combining sutured uniaxial and burst pressure test data, pulmonary homograft failure strength {0.61 [interquartile range (IQR): 0.44, 0.78] N/mm, n = 21} was less than half that of autologous pericardium [2.22 (IQR: 1.65, 2.78) N/mm, n = 15] and CardioCel Neo [1.31 (IQR: 1.20, 1.42) N/mm, n = 20]. Pulmonary homograft burst pressure [245 (IQR: 202, 343) mmHg, n = 7] was significantly lower than autologous pericardium [863 (IQR: 802, 919) mmHg, n = 6] and CardioCel Neo [766 (IQR: 721, 833) mmHg, n = 6]. Our model predicts failure limits for each patch material and outlines safety margins for combinations of aortic diameter and pressure.

CONCLUSIONS

Sutured failure strength of thick-patch pulmonary homograft was significantly lower than autologous pericardium and CardioCel Neo. Patient selection (predicted postoperative arch diameter and haemodynamics) and blood pressure management must be considered when choosing patch material for arch reconstruction. In older children and adolescents, autologous or bovine pericardium may be more suitable materials for aortic patch augmentation to minimize the risk of postoperative patch failure.

Reconstruction of hypoplastic aortic arch in infants, children and young adults often requires patch material to achieve normal aortic size.

INTRODUCTION

Reconstruction of hypoplastic aortic arch in infants, children and young adults often requires patch material to achieve normal aortic size. These patches are exposed to widely variable stresses dependent on vessel size and blood pressure, which scale with patient body surface area (BSA) and age, respectively. However, there is limited data to support evidence-based patch material selection to ensure optimal perioperative and long-term outcomes. Small retrospective cohort studies have compared medium-term outcomes (e.g. freedom from recoarctation, reintervention) among the 3 most commonly used materials: pulmonary homograft, autologous pericardium (AP) and bovine pericardium, but these studies do not relate outcomes to patch mechanical properties [1–4]. Furthermore, patch aortoplasty has been associated with a nontrivial rate of aneurysmal formation and rupture at the repair site [5–12]. Ultimate tensile strength for pericardium and pulmonary homograft has been reported [13–16], but these studies were performed under simple uniaxial loading, which is non-physiologic and does not account for the impact of a surgical suture line. Suture line failure is expected to occur at lower stress compared to unsutured samples, and prior investigators have shown that indirect or derived methods for the evaluation of biological tissue burst pressures are inaccurate and should not be utilized for clinical purposes [17]. Fundamentally, prior patch biomechanics studies are limited and importantly have not translated the data into clinically relevant information to guide patch selection. The tension in an aortic arch patch is proportional to diameter and pressure. It is critical for surgeons to understand the risk of sutured patch failure at different aortic diameters and pressures. This study assessed the sutured mechanical failure strength of patch materials used in aortic arch reconstruction (AAR) to identify a threshold for safe use and to develop a failure prediction model to help guide clinical practice based on material limits.

MATERIALS AND METHODS

Ethics statement

Ex vivo experiments were performed under BCH IRB exemption.

Patch materials evaluated

Samples of thick-patch pulmonary homograft (TPH, n = 3), AP fixed in glutaraldehyde for 5 min (AP, n = 4) and CardioCel Neo (CN) bovine pericardium (n = 4) were tested. Unused patch materials from cardiac surgery cases were collected from the operating room under IRB exemption. Samples were stored in saline at 4°C until testing.

For each sample collected, a pair of orthogonal axes was defined according to visible anatomical landmarks (TPH and AP) or geometric shape (CN). Primary direction (D1) of TPH was defined as the speculated axial direction, identified by tapering thickness from the proximal to distal end, coupled with a thickened region on the distal end corresponding to the suspected branch pulmonary artery confluence (Fig. 1a). Primary direction of AP was defined as the primary direction of the visible vasculature (Fig. 1b). For CN, primary direction was defined along the longer direction of the 5 cm × 8 cm rectangular patch (Fig. 1c). All testing was conducted with patches oriented along these predefined axes to determine whether tissue anisotropy affected failure strength. Of note, this technique of selecting primary direction for the pericardial materials provided consistency but does not definitively correlate with principal directions of material anisotropy.

Figure 1:

Photos of test samples. All images are sized to the same relative scaling. (a–c) Example discarded patch materials received from the operating room, with patch orientation conventions shown by direction 1 (D1) and 2 (D2) arrows in the bottom left of each photo. (d–f) Dogbone-shaped samples for unsutured uniaxial testing. Square 7.5-mm patch samples sewn to (g) porcine aortic tissue or (h and i) Dacron tabs for sutured uniaxial testing. Oval patch samples sewn into (j) porcine aortic arch or (k and l) 25-mm Dacron tube graft for burst pressure testing. Each patch material is shown with thick-patch pulmonary homograft in (a), (d), (g) and (j), autologous pericardium in (b), (e), (h) and (k) and CardioCel Neo in (c), (f), (i) and (l).

Unsutured uniaxial failure test

To assess baseline strength of the patch materials, uniaxial pull-to-failure testing was conducted on unsutured dogbone-shaped samples. Samples were cut into rectangles (TPH 7.5 mm × 15 mm; AP/CN 15 mm × 25 mm for larger gripping area), and an 8-mm punch was used to create a central 3- to 5-mm wide region for failure initiation (Fig. 1d–f). The narrowest width $w$ and thickness $h$ were measured for each cut sample using digital callipers.

Testing was conducted with a tensile pull configuration on a Universal Testing System, instrumented with a 50-N load cell and 250-N pneumatic grips. The grip surfaces were covered with sandpaper to prevent sample slippage. Each sample was preconditioned with 3 cycles of 10% strain at a rate of 10 mm/min, followed by a pull at the same rate that continuously stretched the tissue until failure, as noted by a sharp drop in the loading force (Fig. 2a). Force at failure $F_{\max }$ was recorded for calculation of the normalized loading metric membrane tension (MT) [Eq. (1)]:

Figure 2:

(a) Example force versus displacement data for each material from unsutured uniaxial testing. (b) Example force versus displacement data for each material from sutured uniaxial testing. (c) Example pressure versus time data for each material from the burst pressure testing. AP: autologous pericardium (red dotted line); CN: CardioCel Neo (green dashed line); TPH: thick-patch pulmonary homograft (blue solid line). Marked ‘x’ on each curve corresponds to the failure point of the sample and denotes the force (unsutured and sutured uniaxial test) or pressure (burst pressure test) recorded at failure.

$$\mathrm{MT}=\mathrm{stress}\times \mathrm{thickness}=\frac{F_{\max }}{wh}h=\frac{F_{\max }}{w}.$$ (1)

Sutured uniaxial failure test

To assess strength of sutured patch materials, the same uniaxial test system was used to conduct pull-to-failure testing for TPH, AP and CN samples sutured to tabs of a stronger material. For each patch material, 7.5 × 7.5 mm² were cut and sutured to 15 mm × 20 mm tabs of porcine aorta (TPH) or Dacron tube graft (AP/CN) with 5–0 Prolene using 2-mm suture spacing and bite depth (Fig. 1g–i). Dacron was used for AP and CN given the predicted strength of these materials exceeding the strength of the aortic model. The sutured tabs were gripped and the same preconditioning plus failure test protocol was conducted as described for unsutured uniaxial testing. Failure mode was noted for each sample as (i) suture pull-through: intact suture pulls through the sutured edge of the patch along the direction of pull, (ii) needle-hole tear: intact suture causes needle-hole enlargement and eventual tear propagation between adjacent needle holes, forming an internal patch defect perpendicular to the direction of loading or (iii) patch rupture: patch tears internally or from an edge without suture line involvement. Tests with failure of the suture itself or tab materials were excluded from analysis. Force at failure was recorded (Fig. 2b), and MT was calculated using Eq. (1).

Burst pressure failure test

Inflation-to-failure testing was conducted to determine the burst pressure for each patch material under realistic loading conditions. Oval patches measuring ∼11 mm × 15 mm were sewn into a hole of matching size excised from a test conduit: porcine aortic arch (TPH) or 25-mm Dacron tube graft (AP/CN). Patches were sewn with the same technique as in sutured uniaxial tests. The patched conduit was attached with barbed fittings to a pneumatic pressure transducer (BSP000W, Balluff, Florence, KY) and saline bag. Pressure was recorded to a LabJack T4 DAQ system (LabJack Corporation, Lakewood, CO). The system was filled with water and the saline bag slightly compressed to maintain a low pressure (∼50 mmHg) while baseline circumference measurement of the conduit was taken at the patch location using a silk tie.

To induce failure, a systolic pressure impulse was applied to the conduit by rapidly compressing the filled saline bag between 2 metal plates, mimicking a heart rate of 42 [interquartile range (IQR): 35, 80] bpm. A camera tripod positioned directly in front of the patched conduit captured high-speed video of the burst (1080 p at 240 fps). A ruler attached to the end of the conduit was used for video calibration. The final video frame prior to burst was imported into ImageJ [18] to measure the maximum conduit diameter $d_{\max }$, which was paired with the maximum recorded pressure $P_{\max }$ (Fig. 2c) to calculate the MT at failure from an adjusted version of the Law of Laplace [Eq. (2)]. Mode of failure was noted for each sample as suture pull-through, needle-hole tear or excessive fluid loss (loss of pressure due to significant leakage at the suture line and/or through the needle holes).

$$\mathrm{MT}=\frac{P_{\max }{d}_{\max }}{2}.$$ (2)

Statistical analysis

All data are reported as median (Q1, Q3). Results from different testing modalities, patch materials and material orientations were compared using Wilcoxon rank sum tests. All analyses were conducted using MATLAB 2021b (MathWorks, Natick, MA). All tests were two-sided, and P-values <0.05 were considered statistically significant.

Failure prediction model

A failure prediction model was created to relate vessel diameter and blood pressure to patch failure risk. Using the relationship described in Eq. (2), a plot of MT versus diameter was generated. For a given blood pressure value, MT increases linearly with diameter. Three representative patient categories (neonate, toddler, school-age) were shown as grey-shaded regions on the plot, with the bounds corresponding to the range of z-score 0 diameters from ascending aorta to isthmus and z-score −2 to maximum expected systolic pressures for that age range. The patient categories were defined as: neonate (0- to 28-day old, BSA 0.25–0.26 m², vessel diameters 6.0–9.3 mm, systolic pressures 52–150 mmHg), toddler (1- to 4-year old, BSA 0.39–0.69 m², diameters 7.6–15.4 mm, pressures 70–200 mmHg) and school-age (5- to 18-year old, BSA 0.79–1.84 m², diameters 11.0–25.5 mm, pressures 79–250 mmHg). Sutured uniaxial and burst pressure test data were combined, and the median and range for each patch material were plotted to represent MT levels that are likely to lead to material failure based on our experimental results.

In engineering, a factor of safety is defined as the load-carrying capacity designed into a system beyond its intended loading state. A similar principle was applied to the patch failure predictions with the incorporation of a ‘safety margin’. The maximum expected MT for each patient group across the predefined vessel diameter range was multiplied by factors of 2 and 3, common safety factors for native vascular structures [19], to account for variability and uncertainty in the patch material properties. Considering a safety margin of at least 2 when choosing a patch material will provide the surgeon with higher confidence that the material is safe for that patient size.

RESULTS

In total, 21 unsutured uniaxial (4 TPH, 5 AP, 12 CN), 37 sutured uniaxial (14 TPH, 9 AP, 14 CN) and 19 burst pressure (7 TPH, 6 AP, 6 CN) samples were tested to failure. Representative force versus displacement plots from unsutured and sutured uniaxial tests are shown in Fig. 2a and b, respectively, and a pressure versus time plot from burst pressure testing is shown in Fig. 2c. MT was calculated from force at failure and sample width [Eq. (1)] for uniaxial tests and from pressure and diameter at failure [Eq. (2)] for burst pressure tests. The MTs at failure for all patch materials and test modalities are listed in Supplementary Material, Table S1 and visualized in Fig. 3.

Figure 3:

All failure data for (a) thick-patch pulmonary homograft, (b) autologous pericardium and (c) CardioCel Neo. Samples are divided by orientation D1 and D2 as defined in Fig. 1. Membrane tension at failure was compared between orientations and among test modalities. Black * denotes P < 0.05 between testing modalities. Red ** denotes P < 0.05 between material directions.

Unsutured uniaxial failure test

Only CN demonstrated orientation-dependent strength that reached statistical significance. For CN, MT at failure was 2.7× higher in the D2 direction than in the D1 direction (P = 0.026). For TPH, MT at failure in the D2 direction was 2.6× higher than in the D1 direction but not statistically significant (P = 0.33) (Supplementary Material, Table S1). For AP, MT was 1.6× higher in D1 than in D2 (P = 0.200). Combining D1 and D2 data for each material, TPH failure strength was 4.4× lower than AP (P = 0.016) and 5.7× lower than CN (P = 0.004). AP and CN exhibited similar unsutured failure strength (P = 0.88).

Sutured uniaxial failure test

The most common failure mode during sutured uniaxial testing for all 3 materials was suture pull-through (TPH 6/14, AP 9/9, CN 13/14), followed by needle-hole tear (TPH 8/14), then patch rupture (CN 1/14) (Fig. 4a–c). In contrast to unsutured uniaxial testing, none of the patch materials demonstrated orientation-dependent failure strength with this test modality (D1 versus D2: P = 0.32 for TPH, P = 0.29 for AP, P = 0.85 for CN) (Supplementary Material, Table S1). Combining D1 and D2 data, TPH failure strength was 4.2× lower than AP (P < 0.001) and 2.1× lower than CN (P < 0.001). AP sutured uniaxial failure strength was 1.7× higher than CN (P < 0.001).

Figure 4:

Representative failure photos of each patch material for (a–c) sutured uniaxial and (d–i) burst pressure tests. Sutured uniaxial tests: (a) needle-hole tear of a thick-patch pulmonary homograft sample; (b) suture pull-through of an autologous pericardium sample; and (c) patch rupture without suture line involvement in a CardioCel Neo sample. Burst pressure tests: (d and g) needle-hole tear of a thick-patch pulmonary homograft sample; (e and h) failure from excessive fluid loss in an autologous pericardium sample; and (f and i) suture pull-through of a CardioCel Neo sample.

Burst pressure failure test

All TPH patches failed via needle-hole tear (6/7) or suture pull-through (1/7). AP and CN patches failed almost exclusively by excessive fluid loss (AP 6/6, CN 5/6) with only 1 CN demonstrating suture pull-through failure (Fig. 4d–i). TPH burst pressure [245 (IQR: 202, 343) mmHg] was significantly lower than for AP [863 (IQR: 802, 919) mmHg] and CN [765 (IQR: 721, 833) mmHg] (P = 0.001 for each). AP and CN demonstrated similar burst pressures at failure (P = 0.102). Under this test modality, none of the materials demonstrated orientation-dependent strength (Supplementary Material, Table S1). Combining D1 and D2 data, TPH failure strength was 3.4× lower than AP (P = 0.001) and 3× lower than CN (P = 0.001). AP BPT failure strength was higher than CN but did not reach statistical significance (P = 0.065). All patch materials exhibited higher failure strength in unsutured uniaxial testing compared to both sutured tests, though this difference was not significant for TPH (Fig. 3). Unsutured failure strength for TPH was roughly 2× sutured strength (P = 0.28), for AP was >2× (P = 0.004) and for CN was 5× (P < 0.001).

Comparison among sutured material tests

MT at failure was 1.6× higher in sutured uniaxial tests than in burst pressure tests for TPH (P = 0.03) and 1.7× for AP (P < 0.001) while for CN, MT at failure was similar across both tests (P = 0.84). When considering combined sutured uniaxial and burst pressure data, TPH failure strength [0.61 (IQR: 0.44, 0.78) N/mm] was 3.6× lower than AP [2.22 (IQR: 1.54, 2.66) N/mm, P < 0.001] and 2.1× lower than CN [1.31 (IQR: 1.20, 1.42) N/mm, P < 0.001]. AP exhibited higher sutured failure strength compared to CN (P < 0.001) (Fig. 5).

Figure 5:

Comparison of membrane tension at failure for each patch material with combined sutured uniaxial and burst pressure data. Data from both patch orientations are combined for each material. AP: autologous pericardium (red, middle); CN: CardioCel Neo (green, right); TPH: thick-patch pulmonary homograft (blue, left). Outliers are shown as red ‘+’ symbols. Black * denotes P < 0.05 between patch materials.

Failure prediction model

A failure prediction model was generated from sutured failure strength data (sutured uniaxial + burst pressure testing) to guide patch selection based on patient size (Fig. 6a). The model was visualized on MT versus vessel diameter axes (Fig. 6b–f). Expected MT ranges for 3 patient groups were shown as grey-shaded regions (Fig. 6b). The orange lines mark constant blood pressures. Experimental patch failure MT ranges were added (Fig. 6c) as outlined boxes for each material. The point at which a shaded region intersects with the outlined patch failure range predicts patch failure. Safety margins of 2 and 3 are added to the graph to give clinicians a visual representation of these buffer levels (Fig. 6d–f).

Figure 6:

Failure prediction model for aortic arch reconstruction as a function of patient size. (a) Diagram of membrane tension (MT) exerted on the aortic wall calculated by the Law of Laplace. (b) Expected physiologic MT ranges for 3 patient groups (grey boxes) based on z-score zero thoracic aortic diameters and z-score −2 to maximum systolic pressures for that group. Neonate (0- to 28-day old, BSA 0.25–0.26 m²) was defined by vessel diameters 6.0–9.3 mm and systolic pressures 52–150 mmHg. Toddler was defined as a 1- to 4-year old (BSA 0.39–0.69 m²), with vessel diameters 7.6–15.4 mm and pressures 70–200 mmHg. The school-age category was defined as a 5- to 18-year old [body surface area (BSA) 0.79–1.84 m²] with diameters 11–25.5 mm and pressures 79–250 mmHg. Orange lines mark constant blood pressures (100, 200, 300 mmHg). (c) Patch failure MT ranges from sutured experimental data as outlined boxes in solid blue [thick-patch pulmonary homograft (TPH)], dotted red [autologous pericardium (AP)] and dashed green [CardioCel Neo (CN)], with central horizontal lines denoting median values. Note that the vertical axis is expanded compared to (b) to include the maximum AP failure (3.4 N/mm). The point at which a grey box intersects the outlined patch failure range predicts patch failure. For example, if a school-aged patient with an arch diameter of 22 mm achieved a blood pressure of 200 mmHg, the calculated MT exerted on the vessel wall (i.e. intersection of the 22-mm vessel diameter and 200-mmHg orange line) would exceed that of the experimental failure strength of TPH, predicting high risk for patch failure. Isolated (d) neonate, (e) toddler and (f) school-age patient groups with the addition of safety margins (SM) of 2 and 3. Note that the vertical scale for each of these isolated plots differs from (c). The minimum MT at failure for each patch material is shown as horizontal lines. The original dark grey-shaded boxes from (b) and (c) are shown with progressively lighter shading corresponding to increasing SM (more light = SM2 and most light = SM3). The point at which a box intersects with the minimum patch failure line predicts patch failure risk at that SM.

DISCUSSION

Patch material selection for AAR historically has been based on surgeon preference without quantified thresholds of patch failure available to support material selection for the expected physiologic loads in an individual patient. Repair site aneurysmal formation or rupture, albeit rare, are potentially lethal complications secondary to patch failure. The stresses imposed on patches used in AAR vary dramatically depending on the patient’s anatomy and haemodynamics and can have consequential effects on patch integrity. Therefore, this study delineated a threshold for the use of common AAR patch materials based on sutured mechanical failure strength with a range of safety margins.

Unsutured uniaxial failure testing substantially overestimates in situ patch material strength due to the impact of the suture line on patch integrity. Unsutured testing resulted in at least twice the MT at failure when compared to sutured experiments for all 3 materials. Of note, our unsutured data for TPH and CN are similar to the ultimate tensile strength for these materials reported in the literature [13–15, 20]. The burst pressure testing failure strength was lower than that of sutured uniaxial testing for both TPH and AP. The more physiologic loading state in burst testing may be a more accurate method to assess these materials, coinciding with studies suggesting that indirect methods may be inadequate for the evaluation of biological tissue [17]. Suture technique was kept constant across all tests with 2-mm bite depth and spacing; we expect that suture pull-through failure may be delayed with an increased bite depth, while needle-hole tear propagation may be avoided by staggering depth of adjacent bites to increase their effective spacing.

TPH was shown to be significantly weaker than pericardium across all testing modalities. Compared to AP and CN, TPH exhibited failure strength less than half in sutured uniaxial testing and roughly one-third in burst pressure testing. When sutured data were combined, TPH was significantly weaker with a failure strength less than half that of AP and CN. AP demonstrated statistically higher failure strength than CN in sutured uniaxial tests. Although both of these materials had burst pressures significantly higher than physiologically achievable [21, 22], these findings may be clinically relevant when considering a safety margin. These conclusions are based on averages of material strength over a few tested samples. However, as with any biologic tissue, there exists variability of mechanical properties across specimens. With limited availability of autologous and cryopreserved tissue resulting in a relatively small sample size, we were unable to deduce which of the 3 materials demonstrates the greatest degree of variability. The authors suspect that CN may have the most consistent mechanical properties given regulated standards during the manufacturing process, whereas TPH and AP may exhibit significant variability based on donor age, tissue thickness, time from harvest, etc. Surgeons must note that the strength of a given patch material used intraoperatively may be above or below that estimated for a general population of donor tissue.

The combined sutured data were used to develop a failure prediction model to assist surgeons in AAR patch material selection based on mechanical limits. Our model demonstrates how aortic diameter, expected peak blood pressure and patch material strength contribute to patch failure risk. Figure 6b highlights the combination of aortic diameter and expected peak blood pressure (predicted 150 mmHg for neonate, 200 mmHg for toddler and 250 mmHg for school-age child) across 3 paediatric subgroups and the resulting MT that is exerted on the patch. Figure 6c shows the MT at which TPH, CH and AP patches fail over the range seen in our data. This provides the clinician with a quantitative model to understand if the use of a patch in a particular patient will put them at risk for patch failure. These patch failure prediction models are individualized for neonates (Fig. 6d), toddlers (Fig. 6e) and school-age children (Fig. 6f) given the different expected peak blood pressure and aortic diameter ranges for each group.

Given the range of biologic variability, it is not enough to base a clinical decision on the data from the limited number of samples tested in this study. Even with highly characterized materials such as metals and plastics, engineers always incorporate an additional factor of safety when considering if a material is safe to use for an application. Normal arteries incorporate a safety margin with burst pressures of 1405 (standard deviation: 342) mmHg (range 590–1950) for adult aortas, with other reports of 1700–3200mmHg for mid-sized arteries [21, 23–27]. This equates to a native burst pressure safety margin of 2.2–3.6 in adult aortas. Incorporating a safety margin for patch failure is consistent with both nature and modern engineering design. The degree of margin that should be utilized for aortic arch patches has not been standardized. In our model, we plot a safety margin of 1 (baseline, no additional margin), 2 (2× the baseline) and 3 (3× the baseline). These safety margins are at or below the normal native aortic burst pressure margin.

Neonates have a combination of a small aortic diameter and low peak pressure (predicted 150 mmHg), resulting in a peak MT of <0.1 N/mm. In Fig. 6d, the expected patch loading falls below the minimum tested TPH failure strength even if the clinician elects to use a safety margin of 3. For toddlers, peak MT can be as high as 0.2 N/mm for a 15-mm diameter aorta and 200-mmHg pressure (Fig. 6e). This falls below the TPH minimum, but if the clinician selects a safety margin of 2, the combination of aorta diameter above 11 mm and peak pressure of 200 mmHg puts a TPH patch at risk of failure. For school-age children (Fig. 6f), even with no additional safety margin, aortic diameters above 16 mm are at high risk of TPH patch failure. Using this model of patch material failure, if a surgeon adopts a safety margin of 2, TPH should not be considered for use in patients over 1 month with a reconstructed aorta above 11-mm diameter (Fig. 6e) and CN should not be considered for use in any aorta above 20-mm diameter (Fig. 6f).

Patch material selection is multifaceted and includes consideration of several crucial factors. In the short term, suture retention and burst pressure strength, anatomic location, favourable handling characteristics to allow for appropriate tissue reconstruction (i.e. malleability, stretchability, etc.), product availability and thrombogenicity are of importance. In the long term, immunogenicity, risk of calcification, degree of tissue ingrowth (i.e. endothelization, neovascularization) and remodelling and growth potential to prevent stenosis or aneurysmal dilation are critical elements [28–33]. Specifically, TPH has improved availability over autografts, superior handling, growth potential, durability, haemodynamic advantages and resistance to infection. Costliness and material strength are 2 drawbacks with using TPH. AP is non-antigenic, has potential for growth and remodelling and has limited calcification when unfixed but has limited availability in patients requiring multiple operations and limited mechanical strength leading to progressive dilatation in situ and a predisposition to patch aneurysm formation. Glutaraldehyde fixation improves material stiffness, reducing the risk of aneurysm but increases the risk of calcification. CN has increased availability, has favourable surgical characteristics including uniform thickness, flexibility, elasticity and handling, demonstrates potential for re-endothelialization and undergoes a proprietary fixation process leading to decreased rates of calcification. However, CN is costly and maintains antigenicity. In mid- to long-term follow-up, CN may be a risk factor for restentosis requiring reintervention following AAR, whereas TPH and AP appear to demonstrate similar freedom from recoarctation [3, 34, 35].

Aside from these considerations, tissue mechanical properties, specifically material strength, are also essential components in this decision-making algorithm. Knowledge of tissue biomechanics may prevent rare instances of overt patch rupture as well as provide insight into more commonly encountered issues related to patch material failure such as aneurysmal dilatation or stenosis. It is especially important to understand that stresses placed on the selected patch material scale linearly with vessel diameter and pressure. This concept is most clinically relevant in the immediate postoperative period when the patch is most vulnerable to rupture. Over time, the patch integrates into the surrounding tissue, drastically increasing the material tolerance to higher pressures and vessel wall stresses. These presented experiments, results and failure prediction model are intended to mimic the immediate postoperative period, prior to endothelization or fibrosis. During this time, initial reconstructed vessel diameter is usually dictated by the patient’s age and BSA to achieve a z-score 0 size. Therefore, perioperative blood pressure management in this patient population is extremely important to minimize patch MT. This represents a multidisciplinary objective shared amongst the surgeon, intensivist, cardiologists, mid-levels and nurses, with all members of the care team recognizing the significance of strict blood pressure control and review of any transient pressure elevations to determine potential medical management adjustments that may be necessary. Although often considered an engineering principle, surgeons must account for mechanical properties when selecting a patch material to facilitate a safe and durable repair/reconstruction. This study was performed to provide clarity and guidance to the surgeon regarding tissue strength, 1 factor in the complex multifactorial equation that is patch material selection.

Limitations

Given the limited availability of autologous or cryopreserved tissue, our patch sample size was relatively small. There is potential for variation across patch donors, which may not have been fully captured in our dataset that reflects only 3 or 4 different donor patches for each material. Therefore, mean MT at failure may have been over- or underestimated compared to the general population of donor tissue. Our samples were not all tested within the same timeframe due to variability in collection from the OR and staff/researcher availability. Different amounts of storage time may have impacted overall patch strength, as shown in prior mechanical studies [36]. Furthermore, we did not consider the impact of material fatigue due to cyclic loading. In situ, patch materials may fail at lower MT than measured in this study due to the impact of repeated stress insults. Biologic material properties are also rate and orientation dependent. Uniaxial testing appears to slightly overestimate material strength compared to biaxial or burst pressure testing, which applies multidirectional and/or impulse loads to more accurately reflect physiologic environments. Lastly, AP and CN failure MT during burst testing was somewhat inconclusive as no obvious failure zone in the patch was identified in all but 1 sample. Despite this limitation, all AP and CN samples were shown to withstand pressures much higher than physiologically achievable. As a follow-up to this study, we plan to evaluate the effects of cyclic loading, suture bite depth and bite spacing on patch strength.

CONCLUSIONS

Sutured TPH failure strength was significantly lower than that of AP and CN. Patient selection (predicted postoperative aortic diameter and haemodynamics), patch material and blood pressure management must be considered when planning AAR. In older children and adolescents, AP or CN may be more suitable materials for aortic patch augmentation to minimize the risk of postoperative patch failure.

Glossary

ABBREVIATIONS

AAR

Aortic arch reconstruction

AP

Autologous pericardium

BSA

Body surface area

CN

CardioCel Neo

IQR

Interquartile range

MT

Membrane tension

TPH

Thick-patch pulmonary homograft

 

Presented at the 8th WCPCCS, Washington, DC, USA, 28 August 2023.

FUNDING

This work was supported by internal funds.

Conflict of interest: none declared.

DATA AVAILABILITY

All relevant data are within the manuscript and its supporting Information files.

Author contributions

Dominic P. Recco: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing—original draft; Writing—review & editing. Shannen B. Kizilski: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing—original draft; Writing—review & editing. Lauren E. Marshall: Data curation; Investigation; Methodology; Validation; Visualization; Writing—review & editing. Patrick D. Earley: Data curation; Investigation; Methodology; Writing—review & editing. Nicholas E. Kneier: Data curation; Investigation; Methodology; Writing—review & editing. Pedro J. del Nido: Conceptualization; Writing—review & editing. Peter E. Hammer: Conceptualization; Investigation; Methodology; Resources; Writing—review & editing. David M. Hoganson: Conceptualization; Investigation; Methodology; Resources; Supervision; Visualization; Writing—original draft; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Nabil Hussein and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.