Hernia (2011) 15:4752 DOI 10.

1007/s10029-010-0731-7

ORIGINAL ARTICLE

Anisotropic evaluation of synthetic surgical meshes

E. R. Saberski S. B. Orenstein Y. W. Novitsky

Received: 25 May 2010 / Accepted: 26 August 2010 / Published online: 30 September 2010 Springer-Verlag 2010

Abstract Introduction The material properties of meshes used in hernia repair contribute to the overall mechanical behavior of the repair. The anisotropic potential of synthetic meshes, representing a difference in material properties (e.g., elasticity) in different material axes, is not well dened to date. Haphazard orientation of anisotropic mesh material can contribute to inconsistent surgical outcomes. We aimed to characterize and compare anisotropic properties of commonly used synthetic meshes. Methods Six different polypropylene (Trelex, ProLiteTM, UltraproTM), polyester (ParietexTM), and PTFEbased (Dualmesh, Innit) synthetic meshes were selected. Longitudinal and transverse axes were dened for each mesh, and samples were cut in each axis orientation. Samples underwent uniaxial tensile testing, from which the elastic modulus (E) in each axis was determined. The degree of anisotropy (k) was calculated as a logarithmic expression of the ratio between the elastic modulus in each axis. Results Five of six meshes displayed signicant anisotropic behavior. UltraproTM and Innit exhibited approximately
Presented at Hernia Repair 2010, Annual Meeting of the American Hernia Society, March 2010, Orlando, Florida, USA. E. R. Saberski Y. W. Novitsky University of Connecticut School of Medicine, Farmington, Connecticut, USA e-mail: ESaberski@student.uchc.edu S. B. Orenstein Y. W. Novitsky (&) Connecticut Comprehensive Center for Hernia Repair, Department of Surgery, University of Connecticut Health Center, 263 Farmington Avenue, MC 3955, Farmington, Connecticut 06030, USA e-mail: ynovit@gmail.com

12- and 20-fold differences between perpendicular axes, respectively. Trelex, ProLiteTM, and ParietexTM were 2.32.4 times. Dualmesh was the least anisotropic mesh, without marked difference between the axes. Conclusion Anisotropy of synthetic meshes has been underappreciated. In this study, we found striking differences between elastic properties of perpendicular axes for most commonly used synthetic meshes. Indiscriminate orientation of anisotropic mesh may adversely affect hernia repairs. Proper labeling of all implants by manufacturers should be mandatory. Understanding the specic anisotropic behavior of synthetic meshes should allow surgeons to employ rational implant orientation to maximize outcomes of hernia repair. Keywords Anisotropy Synthetic Mesh Hernia Biomechanical test

Introduction Synthetic surgical meshes are used routinely in repairing abdominal hernia defects, and their use is correlated with a signicant reduction in hernia recurrence [1, 2]. Since the rst clinical use of synthetic mesh for herniorrhaphy by Usher in 1958, a wide variety of mesh materials have become available [35]. Because of the dynamic nature of the abdominal wall, the textile properties of a mesh implant dictate its behavior in vivo and may be a major contributor to the overall mechanical behavior of the repaired abdominal wall [6]. Since synthetic meshes differ by material polymers, ber densities, mesh designs, and composite materials, various meshes induce different mechanical behavior in the abdominal wall post-repair [7, 8]. In fact, such post-repair mechanics can be a major

123

48

Hernia (2011) 15:4752

factor in the risk for hernia recurrence, patient discomfort, and overall postoperative quality of life [9, 10]. According to Hookes law, a material generates tensile force to resist deformation when it is stretched [11]. If the relationship between tensile force generation and deformation is linear in nature, the material is considered elastic. The degree to which an elastic material is stretched and the tensile force it generates are related to the elastic modulus (E) through the following equation: E = r/e, where r represents stress (tensile force) and e represents strain (deformation). Stiffer materials possess greater elastic moduli. The elastic modulus can either be dependent or independent of the orientation of the material when strained [12]. Materials that exhibit this dependence are considered anisotropic, and those that exhibit independence are isotropic. In other words, anisotropic materials will respond to being stretched differently if stretched in different directions. Interestingly, while modern meshes have been utilized for over 50 years, investigations into the anisotropic properties of synthetic meshes have been essentially absent. The abdominal wall exhibits anisotropic behavior, with approximately half the resistance to deformation generated in the cranio-caudal axis when compared to the lateral axis [13, 14]. In other words, the abdominal wall behaves almost twice as elastic in the vertical direction versus the horizontal direction. When an anisotropic material is implanted in the abdominal wall, its orientation is likely to determine the implants response to physiologic intraabdominal forces. Although the exact clinical implications of mesh anisotropy are yet to be established, Anurov et al. [15] recently demonstrated a signicant difference in the risk for mesh failure and hernia recurrence between repairs conducted with orthogonal orientations of lightweight polypropylene mesh. Despite potential adverse effects of haphazard orientation of mesh implants, little data exist on

the anisotropic behavior of todays synthetic meshes. The aim of this study was to investigate and compare anisotropic properties of common synthetic meshes used in hernia repair.

Materials and methods Surgical meshes Six synthetic meshes were investigated: Trelex (TX; Boston Scientic, Natick, MA, USA) is composed of heavyweight microporous polypropylene; ProLiteTM (PL; Atrium Medical, Hudson, NH, USA) is composed of a midweight microporous polypropylene; UltraproTM (UP; Ethicon, Somerville, NJ, USA) is composed of lightweight macroporous polypropylene and poliglecaprone 25; ParietexTM (PX; US Surgical/Covidien, Norwalk, CT, USA) is a three-dimensional macroporous polyester mesh; Dualmesh (DM; WL Gore, Flagstaff, AZ, USA), is composed of expanded polytetrauoroethylene (ePTFE); Innit (INF; WL Gore) is a knit PTFE mesh. See Table 1 for additional mesh characteristics. Mechanical testing For each mesh, longitudinal and transverse axes were arbitrarily dened. Four samples of each mesh were cut into 50 9 20-mm strips in both the longitudinal and transverse axes, as shown in Fig. 1. The strips were tested in uniaxial tension using a universal testing machine (ADMET, Norwood, MA, USA) congured with a 20 N static load cell (ADMET). The samples were loaded with a 30-mm gauge length, allowing for 10-mm purchase in the upper and lower pneumatic grips. To prevent mesh

Table 1 Synthetic mesh characteristics. PTFE Polytetrauoroethylene, ePTFE expanded polytetrauoroethylene

Mesh product Trelex ProLiteTM UltraproTM ParietexTM Dualmesh

Material Polypropylene Polypropylene Polypropylene with Poliglecaprone Polyester ePTFE

Weightdensity (g/m2) Heavyweight95 Midweight85 Lightweight28 Midweight78 (Solid laminar sheet)

Pore size (mm) and characteristics Microporous0.6 Multiple pore shapes Microporous0.8 Multiple pore shapes Macroporous2.04.0 Diamond-shaped Macroporous1.8 9 1.5 Hexagonal 2-sided: micro- and macroporousa Solid sheet Macroporousa Octagonal

Innit

PTFE

(Data not available)

Pore size measurements not available

123

Hernia (2011) 15:4752

49

portion of each stress versus strain plot was calculated using the ordinary least squares method. The average slope calculated for each axis in each mesh dened the respective elastic modulus (E). The ratio between the elastic moduli in each axis for a given mesh was related to its degree of anisotropy (k) using the following equation:    EL   k log  E 
T

where EL and ET represent the elastic modulus for the longitudinal and transverse axes, respectively. For statistical analysis, longitudinal and horizontal axes were compared using Students T test, with P \ 0.05 considered statistically signicant.

Results All meshes behaved elastically when strained within the physiologic range, shown as linearity in the stress versus strain curve. Figure 3 shows this linear response, the slope of which represents the elastic modulus, for each axis of each mesh. Table 2 shows the values of the elastic modulus calculated in each axis for each mesh, as well as the ratios of the elastic moduli. The meshes exhibited varying degrees of difference in the elastic modulus in each axis. Five of the six meshes tested exhibited a marked degree of anisotropy between the two axes. InnitTM was the most anisotropic (k = 1.29; P \ 0.05) followed by UltraproTM (k = 1.07; P \ 0.05), representing 20- and 12-fold differences in the elastic modulus of perpendicular axes, respectively. As shown in Fig. 4, ParietexTM, ProLiteTM and Trelex demonstrated similar degrees of anisotropy (k = 0.37; P \ 0.05). Dualmesh did not display a signicant degree of anisotropy (k = 0.04, P = no signicance).

Fig. 1 Mesh directionality. In this example, Ultrapro has been cut in longitudinal and transverse orientations

slippage, the pneumatic grips were set to 80 psi. The mesh strips underwent tensile testing with a constant strain rate of 20 mm/min until failure, as shown in Fig. 2. Crosshead displacement (mm) and load cell force (N) data were recorded during testing and analyzed. Anisotropy calculation Stress and strain were calculated and plotted against each other. In the context of this study, stress was expressed as tensile force per unit width of the mesh. Strain was dened as a relative elongation of the mesh. The slope of the linear

Fig. 2 Uniaxial tension testing. An ADMET universal testing machine is shown with an expanded polytetrauoroethylene (ePTFE) sample undergoing tensile testing

123

50

Hernia (2011) 15:4752

Fig. 3 Stress versus strain curves for meshes. Slopes of these curves were used to calculate elastic moduli for each axis. Solid lines Longitudinal axis, dashed lines transverse axis

Table 2 Elastic moduli Mesh type Dualmesh Trelex Parietex

TM

Discussion
ET (N/mm) 9.78 3.31 5.28 2.54 0.87 0.98 L/T Ratio 1.1 2.3 2.3 2.4 11.7 19.7 P-value nsa \0.05 \0.05 \0.05 \0.05 \0.05

EL (N/mm) 10.73 7.69 12.26 5.99 10.21 19.35

ProLiteTM UltraproTM Innit

EL Elastic modulus for longitudinal axis, ET elastic modulus for transverse axis, L/T longitudinal/transverse axes
a

No statistical signicance

A wide range of synthetic meshes are available to todays surgeons, and the textile properties of these meshes may inuence mesh performance in vivo. Both microscopic and macroscopic features contribute to a given materials elastic and mechanical behavior. Microscopic factors include the chemical composition, molecular geometry, and polymer chain interactions. Macroscopic factors include the gross ber geometry, density, and overall mesh pore size. In addition, mesh orientation at implantation may be critical. Since there are no overt markings to dene

123

Hernia (2011) 15:4752

51

Fig. 4 Degree of anisotropy. k Degree of anisotropy, calculated from the log of elastic moduli ratios

mesh axes and no information on anisotropic behavior is included in the instructions for use (IFU), it has been assumed that meshes are isotropic with equal material properties in all directions. In this study, however, we demonstrated that the vast majority of commonly used synthetic meshes exhibit a marked degree of anisotropy. Mesh anisotropy is an important consideration with potential dramatic implications. By denition, the mechanical behavior of an anisotropic mesh in one direction is different than its behavior in another direction [11, 12, 16]. As a result, the orientation of an anisotropic implant will likely have an impact on the mechanical behavior of the repaired abdominal wall. Based on our data, if, for example, ParietexTM mesh is stretched in what we dened as the longitudinal axis 10% of its original length, it will generate 1.2 Newtons (N) of resistance per millimeter width of the sample. Alternatively, the same sample stretched in the transverse axis will generate 0.5 N resistance per millimeter width of the sample. As another example, the elastic modulus ratio for UltraproTM was over 11 times greater in the longitudinal versus horizontal axis. Because elastic moduli reect the stiffness of a mesh while the inverse of the elastic modulus reects elasticity, stretching the mesh parallel to the blue lines will result in more resistance. Conversely, stretching the mesh perpendicular to the blue lines allows greater compliance of the mesh. Differences between axes are even more pronounced for InnitTM meshes. Appreciation of this inequality between axes we believe is paramount to maximizing mesh performance during hernia repair. Two strategies exist when positioning mesh implants. It can be argued that an ideal mesh implant should either maximize or minimize the force generated when strained in vivo, implying either a high or low elastic modulus respectively. A mesh with a high elastic modulus will not easily stretch. Thus, it will provide a strong mechanical reinforcement, but at the cost of increased shear forces between the

mesh and abdominal wall. Increased shear forces are shown to contribute to chronic local trauma leading to increased brotic remodeling and subsequent decreased patient quality of life [1719]. A mesh with a low elastic modulus will not generate high shear forces but will accommodate a greater degree of deformation. It may be reasonable to conclude that, because the abdominal wall behaves almost twice as elastic vertically compared to horizontally [13, 14], greater mesh elasticity is needed in a cranio-caudal direction for midline defect repairs (e.g., orienting UltraproTM with the blue lines horizontally for ventral hernia repairs). Further in vivo studies are needed, however, to dene the most favorable mesh placement from an anisotropic standpoint. Our data show a marked difference between the elastic modulus found in orthogonal axes for most common synthetic meshes. Random orientation of mesh implants with a marked degree of anisotropy conceivably contributes to inconsistent outcomes. Considering that the elastic modulus of a mesh contributes to the efcacy of a repair in some capacity, and that the elastic modulus in each axis for a given mesh varied anywhere from 1 to 20 times, it is very likely that mesh orientation will inuence repair efcacy. This information should be described in each mesh manufacturers IFU. In addition, we strongly believe that mesh directionality should be labeled on the mesh product during manufacture in order to provide surgeons with ability to strategically use the anisotropic behavior of a given mesh during hernia repair. Depending on the specic hernia location and goals of the repair, the surgeon can orient a mesh to provide either stiff or compliant reinforcement in a specic axis. Considering that the abdominal wall is anisotropic, such mindful orientation of mesh implants may be a powerful tool in optimizing post-surgical outcomes. This study does has several limitations. We employed a uniaxial tensile test to measure anisotropy. While this test does elicit differences in the response in different axes of a mesh, it does not precisely simulate physiologic loading. Physiologic loading is multiaxial and can be simulated using elaborate testing techniques [20]. Elaborate simulation of physiologic loading was unnecessary for the current study, which aimed to demonstrate differences between the two perpendicular axes for a given mesh. Uniaxial testing is commonplace in experiments assessing the integrity of hernia repairs, and our results directly illustrate that mesh orientation must be considered in such studies. Ongoing in vivo experiments will likely provide clinical correlations to our ndings.

Conclusion All synthetic meshes exhibit various degrees of anisotropy. Moreover, we found that lightweight polypropylene and

123

52

Hernia (2011) 15:4752 7. Cobb WS, Burns JM, Peindl RD, Carbonell AM, Matthews BD, Kercher KW, Heniford BT (2006) Textile analysis of heavy weight, mid-weight, and light weight polypropylene mesh in a porcine ventral hernia model. J Surg Res 136:17 8. Klosterhalfen B, Klinge U, Schumpelick V (1998) Functional and morphological evaluation of different polypropylene-mesh modications for abdominal wall repair. Biomaterials 19:22352246 9. DuBay DA, Wang X, Adamson B, Kuzon WM Jr, Dennis RG, Franz MG (2006) Mesh incisional herniorrhaphy increases abdominal wall elastic properties: a mechanism for decreased hernia recurrences in comparison with suture repair. Surgery 140:1424 10. Novitsky YW, Harrell AG, Hope WW, Kercher KW, Heniford BT (2007) Meshes in hernia repair. Surg Technol Int 16:123127 11. Beer FP, Johnston ER, DeWolf JT (2006) Mechanics of materials, 4th edn. McGraw Hill, New York 12. Ozkaya N, Nordin M (1999) Fundamentals of Biomechanics: equilibrium, motion, and deformation, 2nd edn. Springer, New York 13. Junge K, Klinge U, Prescher A, Giboni P, Niewiera M, Schumpelick V (2001) Elasticity of the anterior abdominal wall and impact for reparation of incisional hernias using mesh implants. Hernia 5:113118 14. Grassel D, Prescher A, Fitzek S, Keyserlingk DG, Axer H (2005) Anisotropy of human linea alba: a biomechanical study. J Surg Res 124:118125 15. Anurov M, Titkova S, Oettinger A (2009) Effectiveness of experimental hernia repair depends on orientation of mesh implant with anisotropic structure. 4th International Hernia Congress: Joint Meeting of the AHS and EHS, Berlin, Germany, Hernia 13(Suppl 1)1:S1415 16. Callister WD (2005) Materials science and engineering: an integrated approach, 2nd edn. Wiley, Hoboken 17. Cobb WS, Kercher KW, Heniford BT (2005) The argument for lightweight polypropylene mesh in hernia repair. Surg Innov 12:6369 18. Hollinsky C, Hollinsky KH (1999) Static calculations for mesh xation by intraabdominal pressure in laparoscopic extraperitoneal herniorrhaphy. Surg Laparosc Endosc Percutan Tech 9:106109 19. Majercik S, Tsikitis V, Iannitti DA (2006) Strength of tissue attachment to mesh after ventral hernia repair with synthetic composite mesh in a porcine model. Surg Endosc 20:16711674 20. Klinge U, Klosterhalfen B, Conze J, Limberg W, Obolenski B, Ottinger AP, Schumpelick V (1998) Modied mesh for hernia repair that is adapted to the physiology of the abdominal wall. Eur J Surg 164:951960

polytetrauoroethylene meshes exhibited approximately 12- and 20-fold differences between orthogonal axes, respectively. Although the exact clinical implications of our ndings are yet to be proven, we believe that the mechanical behavior of implanted synthetic meshes may be dependent on the mesh orientation during hernia repair. In light of our ndings, we urge that anisotropic behavior for all surgical mesh products should be qualied and included in the products IFU. In addition, mesh axes should be marked as part of the manufacturing process in order to ensure consistent implant orientation. Thorough knowledge and understanding of the anisotropic behavior of both synthetic mesh and abdominal wall is paramount to allow surgeons to strategically orient the implant to optimize post-surgical outcomes.
Acknowledgment This study was funded by institutional support from the University of Connecticut Health Center. Conicts of interest Dr. Y. Novitsky has received consulting and/ or speaking fees from Ethicon Inc., W.L. Gore Inc., and Covidien. Mr. E. Saberski and Dr. S. Orenstein have no conicts of interest or nancial ties to disclose for this study.

References
1. Cassar K, Munro A (2002) Surgical treatment of incisional hernia. Br J Surg 89:534545 2. McCormack K, Scott NW, Go PM, Ross S, Grant AM (2003) Laparoscopic techniques versus open techniques for inguinal hernia repair. Cochrane Database Syst Rev:CD001785 3. Usher FC, Ochsner J, Tuttle LL Jr (1958) Use of marlex mesh in the repair of incisional hernias. Am Surg 24:969974 4. Bachman S, Ramshaw B (2008) Prosthetic material in ventral hernia repair: how do I choose? Surg Clin North Am 88:101112 ix 5. Udwadia T (2006) Inguinal hernia repair: the total picture. J Min Access Surg 2:144146 6. Welty G, Klinge U, Klosterhalfen B, Kasperk R, Schumpelick V (2001) Functional impairment and complaints following incisional hernia repair with different polypropylene meshes. Hernia 5:142147

123