Summary
Background
Percutaneous pulmonary valve implantation (PPVI) using the bovine jugular vein Melody ® valve (Medtronic Inc., Minneapolis, MN, USA) is safe and effective. However, post-procedural complications have been reported, the reasons for which are unclear.
Objective
To assess the impact of PPVI procedural steps on valvular histology and leaflet mechanical behaviour.
Methods
Three different valved stents (the Melody ® valve and two homemade stents with bovine and porcine pericardium) were tested in vitro under four conditions: (1) control group; (2) crimping; (3) crimping plus inflation of low-pressure balloon; (4) condition III plus post-dilatation (high-pressure balloon). For each condition, valvular leaflets (and a venous wall sample for Melody ® stents) were taken for histological analysis and mechanical uniaxial testing of the valve leaflets.
Results
Among the Melody ® valves, the incidence of transverse fractures was significantly higher in traumatized samples compared with the control group ( P < 0.05), whereas the incidence and depth of transverse fractures were not statistically different between the four conditions for bovine and porcine pericardial leaflets. No significant modification of the mechanical behaviour of in vitro traumatized Melody ® valvular leaflets was observed. Bovine and porcine pericardia became more elastic and less resilient after balloon expansion and post-dilatation (conditions III and IV), with a significant decrease in elastic modulus and stress at rupture.
Conclusion
Valved stent implantation procedural steps induced histological lesions on Melody ® valve leaflets. Conversely, bovine and porcine pericardial valved stents were not histologically altered by in vitro manipulations, although their mechanical properties were significantly modified. These data could explain some of the long-term complications observed with these substitutes.
Résumé
Contexte
Le remplacement valvulaire pulmonaire percutané utilisant le stent valvé Melody ® est efficace et sécurisé, cependant, des endocardites infectieuses surviennent sur ces prothèses sans explication évidente.
Objectif
Évaluer l’impact des manipulations pré-implantatoires sur la structure histologique et le comportement mécanique des feuillets valvulaires.
Méthodes
Nous avons testé in vitro 3 types de stents valvés (prothèse Melody ® , stents valvés en péricarde bovin et porcin fabriqués manuellement) dans 4 conditions différentes : (1) groupe témoin ne subissant aucune manipulation ; (2) sertissage sur un ballonnet ; (3) sertissage + inflation du ballonnet à basse pression ; (4) groupe III + surdilatation par ballonnet à haute pression. À l’issu de chaque manipulation, les feuillets valvulaires étaient prélevés sur les stents, puis analysés sur le plan histologique et mécaniques (test de traction uniaxiale).
Résultats
Pour les valves Melody ® , on retrouvait plus de lésions histologiques sur les feuillets valvulaires dans les groupes II, III et IV par rapport au groupe témoin ( p < 0,05). L’incidence de ces lésions n’était pas différente entre les 4 conditions pour les stents valvés péricardiques. Les propriétés mécaniques des valves Melody ® traumatisées n’étaient pas modifiées. Les péricardes bovin et porcin devenaient plus élastiques et moins résistants dans les conditions III et IV, avec une diminution du module d’élasticité et du stress à la rupture.
Conclusion
Les manipulations réalisées en salle de cathétérisme entraînent des lésions histologiques significatives sur les feuillets valvulaires des prothèses Melody ® . Les stents valvés en péricarde bovin et porcin ne sont pas altérés histologiquement par ces manipulations mais voient leurs propriétés mécaniques se modifier significativement. Ces données pourraient expliquer certaines complications observées à long terme avec ces substituts.
Background
Patients undergoing surgical right ventricular outflow tract (RVOT) reconstruction are subject to conduit degeneration later in life, requiring further interventions to alleviate the pulmonary stenosis and/or regurgitation that ensues. Since the first reported case in 2000, percutaneous pulmonary valve replacement (PPVI) using the Melody ® valve (Medtronic Inc., Minneapolis, MN, USA) – a glutaraldehyde fixed bovine jugular vein (BJV) valve mounted on a balloon-expandable stent – is now recognized as an alternative to surgical pulmonary valve replacement in patients with failing RVOT . Recent reports showed that PPVI was feasible at a relatively low risk, and mid-term follow-up demonstrated a sustained improvement in haemodynamics up to 7 years after implantation . Despite these promising results, various mid-term and long-term complications have been described, including cases of infective endocarditis (IE) . The reported annualized rate of IE ranges from 2.4% to 3.9% per patient-year . We and others have recently shown that IE occurs more frequently after PPVI than surgical pulmonary valve replacement . IE also involves other valved stents made with different valvular substrates – i.e. the Edwards SAPIEN ® valve (Edwards Lifesciences, Irvine, CA, USA), made with bovine pericardium, and the CoreValve ® valve (Medtronic Inc.), made with porcine pericardium . These results suggest that IE might be related to the implantation technique (i.e. percutaneous or surgical) used for valvular placement. One of the main differences between surgical and transcatheter valve replacement is that percutaneous valves undergo several manipulations before implantation (i.e. crimping) and during implantation (i.e. balloon expansion), whereas surgical prostheses are placed directly in the pulmonary pathway without valvular damage, theoretically. Traumatic injury to biological valve leaflets has been reported during valved stent preparation . In a recent work, we demonstrated that there was selective adhesion of Staphylococcus aureus and S. sanguinis pathogenic strains on healthy Melody ® valve tissue, which increased after implantation procedural steps .
In this in vitro study, we aimed to assess the effects of PPVI procedural steps on the histological and mechanical properties of Melody ® valve leaflets, and to compare these results with other tissues used for valved stent fabrication (i.e. bovine and porcine pericardium).
Methods
Valvular substrates
Three types of valved stents were tested experimentally. The Melody ® valve was obtained from Medtronic Inc. and stored in its commercial packaging. For the bovine pericardium valve, valvular leaflets were obtained from a bovine pericardial patch (10 × 15 cm; Edwards Lifesciences), cut onto a 21-mm homemade three-leaflet valvular mould, and sutured into a vascular stent (CP8Z34; NuMED Canada Inc., Cornwall, ON, Canada); valved stents were then stored in 0.625% glutaraldehyde until use. For the porcine pericardium valve, valvular leaflets were obtained from a porcine pericardial patch (8 × 6 cm; Vascutek Terumo Ltd., Swillington, Leeds, UK); porcine pericardial valved stents were then prepared similarly and stored in 0.625% glutaraldehyde until use.
In vitro manipulations
For each valved stent, we compared four experimental conditions, reproducing the sequential procedural steps leading to conventional PPVI ( Fig. 1 ). Before manipulation, valved stents were rinsed twice for 2 minutes each in 500-mL saline baths to remove the glutaraldehyde.
Condition I: control group
Valved stents were not manipulated.
Condition II: compression group
Valved stents were crimped manually on 5-mL and 2.5-mL sterile syringes, and then onto the 22-mm balloon of the 22F Ensemble ® delivery system (Medtronic Inc.) ( Fig. 1 A). The sheath was advanced to cover the balloon-mounted valved stent over a 5-minute period; this duration was chosen arbitrarily, and aimed to reproduce the duration of crimping during conventional PPVI. The compressed prostheses were flushed regularly with a saline solution. The sheath was then drawn back, and the valved stent was manually enlarged and removed, avoiding damage to the valve.
Condition III: compression/expansion group
Valved stents were first prepared as in condition II. After the sheath was drawn back, the valved stents were deployed in a 20-mm GORE-TEX ® (W. L. Gore & Associates Ltd., Dundee, UK) conduit by inflation of the inner and outer balloons of the delivery system. The balloons were then deflated and the delivery system was removed ( Fig. 1 A and B).
Condition IV: compression/expansion/post-dilatation
Valved stents were first prepared as in condition III. After valve deployment, post-dilatation was performed, using a 22-mm high-pressure balloon (Atlas Gold; Bard Peripheral Vascular, Inc., Tempe, AZ, USA), inflated at 20 atm for 5 seconds ( Fig. 1 A–C).
The valved stents were analysed in each in vitro condition. For each Melody ® valved stent in each condition, three additional samples of the BJV wall adjacent to the leaflets within the sinuses were taken, using an 8-mm diameter (i.e. 0.5 cm 2 ) trepan, for histological tests. After sampling, valvular leaflets and BJV wall fragments were stored in 0.625% glutaraldehyde until histological processing (within 24 hours) or in a saline solution before immediate mechanical testing ( Fig. 1 D) .
For each substrate, five leaflets from two valved stents underwent histological and mechanical evaluation in each of the four conditions.
Uniaxial tensile test
We determined the mechanical properties of the leaflets using uniaxial tensile tests with a universal testing machine (Adamel Lhomargy MTS 100; MTS Systems Corp., Eden Prairie, MN, USA) equipped with TestWorks 4 software (MTS Systems Corp.). The mechanical properties of native and prosthetic valvular leaflets have been published previously with validated methods .
Five leaflets were tested for each substrate in each of the four conditions. Tissue thickness was measured with a caliper with a precision of 0.01 mm. A 100-N load cell was used to apply a tensile force to the tissue samples, and the tissue was stretched at a constant rate of 0.5 mm/min to obtain a stress-strain curve, on which stress at break, elongation at break, ultimate tensile strength (the maximum stress that a material can withstand while being stretched before breaking) and elastic modulus (the slope of the stress-strain curve in the elastic deformation region) were recorded.
Histological analysis
Macroscopic analysis preceded microscopic evaluation. After paraffin embedding, 5 μm-thick samples were stained with haematoxylin and eosin (H&E), and digitalized pictures were obtained at × 5 and × 20 magnifications. Transverse tissue fracture, the basic lesion previously described for bovine pericardial valved stents, originated from one surface of the sample, deep inside the tissue ; it was considered as arbitrarily significant when its depth exceeded 25% of the sample’s thickness. The depth of the biggest fracture was calculated as a percentage (fracture length/sample thickness). The number of fractures and the depth of the biggest fracture were determined at × 5 magnification. The pathologist was blinded to the type of in vitro manipulation that the sample underwent.
Statistical analysis
Results are expressed as mean (standard deviation) or median (range) for continuous variables, or as number (percentage) for categorical variables. The data from these experiments were analysed with non-parametric Mann-Whitney or Kruskal-Wallis tests; these tests were performed to compare variables between two groups (i.e. conditions I vs II; I vs III; I vs IV; II vs III; II vs IV; and III vs IV). The value of statistical significance was set at P ≤ 0.05.
Results
Uniaxial tensile tests
Bovine, porcine and Melody ® valve leaflets had median thicknesses of 0.57 mm (0.45–0.7 mm), 0.23 mm (0.18–0.25 mm) and 0.1 mm (0.08–0.1 mm), respectively, and median widths of 8 mm (5–12 mm), 9 mm (5–11 mm) and 6 mm (5–9 mm), respectively. Uniaxial measurements showed typical non-linear J-shaped stress-strain curves. Table 1 shows the mechanical behaviour of each substrate obtained in the different in vitro conditions.
| Condition I: control ( n = 5) | Condition II: compression ( n = 5) | Condition III: expansion ( n = 5) | Condition IV: post-dilatation ( n = 5) | P | |
|---|---|---|---|---|---|
| Melody ® leaflet | |||||
| Stress at break (MPa) | 5.7 (2.6–8) | 5.7 (2.8–7.9) | 4.6 (3.7–13) | 5.5 (3.7–11.5) | I vs II, P = 0.48 |
| I vs III, P = 0.89 | |||||
| I vs IV, P = 0.87 | |||||
| II vs III, P = 0.66 | |||||
| II vs IV, P = 0.65 | |||||
| III vs IV, P = 0.79 | |||||
| Elongation at break (%) | 34.6 (31–54) | 44 (32–49) | 66 (19–83) | 58 (30–67) | I vs II, P = 0.74 |
| I vs III, P = 0.45 | |||||
| I vs IV, P = 0.38 | |||||
| II vs III, P = 0.48 | |||||
| II vs IV, P = 0.43 | |||||
| III vs IV, P = 0.7 | |||||
| Ultimate tensile strength (MPa) | 5.7 (2.7–8.8) | 5.6 (2.5–8.9) | 5.2 (3.7–13) | 5.6 (3.9–11.8) | I vs II, P = 0.64 |
| I vs III, P = 0.68 | |||||
| I vs IV, P = 0.66 | |||||
| II vs III, P = 0.67 | |||||
| II vs IV, P = 0.65 | |||||
| III vs IV, P = 0.79 | |||||
| Elastic modulus (MPa) | 0.16 (0.1–0.4) | 0.19 (0.18–0.4) | 0.2 (0.16–0.32) | 0.24 (0.18–0.31) | I vs II, P = 0.34 |
| I vs III, P = 0.91 | |||||
| I vs IV, P = 0.71 | |||||
| II vs III, P = 0.71 | |||||
| II vs IV, P = 0.90 | |||||
| III vs IV, P = 0.14 | |||||
| Bovine pericardium | |||||
| Stress at break (MPa) | 8.7 (8.2–10) | 10.2 (7–12.9) | 6 (5–8) | 7 (5.7–9.1) | I vs II, P = 0.5 |
| I vs III, P = 0.02 | |||||
| I vs IV, P = 0.005 | |||||
| II vs III, P = 0.18 | |||||
| II vs IV, P = 0.5 | |||||
| III vs IV, P = 0.30 | |||||
| Elongation at break (%) | 45 (39–61) | 48 (24–60) | 45.8 (40–49) | 42.6 (35–58) | I vs II, P = 0.72 |
| I vs III, P = 0.62 | |||||
| I vs IV, P = 0.58 | |||||
| II vs III, P = 0.84 | |||||
| II vs IV, P = 0.99 | |||||
| III vs IV, P = 0.75 | |||||
| Ultimate tensile strength (MPa) | 9.7 (6.3–12) | 10.7 (7.2–14) | 7 (6.3–8.6) | 8.4 (6–10.2) | I vs II, P = 0.85 |
| I vs III, P = 0.08 | |||||
| I vs IV, P = 0.13 | |||||
| II vs III, P = 0.14 | |||||
| II vs IV, P = 0.12 | |||||
| III vs IV, P = 0.37 | |||||
| Elastic modulus (MPa) | 0.42 (0.37–0.54) | 0.52 (0.41–0.54) | 0.36 50.32–0.38) | 0.38 (0.33–0.58) | I vs II, P = 0.33 |
| I vs III, P = 0.042 | |||||
| I vs IV, P = 0.19 | |||||
| II vs III, P = 0.001 | |||||
| II vs IV, P = 0.22 | |||||
| III vs IV, P = 0.43 | |||||
| Porcine pericardium | |||||
| Stress at break (MPa) | 9 (7.5–10.8) | 9.4 (7.1–12) | 7.6 (2.8–8) | 4.9 (1.7–7) | I vs II, P = 0.35 |
| I vs III, P = 0.15 | |||||
| I vs IV, P = 0.032 | |||||
| II vs III, P = 0.06 | |||||
| II vs IV, P = 0.33 | |||||
| III vs IV, P = 0.28 | |||||
| Elongation at break (%) | 31 (25–36) | 29 (21–37) | 36 (32–58) | 24 (20–35) | I vs II, P = 0.54 |
| I vs III, P = 0.17 | |||||
| I vs IV, P = 0.25 | |||||
| II vs III, P = 0.17 | |||||
| II vs IV, P = 0.66 | |||||
| III vs IV, P = 0.09 | |||||
| Ultimate tensile strength (MPa) | 9.2 (7.6–11) | 10.5 (7–12) | 8.2 (3–8.7) | 5 (2–7) | I vs II, P = 0.41 |
| I vs III, P = 0.24 | |||||
| I vs IV, P = 0.03 | |||||
| II vs III, P = 0.12 | |||||
| II vs IV, P = 0.27 | |||||
| III vs IV, P = 0.23 | |||||
| Elastic modulus (MPa) | 0.49 (0.4–0.8) | 0.7 (0.36–1) | 0.31 (0.24–0.44) | 0.43 (0.15–0.46) | I vs II, P = 0.28 |
| I vs III, P = 0.04 | |||||
| I vs IV, P = 0.031 | |||||
| II vs III, P = 0.35 | |||||
| II vs IV, P = 0.31 | |||||
| III vs IV, P = 0.92 | |||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
