- Email: [email protected]
UHMWPE fiber constructs
Journal of Biomechanics 46 (2013) 1463–1470
Contents lists available at SciVerse ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Hydrogel fibers for ACL prosthesis: Design and mechanical evaluation of PVA and PVA/UHMWPE fiber constructs Jason S. Bach a,1, Fabrice Detrez b,2, Mohammed Cherkaoui c,3, Sabine Cantournet b,4, David N. Ku a,n, Laurent Corte´ b,5 a
George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, 30332-0405, USA Centre des Mate´riaux, Mines ParisTech, CNRS UMR 7633, BP 87, F-91003 Evry Cedex, France c George W. Woodruff School of Mechanical Engineering, Georgia Tech Lorraine, 2 Rue Marconi, 57070 Metz, France b
a r t i c l e i n f o
abstract
Article history: Accepted 26 February 2013
Prosthetic devices for anterior cruciate ligament (ACL) reconstruction have been unsuccessful due to mechanical failure or chronic inflammation. Polymer hydrogels combine biocompatibility and unique low friction properties; however, their prior use for ligament reconstruction has been restricted to coatings due to insufficient tensile mechanics. Here, we investigate new constructs of polyvinyl alcohol (PVA) hydrogel fibers. In water, these fibers swell to an equilibrium water content of 50% by weight, retaining a tensile modulus greater than 40 MPa along the fiber axis at low strain. Rope constructs were assembled for ACL replacement and mechanical properties were compared with data from the literature. Pure PVA hydrogel constructs closely reproduce the non-linear tensile stiffness of the native ACL with an ultimate strength of about 2000 N. An additional safety factor in tensile strength was achieved with composite braids by adding ultrahigh molecular weight polyethylene (UHMWPE) fibers around a core of PVA cords. Composition and braiding angle are adjusted to produce a non-linear tensile behavior within the range of the native ligament that can be predicted by a simple rope model. This design was found to sustain over one million cycles between 50 and 450 N with limited damage and less than 20% creep. The promising mechanical performances of these systems provide justification for more extensive in vivo evaluation. & 2013 Published by Elsevier Ltd.
Keywords: Ligament Anterior cruciate ligament Artificial ligament Hydrogel fibers Polyvinyl alcohol Ultrahigh molecular weight polyethylene
1. Introduction Anterior cruciate ligament (ACL) tears do not heal on their own nor respond well to primary repair due to a poor blood supply (Bray et al., 2002). Tendon autografts are currently the standard for ACL replacement (Woo and Adams, 1990) yielding good results. Drawbacks such as donor site morbidity, longer recovery times, and the requirement of the harvest operation with autografts would make an off-the-shelf prosthetic attractive for surgeons and patients (Kartus et al., 2001). Beginning in the
n
Corresponding author. Tel.: þ404 894 6827; fax: þ 404 894 2291. E-mail addresses: [email protected] (J.S. Bach), [email protected] (F. Detrez), [email protected] (M. Cherkaoui), [email protected] (S. Cantournet), [email protected] (D.N. Ku), [email protected] (L. Corte´). 1 Tel.: þ1 404 894 6827; fax: þ1 404 894 2291. 2 Tel.: þ33 1 60 76 30 70; fax: þ33 1 60 76 31 50. 3 Tel.: þ33 3 87 20 3939; fax: þ 33 3 87 20 3940. 4 Tel.: þ33 1 60 76 30 52; fax: þ 33 1 60 76 31 50. 5 Tel.: þ33 1 60 76 31 42; fax: þ 33 1 60 76 31 50. 0021-9290/$ - see front matter & 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jbiomech.2013.02.020
1970s, multiple prosthetics have been developed; however, due to high rates of mechanical failure and immunogenic particulation, the field has turned away from these devices (Mascarenhas and MacDonald, 2008). As surgical techniques and tools have since improved and a wider range of biomaterials have become available, the question of whether a functional prosthetic can be developed is worth exploring again. Polymer hydrogels, consisting of water-swollen networks of cross-linked macromolecules are particularly relevant materials as they are similar to the extra-cellular matrix of native tissues. They have shown promising results as biodegradable scaffolds for tissue-engineered ligaments (Drury and Mooney, 2003). Nevertheless, achieving proper mechanical performance with tissueengineered systems during scaffold degradation and in newly formed tissue remains a challenge. Non-degradable hydrogels could bring a more immediate solution for ligament implants. In particular, polyvinyl alcohol (PVA) can be processed to form a non-degradable hydrogel with excellent biocompatibility (Hassan and Peppas, 2000a). These materials exhibit a unique selflubrication property by forming a thin water layer at the contact interface resulting in low surface friction (Gong, 2006; Mamada et al., 2011). As a result, PVA has shown potential in several
8 mm
o 20% elongation after 1 M cycles from 50 N to 450 N NA
10 mm
Functional stiffness maintained (140 N/% strain) NA
Measurement of length during of fatigue testing
Ruler (Caborn et al., 1998; Duthon et al., 2006)
soft-tissue replacement applications including knee cartilage and vein valve repair (Hassan and Peppas, 2000b; Noguchi et al., 1991; Sathe and Ku, 2007). In this article, we investigate how PVA hydrogels can be used to produce artificial ligaments having mechanical performance suitable for ACL replacement. For ligament implants, hydrogels are usually limited by their insufficient tensile properties and are used as the coating of a stiffer fibrous polymeric construct (Freeman et al., 2011). Nevertheless, the highly oriented organization of the collagen matrix of ligaments suggests that appropriate tensile properties might be attained with synthetic hydrogels having a strong macromolecular orientation along the ligament axis. Such an orientation can be achieved in PVA using fiber forming processes including solutionspinning (Yamaura and Kamakura, 2000), melt-spinning (Wu et al., 2012), or electrospinning (Yao et al., 2003). Water swelling to form hydrogel fibers from spun fibers may release macromolecular orientation and thus prevent access to sufficient tensile stiffness. Here, we show that appropriate tensile properties can be achieved from PVA hydrogel fibers having a water content of 50% by weight, comparable to that of native ligament tissue. In the first part of this study, prosthetic ACLs were assembled from PVA only and compared to the mechanical design inputs obtained from the literature and outlined in Table 1. Particular attention was paid to the proper reproduction of the non-linear tensile response of native ligaments. A periodic crimp pattern of collagen fibrils confers an elastic response to the ligaments with a low stiffness toe regime at low strain and a higher stiffness linear regime at greater strain (Woo et al., 1999). This low stiffness toe likely helps prevent creep and fatigue (Freeman and Kwansa, 2008; Laurencin and Freeman, 2005). A toe region has been integrated in several scaffolds for tissue engineered ACL (Freeman et al., 2007; Gentleman et al., 2003; Horan et al., 2006; Sahoo et al., 2006). In the second part, we explore how high-strength fibers of ultrahigh molecular weight polyethylene (UHMWPE) may be added to provide a safety factor in tensile strength (Cook et al., 2010; Kurtz, 2009). In such composite ropes using two types of fibers, complex structures are required to reproduce the nonlinear tensile behavior of ligaments (Hopper et al., 1995). A hollowbraided design is proposed consisting of a core of PVA hydrogel fibers surrounded by a braided layer of UHMWPE threads. With this structure, the low strain response is governed by the core while the outside layer provides stiffening at larger strains. The relationship between rope composition, geometrical parameters and tensile response is investigated using both experimental measurements and a simple rope model. Based on those results, we discuss the potential of both pure PVA hydrogel fiber constructs or PVA/UHMWPE constructs for ACL replacement.
2. Materials and methods 2.1. Materials PVA that forms a stable insoluble hydrogel below 90 1C was purchased in the form of 675-dtex threads composed of 15 twisted continuous fibers. This form of PVA is over 95% hydrolyzed, similar to other biocompatible PVA grades used for implant applications (Baker et al., 2012). PVA cords of 15 twisted threads were assembled by a textile manufacturer (Morel-Journel, Lyon, France). UHMWPE was purchased in the form of 1200-dtex threads composed of 4 braided bundles of 60 continuous fibers.
2.2. Fabrication of fiber constructs Diameter
Permanent elongation
(FDA, 1993)
Functional stiffness maintained for 41 M load cycles from 50 N to 450 N o 20% permanent elongation after 1 M cycles from 50 N to 450 N o 10 mm Fatigue
Fatigue testing
4 30% 4 12% Uniaxial tensile test
4 40%
6–15% 2–15% Uniaxial tensile test
5–15%
44000 N 4 2000 N 41730 N
Ultimate Load Toe Transition Strain Ultimate Strain
Uniaxial tensile test
25–60 N/% strain 35–400 N/% strain Uniaxial tensile test Stiffness
(Hamner et al., 1999; Noyes et al., 1984; Yagi et al., 2002) (Brown, 1995; Hamner et al., 1999; Noyes et al., 1984; Yagi et al., 2002) (Butler et al., 1986; Grood and Noyes, 1976; Noyes et al., 1984; Noyes and Grood, 1976; Woo and Adams, 1990; Woo et al., 1991) (Butler et al., 1986; Grood and Noyes, 1976; Noyes et al., 1984; Noyes and Grood, 1976; Woo and Adams, 1990; Woo et al., 1991) (FDA, 1993)
PVA construct (Braided 20 PVA cords) Target value Verification test Ref. Design input
Table 1 Design inputs for a prosthetic ACL, verification test results and characteristics of 20 braided PVA cords and hollow-braid constructs with 16 PVA hydrogel cords and 12 UHMWPE threads.
60–180 N/% strain
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
PVA/UHMWPE construct (Hollow-braid/16 PVA cords/12 UHMWPE threads/ h ¼451)
1464
Full-scale constructs for ACL replacement were fabricated manually. Both twisted and diamond braided PVA-only constructs were produced. Twisted samples were obtained from 16 PVA cords by first twisting 4 groups of 4 cords together with a left-hand twist and then, twisting these 4 ropes into a larger one
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470 with a right-hand twist. For the diamond braided samples, 20 PVA cords were braided together using a 1/1 overlap-underlap technique forming a cylinder (Brunnschweiler, 1954). For PVA/UHMWPE constructs, a hollow-braid design was used in which PVA cords were aligned in parallel forming a core, which was then surrounded with a tubular diamond braid of 12 UHMWPE threads. After assembly, the diamond braid was left free to slide at one end to allow swelling of the PVA core after which the braid was tightened around the core and fixed at both ends. A first series was obtained by varying the number of PVA cords in the core to 8, 12, and 16. In a second series, the braid angle of the tubular braid was modified to 407 21, 45 721, 507 21, and 557 21.
1465
80µm
100µm 1/2turn ~ 4mm
2.3. Tensile testing Tensile testing was performed at room temperature in air on wet samples using an Instron 5966 apparatus. Samples were immersed in distilled water for at least 12 h and tested less than 10 min after removal from water to ensure negligible drying. Pull-to-failure tests were performed using capstan fixations with a diameter of 7 cm and sub-failure tests, using large wedge clamps. Full scale samples were stretched for 20 cycles between 0 N and 700 N at a rate of 1 mm per second (about 2% s 1), corresponding to the maximum force during strenuous activity at a medium to slower rate in vivo (Chen and Black, 1980; Noyes et al., 1974, 1984; Pioletti et al., 1999). The tensile curves corresponding to the last loading cycle are presented. Strain was measured using a video extensometer at a rate of 2 images per second. Each test was repeated on at least 3 samples. 2.4. Fatigue testing Samples were first immersed in distilled water for at least 12 h. Fatigue tensile testing was performed on a custom apparatus with an immersion chamber. Sample length was 15 cm including a gauge length of 5 cm between two 5 cm long sections for gripping at each end. Samples were stretched to 500 N for 10 min prior to tightening in this stretched state. This provided a stronger grip and reduced deformation and wear at the clamps. After unloading back to zero, samples were completely submerged in water at room temperature and left to swell for 1 h prior to testing. Fatigue testing simulated a normal non-strenuous activity, operating in force-controlled mode between 50 N and 450 N at 1 Hz, while constantly recording force and displacement data at 100 points per second (Noyes et al., 1984).
3. Results
500µm 1/2turn ~ 1cm
1mm
UHMWPE thread
3mm
PVA cord
Fig. 1. Observations of constitutive fibers and threads: (a) dry PVA fibers (SEM observation), (b) dry PVA thread, (c) dry PVA cord, (d) UHMWPE thread and PVA cord.
3.1. Mechanical evaluation of PVA hydrogel cords Dry PVA fibers, threads, and cords were characterized by electronic and optical microscopy as shown in Fig. 1. SEM observations (Fig. 1a) show that PVA fibers have a roughly circular cross-section with a diameter of 80 71 mm. They are assembled into threads of 15 fibers with a left-hand twist and an 8 mm pitch (Fig. 1b). Threads were themselves assembled into cords of 15 threads with a right-hand twist and a 2 cm pitch (Fig. 1c). Upon immersion in water or bodily fluids, PVA fibers swell becoming hydrogel fibers. Here, PVA is fully hydrolyzed. Its crosslinking is physical and results from the formation of small crystallites that melt in water above 90 1C (Hassan and Peppas, 2000a). Below this melting temperature, no degradation is expected. A stability study confirms that the hydrogel fibers are insoluble at 40 1C and show no change in weight after 3 years of water immersion. Optical microscopic images in Fig. 2a show the swelling of one single fiber. Corresponding diameter measurements in Fig. 2b show that after approximately 10 min, an equilibrium water content of 50% by weight is reached and the diameter increases to 110710 mm. The diameter of swollen threads and cords are 0.5 70.1 mm and 370.5 mm, respectively. The tensile properties of swollen PVA hydrogel cords are shown in Fig. 3. These measurements integrate both the material intrinsic behavior and the deformation of the twisted cord structures. Swollen cords exhibit the characteristic rubbery behavior of PVA hydrogels. They can elastically sustain large strains over 30% with little dissipation as shown by loading–unloading curves in Fig. 3a. They exhibit a non-linear response with a
stiffness of 0.8 N/% strain at low strain and over 2 N/% strain above 10%. Their ultimate strain and load are of the order of 3773% and 100720N. 3.2. PVA hydrogel fiber constructs for ACL replacement From these results, a first design was proposed for PVA-based constructs exhibiting the stiffness and strength adequate for ACL replacement while remaining within size constraints. About 250 to 300 PVA threads, i.e. 3500–4500 fibers, were necessary. Manipulation of fewer cords was preferred and 15 to 20 PVA cords were required to provide the minimal stiffness requirement of 35 N/%strain as reported in the literature (Table 1). Cord assembly was performed manually with two patterns: 16 twisted PVA cords (Fig. 4a) and 20 braided PVA cords (Fig. 4b). All tested constructs were assembled with dry PVA cords and were submerged in distilled water at room temperature for at least 12 h prior to testing. Both designs were 10 mm in diameter, the maximum allowable before exceeding the size constraints and slightly larger than the 8 mm of an average hamstring graft (Caborn et al., 1998). Tensile curves up to failure are given in Fig. 5 and compared to the range of behavior of native ACL as estimated from force– displacement data by Woo et al. considering a ligament length between 20 and 40 mm (Caborn et al., 1998; Woo et al., 1991). Twisted 16 PVA cords have stiffness close to the lower bound of normal elasticity and strength of 1800 7100 N, slightly inferior to the native ACL. Braided 20 PVA cords provide a non-linear
1466
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Fig. 2. Characterization of fiber swelling. (a) Single fiber observation by optical microscopy and (b) time resolved measurement of fiber diameter during swelling.
response within the range for native ACL with an ultimate load of 2200 7100 N, equivalent to the average ACL in the young adult (Woo et al., 1991). Nevertheless, at low strain ( o5%), the braided 20 PVA cords provide a rather high stiffness of about 25 N/% strain instead of the low stiffness toe of native ACL as shown in Fig. 6. 3.3. Hollow-braided constructs of PVA/UHMWPE fibers A safety factor in terms of ultimate load would be advantageous and is achievable by adding UHMWPE fibers of 1200dtex shown in Fig. 1d. An UHMWPE thread has a high stiffness close to 100 N/% strain and an ultimate load of 375 N as indicated in Fig. 3. The coarse assembly of 12 UHMWPE threads twisted with 16 PVA cords (Fig. 4c) provides an ultimate load twice that of the native ACL; however, it results in a very high stiffness when the response is dominated by the UHMWPE threads under tension as indicated in Fig. 5. More complex fiber assemblies should allow retention of high tensile strength while reducing stiffness closer to the level of native ACL to avoid over-constraint of the knee (Suggs et al., 2003). For PVA/UHMWPE assemblies, a hollow-braided construct was designed where a core of parallel PVA cords is wrapped by a diamond braided structure of UHMWPE threads as shown in Fig. 4d. In this construct, the tensile response at low strain is governed by the PVA core. As strain increases, the diamond braided structure deforms and compresses the core. UHMWPE threads are progressively put under tension and gradually stiffen the response. Two design parameters were modified to adjust the shape of the tensile response: the number of PVA cords in the core and the braid angle, y, between the UHMWPE thread axis and the ligament axis, as illustrated in Fig. 4d.
Fig. 3. Tensile force versus strain for PVA hydrogel cords. (a) Four loadingunloading cycles (first cycle not shown), (b) loading to failure curves for PVA hydrogel cords (3 samples) and UHMWPE thread. The end of the curves corresponds to failure.
Fig. 4. Full-scale constructs after equilibration in water for ACL replacement. (a) Sixteen twisted PVA cords, (b) 20 braided PVA cords, (c) 16 twisted PVA cords and 12 UHMWPE threads and (d) Hollow-braided assembly of 16 PVA cords and 12 UHMWPE threads.
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
1467
Fig. 5. Tensile forceversus strain for 16 twisted PVA hydrogel cords (blue), 20 braided PVA hydrogel cords (green) and a mixed construct of 16 PVA hydrogel cords twisted along with 12 UHMWPE threads (red). Curves are compared with the native ACL and tendons used for autografts shaded in gray (Duthon et al., 2006; Kim et al., 2003; Noyes et al., 1984; Woo et al., 1991). The ends of the curves correspond to failure. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 7. Role of geometrical parameters in hollow-braided constructs: (a) Force versus strain for samples with 8, 12, and 16 PVA hydrogel cords in the core and braid angle of 451. Inset: close-up on the low strain toe region. Symbols are experimental data and thin lines show model predictions and (b) force versus strain for samples with 16 PVA hydrogel cords in the core and braid angles of 401, 451, 501, and 551. Thin lines show model predictions.
In a simple model assuming incompressibility of the core, the composition, initial braid angle (y0), and mechanical stiffness of PVA cords and UHMWPE threads are utilized as inputs to predict the tensile behavior of the structure (see supporting information). At low strains, the tensile behavior is given by the response of the core while the contribution of the outer braid is added above a critical elongation l* which corresponds to the positive root of equation Fig. 6. Tensile force versus strain (Close-up on 0–700 N range) for 16 twisted PVA hydrogel cords (blue), 20 braided PVA hydrogel cords (green) and hollow-braided construct (HB) with a core of 16 PVA hydrogel cords and an outer braid of 12 UHMWPE threads with y ¼ 451 (red). Shaded regions show typical range for the native ACL and tendons used for autografts (Duthon et al., 2006; Kim et al., 2003; Noyes et al., 1984; Woo et al., 1991). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 7a shows the behavior of hollow-braids with 12 UHMWPE threads braided at y ¼451 and cores composed of 8, 12, and 16 PVA cords with diameters after swelling of 5.5 mm, 7 mm, and 8 mm, respectively. Increasing the number of PVA cords stiffens the toe and smoothes the transition from lower to higher stiffness. Fig. 7b shows the force–strain curves for hollow-braided samples with 16 PVA cords in the core and 12 UHMWPE threads braided at y ¼40 721, 45721, 50 721, and 55721. Increasing the braid angle implies that UHMWPE threads are recruited at greater tensile strains, therefore increasing the size of the low stiffness toe and decreasing the overall stiffness of the constructs. The best match to the native ACL is provided with a braid angle between 401 and 501 and is reproduced in Fig. 6 for y ¼451.
l3 lð1 þ ðtany0 Þ2 Þ þðatany0 Þ2 ¼ 0 The only fitting parameter, the ratio (a) of the initial braid radius to the initial core radius, is estimated to be 1.15, which is a realistic value given the assembly process. Stiffness of the PVA cords and UHMWPE threads is taken to be 0.8 N/% strain and 100 N/% strain, respectively. The obtained predictions are in fairly good agreement with the experimental data shown in Fig. 7a and b. Dependence of the toe stiffness and on the core size and of the toe size on the initial braid angle are well captured. Predictions indicate that braid angles lower than 501 are required to obtain significant stiffening from the UHMWPE outer braid in the 0–30% strain range. Resistance to fatigue was evaluated on a set of three hollowbraided samples made of 16 PVA cords and 12 UHMWPE threads (y ¼451). These constructs were able to withstand cyclic loading over the physiological tensile force range of 50–450 N up to 1 million caycles. Loading–unloading loops in Fig. 8a show that the stiffness remained within the acceptable range and slightly increased over the course of loading due to bedding in of the UHMWPE fibers into the PVA core. From the minimum strain in
1468
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Novakovic et al., 2004). Though the proposed design is slightly inferior in strength compared to the Gore-tex graft, which had an ultimate load of 5300 N, rupture was not a prominent failure mode for that device (Mascarenhas and MacDonald, 2008; Moyen and Lerat, 1994; Vunjak-Novakovic et al., 2004). The addition of UHMWPE fibers may raise again the issue of particulation of dry polymer fibers for ACL reconstruction (Guidoin et al., 2000). Nevertheless, the number of friction points is greatly reduced in the PVA/UHMWPE hollow-braid structure as compared to full UHMWPE constructs. Several approaches are being investigated to alleviate this possible limitation, including the embedding of UHMWPE fibers in a hydrogel matrix and the use of parallel helices of UHMWPE instead of diamond braid to suppress thread-thread contacts. Our fatigue study was limited to a maximum load of 450N, representative of normal activities; however, loads as high as 700N may occur in vivo (Chen and Black, 1980; Noyes et al., 1984). An ongoing study involves loading at higher and lower magnitudes continuing until rupture. Nevertheless, for ligament replacement, in vitro tests are still unsatisfactory to predict longterm in vivo use (Yahia et al., 1996). Major deviations between in vitro predictions and the observed clinical behavior have been reported on Gore-tex ligaments and bring to light the importance of long-term in vivo studies in the next stage of device development (Bolton and Bruchman, 1985).
Fig. 8. Fatigue study of a hollow-braided sample in water (outside braid: 12 UHMWPE threads with y ¼ 451, core: 16PVA cords). (a) Load versus deformation curves over the course of cyclic loading and (b) minimum and maximum strains in one loading cycle as a function of the number of cycles.
each cycle, an upper bound for permanent elongation of after 1 million cycles was about 20% (Fig. 8b).
5. Conclusion This mechanical evaluation shows that PVA hydrogel fibers may be the basis for novel ACL replacements that would exhibit the biocompatibility and low wear properties of hydrogels as well as an ACL-like tensile response. We find that the addition of high strength fibers and a hollow-braid design provide ways to modulate the performances of such constructs while satisfying the mechanical requirements for a prosthetic anterior cruciate ligament.
4. Discussion These results show that hydrogel fiber constructs can closely reproduce the tensile behavior of native ACL (see Fig. 6 and Table 1). Ultimate strength remains a limitation for pure hydrogel fiber constructs with a diameter below 10 mm. Application of previously developed fiber assemblies for ACL scaffolds (Freeman et al., 2007; Gentleman et al., 2003; Horan et al., 2006; Sahoo et al., 2006) or PVA modification techniques (Choi et al., 2012; Wu et al., 2012) may improve the performances of these systems. A straight-forward solution to gain a safety factor in tensile strength is to introduce high modulus fibers of UHMWPE. Using a hollow-braid design, PVA/UHMWPE constructs provided a controlled toe region, which could be matched to physiological needs by adjusting the core size and braid angle. 12 UHMWPE threads wound at a braid angle from 401 to 501 around a core of 16 PVA cords in parallel fulfilled the requirements for stiffness, ultimate strain, toe region effect, and resistance to fatigue and permanent elongation (see Table 1). The ultimate load of the proposed design with UHMWPE is greater than scaffolds used for tissue engineered ligaments such as those made of polylactide-co-glycolide with an ultimate load from 525 to 907 N and those of silk with ultimate loads as high as 2337 N (Altman et al., 2002; Cooper et al., 2005). It is also greater than other previous permanent prosthetic devices that were susceptible to rupture as a failure mode such as the carbon fiber, Leeds-Keio, Dacron, and ABC prosthetic ligaments with ultimate loads ranging from 660 to 3631 N.(Mascarenhas and MacDonald, 2008; Moyen and Lerat, 1994; Petrigliano et al., 2006; Vunjak-
Conflict of interest statement The authors report they have applied for an International Patent on the ACL prosthesis. There has been no significant financial support for this work that could have influenced its outcome.
Acknowledgments The authors would like to thank Yann Auriac and JeanChristophe Tesseidre for technical support on mechanical testing, for the video extensometry software. Funding from the French National Research Agency (ANR) through Emergence-TEC program (ANR-08-ETEC-003) and Georgia Tech-Lorraine is gratefully acknowledged.
Appendix A See Fig. A1
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jbiomech. 2013.02.020.
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Fig. A1. (a) Schematic representation of a hollow-braid construct, (b) evolution of the core radius (full line) and outer braid radius (dashed line) as a function of stretching ratio. The cross-over gives the locking ratio when the outer braid comes into contact with the core. (Rb0/Rc0 ¼ 1.08, y0 ¼451) and (c) stretching ratio at locking as a function of braid angle (Rb0/Rc0 ¼ 1.08).
References Altman, G.H., Horan, R.L., Lu, H.H., Moreau, J., Martin, I., Richmond, J.C., Kaplan, D.L., 2002. Silk matrix for tissue engineered anterior cruciate ligaments. Biomaterials 23 (20), 4131–4141. Baker, M.I., Walsh, S.P., Schwartz, Z., Boyan, B.D., 2012. A review of polyvinyl alcohol and its uses in cartilage and orthopedic applications. Journal of Biomedical Materials Research Part B: Applied Biomaterials 100B (5), 1451–1457. Bolton, C.W., Bruchman, W.C., 1985. The gore-tex expanded polytetrafluoroethylene prosthetic ligament. An in vitro and in vivo evaluation. Clinical Orthopaedics and Related Research 196, 202–213. Bray, R.C., Leonard, C.A., Salo, P.T., 2002. Vascular physiology and long-term healing of partial ligament tears. Journal of Orthopaedic Research 20 (5), 984–989. Brown, C. H., 1995. Biomechanics of the Semitendinosus and Gracilus Tendon Grafts. In: Procedings of the American Orthopaedic Society for Sports Medicine Conference, Toronto, Canada. Brunnschweiler, D., 1954. 5-The structure and tensile properties of braids. Journal of the Textile Institute Transactions 45 (1), T55–T77. Butler, D.L., Kay, M.D., Stouffer, D.C., 1986. Comparison of material properties in fascicle-bone units from human patellar tendon and knee ligaments. Journal of Biomechanics 19 (6), 425–432. Caborn, D.N.M., Coen, M., Neef, R., Hamilton, D., Nyland, J., Johnson, D.L., 1998. Quadrupled semitendinosus-gracilis autograft fixation in the femoral tunnel: A comparison between a metal and a bioabsorbable interference screw. Arthroscopy: The Journal of Arthroscopic and Related Surgery 14 (3), 241–245. Chen, E.H., Black, J., 1980. Materials design analysis of the prosthetic anterior cruciate ligament. Journal of Biomedical Materials Research 14 (5), 567–586. Choi, J., Kung, H.J., Macias, C.E., Muratoglu, O.K., 2012. Highly lubricious poly(vinyl alcohol)-poly(acrylic acid) hydrogels. Journal of Biomedical Materials Research Part B: Applied Biomaterials 100 (2), 524–532. Cook, J.L., Luther, J.K., Beetem, J., Karnes, J., Cook, C.R., 2010. Clinical comparison of a novel extracapsular stabilization procedure and tibial plateau leveling
1469
osteotomy for treatment of cranial cruciate ligament deficiency in dogs. Veterinary Surgery 39 (3), 315–323. Cooper, J.A., Lu, H.H., Ko, F.K., Freeman, J.W., Laurencin, C.T., 2005. Fiber-based tissue-engineered scaffold for ligament replacement: design considerations and in vitro evaluation. Biomaterials 26 (13), 1523–1532. Drury, J.L., Mooney, D.J., 2003. Hydrogels for tissue engineering: Scaffold design variables and applications. Biomaterials 24 (24), 4337–4351. Duthon, V.B., Barea, C., Abrassart, S., Fasel, J.H., Fritschy, D., Me´ne´trey, J., 2006. Anatomy of the anterior cruciate ligament. Knee Surgery, Sports Traumatology, Arthroscopy 14 (3), 204–213. Freeman, J.W., Kwansa, A.L., 2008. Recent advancements in ligament tissue engineering: The use of various techniques and materials for ACL repair. Recent Patents on Biomedical Engineering 1, 18–23. Freeman, J.W., Woods, M.D., Cromer, D.A., Ekwueme, E.C., Andric, T., Atiemo, E.A., Bijoux, C.H., Laurencin, C.T., 2011. Evaluation of a hydrogel-fiber composite for ACL tissue engineering. Journal of Biomechanics 44 (4), 694–699. Freeman, J.W., Woods, M.D., Laurencin, C.T., 2007. Tissue engineering of the anterior cruciate ligament using a braid-twist scaffold design. Journal of Biomechanics 40 (9), 2029–2036. Gentleman, E., Lay, A.N., Dickerson, D.A., Nauman, E.A., Livesay, G.A., Dee, K.C., 2003. Mechanical characterization of collagen fibers and scaffolds for tissue engineering. Biomaterials 24 (21), 3805–3813. Gong, J.P., 2006. Friction and lubrication of hydrogels—its richness and complexity. Soft Matter 2 (7), 544–552. Guidoin, M.F., Marios, Y., Bejui, J., Poddevin, N., King, M.W., Guidoin, R., 2000. Analysis of retrieved polymer fiber based replacements for the ACL. Biomaterials 21 (23), 2461–2474. Hassan, C.M., Peppas, N.A., 2000a. Structure and applications of poly(vinyl alcohol) hydrogels produced by conventional crosslinking or by freezing/thawing methods. Advances in Polymer Science 153, 37–66. Hassan, C.M., Peppas, N.A., 2000b. Structure and morphology of freeze/thawed PVA hydrogels. Macromolecules 33 (7), 2472–2479. Hamner, D.L., Brown, C.H., Steiner, M.E., Hecker, A.T., Hayes, W.C., 1999. Hamstring tendon grafts for reconstruction of the anterior cruciate ligament: biomechanical evaluation of the use of multiple strands and tensioning techniques. Journal of Bone and Joint Surgery, American Volume 81 (4), 549–557. Hopper, R.H., Grant, J.W., Popper, P., 1995. Mechanics of a hybrid circular braid with an elastic core. Textile Research Journal 65 (12), 709–722. Horan, R.L., Collette, A.L., Lee, C., Antle, K., Chen, J., Altman, G.H., 2006. Yarn design for functional tissue engineering. Journal of Biomechanics 39 (12), 2232–2240. Kartus, J., Movin, T., Karlsson, J., 2001. Donor-site morbidity and anterior knee problems after anterior cruciate ligament reconstruction using autografts. Arthroscopy: The Journal of Arthroscopic and Related Surgery 17 (9), 971–980. Kim, D.H., Wilson, D.R., Hecker, A.T., Jung, T.M., Brown, C.H., 2003. Twisting and braiding reduces the tensile strength and stiffness of human hamstring tendon grafts used for anterior cruciate ligament reconstruction. The American Journal of Sports Medicine 31 (6), 861. Kurtz, S.M., 2009. UHMWPE Biomaterials Handbook: Ultra High Molecular Weight Polyethylene in Total Joint Replacement and Medical Devices. Academic Press, London. Laurencin, C.T., Freeman, J.W., 2005. Ligament tissue engineering: An evolutionary materials science approach. Biomaterials 26 (36), 7530–7536. Mamada, K., Fridrici, V., Kosukegawa, H., Kapsa, P., Ohta, M., 2011. Friction properties of poly(vinyl alcohol) hydrogel: Effects of degree of polymerization and saponification value. Tribology Letters 42 (2), 241–251. Mascarenhas, R., MacDonald, P.B., 2008. Anterior cruciate ligament reconstruction: A look at prosthetics—past, present and possible future. McGill Journal of Medicine 11 (1), 29–37. Moyen, B., Lerat, J.-L., 1994. Artificial ligaments for anterior cruciate ligament replacement. Journal of Bone and Joint Surgery 76-B (2), 173–175. Noguchi, T., Yamamuro, T., Oka, M., Kumar, P., Kotoura, Y., Hyon, S.-H., Ikada, Y., 1991. Poly(vinyl alcohol) hydrogel as an artificial articular cartilage: Evaluation of biocompatibility. Journal of Applied Biomaterials 2 (2), 101–107. Noyes, F.R., Butler, D.L., Grood, E.S., Zernicke, R.F., Hefzy, M.S., 1984. Biomechanical analysis of human ligament grafts used in knee-ligament repairs and reconstructions. Journal of Bone and Joint Surgery, American 66 (3), 344–352. Noyes, F.R., DeLucas, J.L., Torvik, P.J., 1974. Biomechanics of anterior cruciate ligament failure: An analysis of strain-rate sensitivity and mechanisms of failure in primates. Journal of Bone and Joint Surgery 56-A (2), 236–253. Noyes, F.R., Grood, E.S., 1976. The strength of the anterior cruciate ligament in humans and rhesus monkeys. Journal of Bone and Joint Surgery, American Volume 58 (8), 1074–1082. Petrigliano, F.A., McAllister, D.R., Wu, B.M., 2006. Tissue engineering for anterior cruciate ligament reconstruction: A review of current strategies. Arthroscopy: The Journal of Arthroscopic and Related Surgery 22 (4), 441–451. Pioletti, D.P., Rakotomanana, L.R., Leyvraz, P.-F., 1999. Strain rate effect on the mechanical behavior of the anterior cruciate ligament-bone complex. Medical Engineering & Physics 21 (2), 95–100. Sahoo, S., Ouyang, H., Goh, J.C.-H., Tay, T.E., Toh, S.L., 2006. Characterization of a novel polymeric scaffold for potential application in tendon/ligament. Tissue Engineering. Tissue Engineering 12 (1), 1–9. Sathe, R.D., Ku, D.N., 2007. Flexible prosthetic vein valve. Journal of Medical Devices 1 (2), 105–112. Suggs, J., Wang, C., Li, G., 2003. The effect of graft stiffness on knee joint biomechanics after ACL reconstruction—A 3D computational simulation. Clinical Biomechanics 18 (1), 35–43.
1470
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Vunjak-Novakovic, G., Altman, G., Horan, R., Kaplan, D.L., 2004. Tissue engineering of ligaments. Annual Review of Biomedical Engineering 6, 131–156. Woo, S.L.-Y.,Adams, D.J. 1990. The tensile properties of the human anterior cruciate ligament (ACL) and ACL graft tissues. Knee Ligaments: Structure, Funtion, Injury, and Repair. In: D. Daniel, W. H. Akeson, J. J. O’Connor. (Eds.), New York, Raven Place, pp. 279–289. Woo, S.L.-Y., Hollis, J.M., Adams, D.J., Lyon, R.M., Takai, S., 1991. Tensile properties of the human femur-anterior cruciate ligament-tibia complex: The effects of specimen age and orientation. American Journal of Sports Medicine 19 (3), 217–225. Woo, S.L., Debski, R.E., Withrow, J.D., Janaushek, M.A., 1999. Biomechanics of knee ligaments. American Journal of Sports Medicine 27 (4), 533–543. Wu, Q., Chen, N., Li, L., Wang, Q., 2012. Structure evolution of melt spun poly(vinyl alcohol) fibers during hot drawing. Journal of Applied Polymer Science 124 (1), 421–428.
Yagi, M., Wong, E.K., Kanamori, A., Debski, R.E., Fu, F.H., Woo, S.L.-Y., 2002. Biomechanical analysis of an anatomic anterior cruciate ligament reconstruction. The American Journal of Sports Medicine 30 (5), 660–666. Yahia, L.H., Hagemeister, N., Drouin, G., Rivard, C.H., Rhalmi, S.,Newman, N. 1996. Evaluation of crutiate ligament prostheses—Critique of current concepts and Procedures, Ligaments and Ligamentoplasties. In: L. H. Yahia (Ed.), Berlin, Springer-Verlag. 1, pp. 1–16. Yamaura, K., Kamakura, R., 2000. Gel-spinning of partially saponificated poly(vinyl alcohol). Journal of Applied Polymer Science 77 (13), 2872–2876. Yao, L., Haas, T.W., Guiseppi-Elie, A., Bowlin, G.L., Simpson, D.G., Wnek, G.E., 2003. Electrospinning and stabilization of fully hydrolyzed poly(vinyl alcohol) fibers. Chemistry of Materials 15 (9), 1860–1864.
Contents lists available at SciVerse ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Hydrogel fibers for ACL prosthesis: Design and mechanical evaluation of PVA and PVA/UHMWPE fiber constructs Jason S. Bach a,1, Fabrice Detrez b,2, Mohammed Cherkaoui c,3, Sabine Cantournet b,4, David N. Ku a,n, Laurent Corte´ b,5 a
George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, 30332-0405, USA Centre des Mate´riaux, Mines ParisTech, CNRS UMR 7633, BP 87, F-91003 Evry Cedex, France c George W. Woodruff School of Mechanical Engineering, Georgia Tech Lorraine, 2 Rue Marconi, 57070 Metz, France b
a r t i c l e i n f o
abstract
Article history: Accepted 26 February 2013
Prosthetic devices for anterior cruciate ligament (ACL) reconstruction have been unsuccessful due to mechanical failure or chronic inflammation. Polymer hydrogels combine biocompatibility and unique low friction properties; however, their prior use for ligament reconstruction has been restricted to coatings due to insufficient tensile mechanics. Here, we investigate new constructs of polyvinyl alcohol (PVA) hydrogel fibers. In water, these fibers swell to an equilibrium water content of 50% by weight, retaining a tensile modulus greater than 40 MPa along the fiber axis at low strain. Rope constructs were assembled for ACL replacement and mechanical properties were compared with data from the literature. Pure PVA hydrogel constructs closely reproduce the non-linear tensile stiffness of the native ACL with an ultimate strength of about 2000 N. An additional safety factor in tensile strength was achieved with composite braids by adding ultrahigh molecular weight polyethylene (UHMWPE) fibers around a core of PVA cords. Composition and braiding angle are adjusted to produce a non-linear tensile behavior within the range of the native ligament that can be predicted by a simple rope model. This design was found to sustain over one million cycles between 50 and 450 N with limited damage and less than 20% creep. The promising mechanical performances of these systems provide justification for more extensive in vivo evaluation. & 2013 Published by Elsevier Ltd.
Keywords: Ligament Anterior cruciate ligament Artificial ligament Hydrogel fibers Polyvinyl alcohol Ultrahigh molecular weight polyethylene
1. Introduction Anterior cruciate ligament (ACL) tears do not heal on their own nor respond well to primary repair due to a poor blood supply (Bray et al., 2002). Tendon autografts are currently the standard for ACL replacement (Woo and Adams, 1990) yielding good results. Drawbacks such as donor site morbidity, longer recovery times, and the requirement of the harvest operation with autografts would make an off-the-shelf prosthetic attractive for surgeons and patients (Kartus et al., 2001). Beginning in the
n
Corresponding author. Tel.: þ404 894 6827; fax: þ 404 894 2291. E-mail addresses: [email protected] (J.S. Bach), [email protected] (F. Detrez), [email protected] (M. Cherkaoui), [email protected] (S. Cantournet), [email protected] (D.N. Ku), [email protected] (L. Corte´). 1 Tel.: þ1 404 894 6827; fax: þ1 404 894 2291. 2 Tel.: þ33 1 60 76 30 70; fax: þ33 1 60 76 31 50. 3 Tel.: þ33 3 87 20 3939; fax: þ 33 3 87 20 3940. 4 Tel.: þ33 1 60 76 30 52; fax: þ 33 1 60 76 31 50. 5 Tel.: þ33 1 60 76 31 42; fax: þ 33 1 60 76 31 50. 0021-9290/$ - see front matter & 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jbiomech.2013.02.020
1970s, multiple prosthetics have been developed; however, due to high rates of mechanical failure and immunogenic particulation, the field has turned away from these devices (Mascarenhas and MacDonald, 2008). As surgical techniques and tools have since improved and a wider range of biomaterials have become available, the question of whether a functional prosthetic can be developed is worth exploring again. Polymer hydrogels, consisting of water-swollen networks of cross-linked macromolecules are particularly relevant materials as they are similar to the extra-cellular matrix of native tissues. They have shown promising results as biodegradable scaffolds for tissue-engineered ligaments (Drury and Mooney, 2003). Nevertheless, achieving proper mechanical performance with tissueengineered systems during scaffold degradation and in newly formed tissue remains a challenge. Non-degradable hydrogels could bring a more immediate solution for ligament implants. In particular, polyvinyl alcohol (PVA) can be processed to form a non-degradable hydrogel with excellent biocompatibility (Hassan and Peppas, 2000a). These materials exhibit a unique selflubrication property by forming a thin water layer at the contact interface resulting in low surface friction (Gong, 2006; Mamada et al., 2011). As a result, PVA has shown potential in several
8 mm
o 20% elongation after 1 M cycles from 50 N to 450 N NA
10 mm
Functional stiffness maintained (140 N/% strain) NA
Measurement of length during of fatigue testing
Ruler (Caborn et al., 1998; Duthon et al., 2006)
soft-tissue replacement applications including knee cartilage and vein valve repair (Hassan and Peppas, 2000b; Noguchi et al., 1991; Sathe and Ku, 2007). In this article, we investigate how PVA hydrogels can be used to produce artificial ligaments having mechanical performance suitable for ACL replacement. For ligament implants, hydrogels are usually limited by their insufficient tensile properties and are used as the coating of a stiffer fibrous polymeric construct (Freeman et al., 2011). Nevertheless, the highly oriented organization of the collagen matrix of ligaments suggests that appropriate tensile properties might be attained with synthetic hydrogels having a strong macromolecular orientation along the ligament axis. Such an orientation can be achieved in PVA using fiber forming processes including solutionspinning (Yamaura and Kamakura, 2000), melt-spinning (Wu et al., 2012), or electrospinning (Yao et al., 2003). Water swelling to form hydrogel fibers from spun fibers may release macromolecular orientation and thus prevent access to sufficient tensile stiffness. Here, we show that appropriate tensile properties can be achieved from PVA hydrogel fibers having a water content of 50% by weight, comparable to that of native ligament tissue. In the first part of this study, prosthetic ACLs were assembled from PVA only and compared to the mechanical design inputs obtained from the literature and outlined in Table 1. Particular attention was paid to the proper reproduction of the non-linear tensile response of native ligaments. A periodic crimp pattern of collagen fibrils confers an elastic response to the ligaments with a low stiffness toe regime at low strain and a higher stiffness linear regime at greater strain (Woo et al., 1999). This low stiffness toe likely helps prevent creep and fatigue (Freeman and Kwansa, 2008; Laurencin and Freeman, 2005). A toe region has been integrated in several scaffolds for tissue engineered ACL (Freeman et al., 2007; Gentleman et al., 2003; Horan et al., 2006; Sahoo et al., 2006). In the second part, we explore how high-strength fibers of ultrahigh molecular weight polyethylene (UHMWPE) may be added to provide a safety factor in tensile strength (Cook et al., 2010; Kurtz, 2009). In such composite ropes using two types of fibers, complex structures are required to reproduce the nonlinear tensile behavior of ligaments (Hopper et al., 1995). A hollowbraided design is proposed consisting of a core of PVA hydrogel fibers surrounded by a braided layer of UHMWPE threads. With this structure, the low strain response is governed by the core while the outside layer provides stiffening at larger strains. The relationship between rope composition, geometrical parameters and tensile response is investigated using both experimental measurements and a simple rope model. Based on those results, we discuss the potential of both pure PVA hydrogel fiber constructs or PVA/UHMWPE constructs for ACL replacement.
2. Materials and methods 2.1. Materials PVA that forms a stable insoluble hydrogel below 90 1C was purchased in the form of 675-dtex threads composed of 15 twisted continuous fibers. This form of PVA is over 95% hydrolyzed, similar to other biocompatible PVA grades used for implant applications (Baker et al., 2012). PVA cords of 15 twisted threads were assembled by a textile manufacturer (Morel-Journel, Lyon, France). UHMWPE was purchased in the form of 1200-dtex threads composed of 4 braided bundles of 60 continuous fibers.
2.2. Fabrication of fiber constructs Diameter
Permanent elongation
(FDA, 1993)
Functional stiffness maintained for 41 M load cycles from 50 N to 450 N o 20% permanent elongation after 1 M cycles from 50 N to 450 N o 10 mm Fatigue
Fatigue testing
4 30% 4 12% Uniaxial tensile test
4 40%
6–15% 2–15% Uniaxial tensile test
5–15%
44000 N 4 2000 N 41730 N
Ultimate Load Toe Transition Strain Ultimate Strain
Uniaxial tensile test
25–60 N/% strain 35–400 N/% strain Uniaxial tensile test Stiffness
(Hamner et al., 1999; Noyes et al., 1984; Yagi et al., 2002) (Brown, 1995; Hamner et al., 1999; Noyes et al., 1984; Yagi et al., 2002) (Butler et al., 1986; Grood and Noyes, 1976; Noyes et al., 1984; Noyes and Grood, 1976; Woo and Adams, 1990; Woo et al., 1991) (Butler et al., 1986; Grood and Noyes, 1976; Noyes et al., 1984; Noyes and Grood, 1976; Woo and Adams, 1990; Woo et al., 1991) (FDA, 1993)
PVA construct (Braided 20 PVA cords) Target value Verification test Ref. Design input
Table 1 Design inputs for a prosthetic ACL, verification test results and characteristics of 20 braided PVA cords and hollow-braid constructs with 16 PVA hydrogel cords and 12 UHMWPE threads.
60–180 N/% strain
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
PVA/UHMWPE construct (Hollow-braid/16 PVA cords/12 UHMWPE threads/ h ¼451)
1464
Full-scale constructs for ACL replacement were fabricated manually. Both twisted and diamond braided PVA-only constructs were produced. Twisted samples were obtained from 16 PVA cords by first twisting 4 groups of 4 cords together with a left-hand twist and then, twisting these 4 ropes into a larger one
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470 with a right-hand twist. For the diamond braided samples, 20 PVA cords were braided together using a 1/1 overlap-underlap technique forming a cylinder (Brunnschweiler, 1954). For PVA/UHMWPE constructs, a hollow-braid design was used in which PVA cords were aligned in parallel forming a core, which was then surrounded with a tubular diamond braid of 12 UHMWPE threads. After assembly, the diamond braid was left free to slide at one end to allow swelling of the PVA core after which the braid was tightened around the core and fixed at both ends. A first series was obtained by varying the number of PVA cords in the core to 8, 12, and 16. In a second series, the braid angle of the tubular braid was modified to 407 21, 45 721, 507 21, and 557 21.
1465
80µm
100µm 1/2turn ~ 4mm
2.3. Tensile testing Tensile testing was performed at room temperature in air on wet samples using an Instron 5966 apparatus. Samples were immersed in distilled water for at least 12 h and tested less than 10 min after removal from water to ensure negligible drying. Pull-to-failure tests were performed using capstan fixations with a diameter of 7 cm and sub-failure tests, using large wedge clamps. Full scale samples were stretched for 20 cycles between 0 N and 700 N at a rate of 1 mm per second (about 2% s 1), corresponding to the maximum force during strenuous activity at a medium to slower rate in vivo (Chen and Black, 1980; Noyes et al., 1974, 1984; Pioletti et al., 1999). The tensile curves corresponding to the last loading cycle are presented. Strain was measured using a video extensometer at a rate of 2 images per second. Each test was repeated on at least 3 samples. 2.4. Fatigue testing Samples were first immersed in distilled water for at least 12 h. Fatigue tensile testing was performed on a custom apparatus with an immersion chamber. Sample length was 15 cm including a gauge length of 5 cm between two 5 cm long sections for gripping at each end. Samples were stretched to 500 N for 10 min prior to tightening in this stretched state. This provided a stronger grip and reduced deformation and wear at the clamps. After unloading back to zero, samples were completely submerged in water at room temperature and left to swell for 1 h prior to testing. Fatigue testing simulated a normal non-strenuous activity, operating in force-controlled mode between 50 N and 450 N at 1 Hz, while constantly recording force and displacement data at 100 points per second (Noyes et al., 1984).
3. Results
500µm 1/2turn ~ 1cm
1mm
UHMWPE thread
3mm
PVA cord
Fig. 1. Observations of constitutive fibers and threads: (a) dry PVA fibers (SEM observation), (b) dry PVA thread, (c) dry PVA cord, (d) UHMWPE thread and PVA cord.
3.1. Mechanical evaluation of PVA hydrogel cords Dry PVA fibers, threads, and cords were characterized by electronic and optical microscopy as shown in Fig. 1. SEM observations (Fig. 1a) show that PVA fibers have a roughly circular cross-section with a diameter of 80 71 mm. They are assembled into threads of 15 fibers with a left-hand twist and an 8 mm pitch (Fig. 1b). Threads were themselves assembled into cords of 15 threads with a right-hand twist and a 2 cm pitch (Fig. 1c). Upon immersion in water or bodily fluids, PVA fibers swell becoming hydrogel fibers. Here, PVA is fully hydrolyzed. Its crosslinking is physical and results from the formation of small crystallites that melt in water above 90 1C (Hassan and Peppas, 2000a). Below this melting temperature, no degradation is expected. A stability study confirms that the hydrogel fibers are insoluble at 40 1C and show no change in weight after 3 years of water immersion. Optical microscopic images in Fig. 2a show the swelling of one single fiber. Corresponding diameter measurements in Fig. 2b show that after approximately 10 min, an equilibrium water content of 50% by weight is reached and the diameter increases to 110710 mm. The diameter of swollen threads and cords are 0.5 70.1 mm and 370.5 mm, respectively. The tensile properties of swollen PVA hydrogel cords are shown in Fig. 3. These measurements integrate both the material intrinsic behavior and the deformation of the twisted cord structures. Swollen cords exhibit the characteristic rubbery behavior of PVA hydrogels. They can elastically sustain large strains over 30% with little dissipation as shown by loading–unloading curves in Fig. 3a. They exhibit a non-linear response with a
stiffness of 0.8 N/% strain at low strain and over 2 N/% strain above 10%. Their ultimate strain and load are of the order of 3773% and 100720N. 3.2. PVA hydrogel fiber constructs for ACL replacement From these results, a first design was proposed for PVA-based constructs exhibiting the stiffness and strength adequate for ACL replacement while remaining within size constraints. About 250 to 300 PVA threads, i.e. 3500–4500 fibers, were necessary. Manipulation of fewer cords was preferred and 15 to 20 PVA cords were required to provide the minimal stiffness requirement of 35 N/%strain as reported in the literature (Table 1). Cord assembly was performed manually with two patterns: 16 twisted PVA cords (Fig. 4a) and 20 braided PVA cords (Fig. 4b). All tested constructs were assembled with dry PVA cords and were submerged in distilled water at room temperature for at least 12 h prior to testing. Both designs were 10 mm in diameter, the maximum allowable before exceeding the size constraints and slightly larger than the 8 mm of an average hamstring graft (Caborn et al., 1998). Tensile curves up to failure are given in Fig. 5 and compared to the range of behavior of native ACL as estimated from force– displacement data by Woo et al. considering a ligament length between 20 and 40 mm (Caborn et al., 1998; Woo et al., 1991). Twisted 16 PVA cords have stiffness close to the lower bound of normal elasticity and strength of 1800 7100 N, slightly inferior to the native ACL. Braided 20 PVA cords provide a non-linear
1466
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Fig. 2. Characterization of fiber swelling. (a) Single fiber observation by optical microscopy and (b) time resolved measurement of fiber diameter during swelling.
response within the range for native ACL with an ultimate load of 2200 7100 N, equivalent to the average ACL in the young adult (Woo et al., 1991). Nevertheless, at low strain ( o5%), the braided 20 PVA cords provide a rather high stiffness of about 25 N/% strain instead of the low stiffness toe of native ACL as shown in Fig. 6. 3.3. Hollow-braided constructs of PVA/UHMWPE fibers A safety factor in terms of ultimate load would be advantageous and is achievable by adding UHMWPE fibers of 1200dtex shown in Fig. 1d. An UHMWPE thread has a high stiffness close to 100 N/% strain and an ultimate load of 375 N as indicated in Fig. 3. The coarse assembly of 12 UHMWPE threads twisted with 16 PVA cords (Fig. 4c) provides an ultimate load twice that of the native ACL; however, it results in a very high stiffness when the response is dominated by the UHMWPE threads under tension as indicated in Fig. 5. More complex fiber assemblies should allow retention of high tensile strength while reducing stiffness closer to the level of native ACL to avoid over-constraint of the knee (Suggs et al., 2003). For PVA/UHMWPE assemblies, a hollow-braided construct was designed where a core of parallel PVA cords is wrapped by a diamond braided structure of UHMWPE threads as shown in Fig. 4d. In this construct, the tensile response at low strain is governed by the PVA core. As strain increases, the diamond braided structure deforms and compresses the core. UHMWPE threads are progressively put under tension and gradually stiffen the response. Two design parameters were modified to adjust the shape of the tensile response: the number of PVA cords in the core and the braid angle, y, between the UHMWPE thread axis and the ligament axis, as illustrated in Fig. 4d.
Fig. 3. Tensile force versus strain for PVA hydrogel cords. (a) Four loadingunloading cycles (first cycle not shown), (b) loading to failure curves for PVA hydrogel cords (3 samples) and UHMWPE thread. The end of the curves corresponds to failure.
Fig. 4. Full-scale constructs after equilibration in water for ACL replacement. (a) Sixteen twisted PVA cords, (b) 20 braided PVA cords, (c) 16 twisted PVA cords and 12 UHMWPE threads and (d) Hollow-braided assembly of 16 PVA cords and 12 UHMWPE threads.
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
1467
Fig. 5. Tensile forceversus strain for 16 twisted PVA hydrogel cords (blue), 20 braided PVA hydrogel cords (green) and a mixed construct of 16 PVA hydrogel cords twisted along with 12 UHMWPE threads (red). Curves are compared with the native ACL and tendons used for autografts shaded in gray (Duthon et al., 2006; Kim et al., 2003; Noyes et al., 1984; Woo et al., 1991). The ends of the curves correspond to failure. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 7. Role of geometrical parameters in hollow-braided constructs: (a) Force versus strain for samples with 8, 12, and 16 PVA hydrogel cords in the core and braid angle of 451. Inset: close-up on the low strain toe region. Symbols are experimental data and thin lines show model predictions and (b) force versus strain for samples with 16 PVA hydrogel cords in the core and braid angles of 401, 451, 501, and 551. Thin lines show model predictions.
In a simple model assuming incompressibility of the core, the composition, initial braid angle (y0), and mechanical stiffness of PVA cords and UHMWPE threads are utilized as inputs to predict the tensile behavior of the structure (see supporting information). At low strains, the tensile behavior is given by the response of the core while the contribution of the outer braid is added above a critical elongation l* which corresponds to the positive root of equation Fig. 6. Tensile force versus strain (Close-up on 0–700 N range) for 16 twisted PVA hydrogel cords (blue), 20 braided PVA hydrogel cords (green) and hollow-braided construct (HB) with a core of 16 PVA hydrogel cords and an outer braid of 12 UHMWPE threads with y ¼ 451 (red). Shaded regions show typical range for the native ACL and tendons used for autografts (Duthon et al., 2006; Kim et al., 2003; Noyes et al., 1984; Woo et al., 1991). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 7a shows the behavior of hollow-braids with 12 UHMWPE threads braided at y ¼451 and cores composed of 8, 12, and 16 PVA cords with diameters after swelling of 5.5 mm, 7 mm, and 8 mm, respectively. Increasing the number of PVA cords stiffens the toe and smoothes the transition from lower to higher stiffness. Fig. 7b shows the force–strain curves for hollow-braided samples with 16 PVA cords in the core and 12 UHMWPE threads braided at y ¼40 721, 45721, 50 721, and 55721. Increasing the braid angle implies that UHMWPE threads are recruited at greater tensile strains, therefore increasing the size of the low stiffness toe and decreasing the overall stiffness of the constructs. The best match to the native ACL is provided with a braid angle between 401 and 501 and is reproduced in Fig. 6 for y ¼451.
l3 lð1 þ ðtany0 Þ2 Þ þðatany0 Þ2 ¼ 0 The only fitting parameter, the ratio (a) of the initial braid radius to the initial core radius, is estimated to be 1.15, which is a realistic value given the assembly process. Stiffness of the PVA cords and UHMWPE threads is taken to be 0.8 N/% strain and 100 N/% strain, respectively. The obtained predictions are in fairly good agreement with the experimental data shown in Fig. 7a and b. Dependence of the toe stiffness and on the core size and of the toe size on the initial braid angle are well captured. Predictions indicate that braid angles lower than 501 are required to obtain significant stiffening from the UHMWPE outer braid in the 0–30% strain range. Resistance to fatigue was evaluated on a set of three hollowbraided samples made of 16 PVA cords and 12 UHMWPE threads (y ¼451). These constructs were able to withstand cyclic loading over the physiological tensile force range of 50–450 N up to 1 million caycles. Loading–unloading loops in Fig. 8a show that the stiffness remained within the acceptable range and slightly increased over the course of loading due to bedding in of the UHMWPE fibers into the PVA core. From the minimum strain in
1468
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Novakovic et al., 2004). Though the proposed design is slightly inferior in strength compared to the Gore-tex graft, which had an ultimate load of 5300 N, rupture was not a prominent failure mode for that device (Mascarenhas and MacDonald, 2008; Moyen and Lerat, 1994; Vunjak-Novakovic et al., 2004). The addition of UHMWPE fibers may raise again the issue of particulation of dry polymer fibers for ACL reconstruction (Guidoin et al., 2000). Nevertheless, the number of friction points is greatly reduced in the PVA/UHMWPE hollow-braid structure as compared to full UHMWPE constructs. Several approaches are being investigated to alleviate this possible limitation, including the embedding of UHMWPE fibers in a hydrogel matrix and the use of parallel helices of UHMWPE instead of diamond braid to suppress thread-thread contacts. Our fatigue study was limited to a maximum load of 450N, representative of normal activities; however, loads as high as 700N may occur in vivo (Chen and Black, 1980; Noyes et al., 1984). An ongoing study involves loading at higher and lower magnitudes continuing until rupture. Nevertheless, for ligament replacement, in vitro tests are still unsatisfactory to predict longterm in vivo use (Yahia et al., 1996). Major deviations between in vitro predictions and the observed clinical behavior have been reported on Gore-tex ligaments and bring to light the importance of long-term in vivo studies in the next stage of device development (Bolton and Bruchman, 1985).
Fig. 8. Fatigue study of a hollow-braided sample in water (outside braid: 12 UHMWPE threads with y ¼ 451, core: 16PVA cords). (a) Load versus deformation curves over the course of cyclic loading and (b) minimum and maximum strains in one loading cycle as a function of the number of cycles.
each cycle, an upper bound for permanent elongation of after 1 million cycles was about 20% (Fig. 8b).
5. Conclusion This mechanical evaluation shows that PVA hydrogel fibers may be the basis for novel ACL replacements that would exhibit the biocompatibility and low wear properties of hydrogels as well as an ACL-like tensile response. We find that the addition of high strength fibers and a hollow-braid design provide ways to modulate the performances of such constructs while satisfying the mechanical requirements for a prosthetic anterior cruciate ligament.
4. Discussion These results show that hydrogel fiber constructs can closely reproduce the tensile behavior of native ACL (see Fig. 6 and Table 1). Ultimate strength remains a limitation for pure hydrogel fiber constructs with a diameter below 10 mm. Application of previously developed fiber assemblies for ACL scaffolds (Freeman et al., 2007; Gentleman et al., 2003; Horan et al., 2006; Sahoo et al., 2006) or PVA modification techniques (Choi et al., 2012; Wu et al., 2012) may improve the performances of these systems. A straight-forward solution to gain a safety factor in tensile strength is to introduce high modulus fibers of UHMWPE. Using a hollow-braid design, PVA/UHMWPE constructs provided a controlled toe region, which could be matched to physiological needs by adjusting the core size and braid angle. 12 UHMWPE threads wound at a braid angle from 401 to 501 around a core of 16 PVA cords in parallel fulfilled the requirements for stiffness, ultimate strain, toe region effect, and resistance to fatigue and permanent elongation (see Table 1). The ultimate load of the proposed design with UHMWPE is greater than scaffolds used for tissue engineered ligaments such as those made of polylactide-co-glycolide with an ultimate load from 525 to 907 N and those of silk with ultimate loads as high as 2337 N (Altman et al., 2002; Cooper et al., 2005). It is also greater than other previous permanent prosthetic devices that were susceptible to rupture as a failure mode such as the carbon fiber, Leeds-Keio, Dacron, and ABC prosthetic ligaments with ultimate loads ranging from 660 to 3631 N.(Mascarenhas and MacDonald, 2008; Moyen and Lerat, 1994; Petrigliano et al., 2006; Vunjak-
Conflict of interest statement The authors report they have applied for an International Patent on the ACL prosthesis. There has been no significant financial support for this work that could have influenced its outcome.
Acknowledgments The authors would like to thank Yann Auriac and JeanChristophe Tesseidre for technical support on mechanical testing, for the video extensometry software. Funding from the French National Research Agency (ANR) through Emergence-TEC program (ANR-08-ETEC-003) and Georgia Tech-Lorraine is gratefully acknowledged.
Appendix A See Fig. A1
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jbiomech. 2013.02.020.
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Fig. A1. (a) Schematic representation of a hollow-braid construct, (b) evolution of the core radius (full line) and outer braid radius (dashed line) as a function of stretching ratio. The cross-over gives the locking ratio when the outer braid comes into contact with the core. (Rb0/Rc0 ¼ 1.08, y0 ¼451) and (c) stretching ratio at locking as a function of braid angle (Rb0/Rc0 ¼ 1.08).
References Altman, G.H., Horan, R.L., Lu, H.H., Moreau, J., Martin, I., Richmond, J.C., Kaplan, D.L., 2002. Silk matrix for tissue engineered anterior cruciate ligaments. Biomaterials 23 (20), 4131–4141. Baker, M.I., Walsh, S.P., Schwartz, Z., Boyan, B.D., 2012. A review of polyvinyl alcohol and its uses in cartilage and orthopedic applications. Journal of Biomedical Materials Research Part B: Applied Biomaterials 100B (5), 1451–1457. Bolton, C.W., Bruchman, W.C., 1985. The gore-tex expanded polytetrafluoroethylene prosthetic ligament. An in vitro and in vivo evaluation. Clinical Orthopaedics and Related Research 196, 202–213. Bray, R.C., Leonard, C.A., Salo, P.T., 2002. Vascular physiology and long-term healing of partial ligament tears. Journal of Orthopaedic Research 20 (5), 984–989. Brown, C. H., 1995. Biomechanics of the Semitendinosus and Gracilus Tendon Grafts. In: Procedings of the American Orthopaedic Society for Sports Medicine Conference, Toronto, Canada. Brunnschweiler, D., 1954. 5-The structure and tensile properties of braids. Journal of the Textile Institute Transactions 45 (1), T55–T77. Butler, D.L., Kay, M.D., Stouffer, D.C., 1986. Comparison of material properties in fascicle-bone units from human patellar tendon and knee ligaments. Journal of Biomechanics 19 (6), 425–432. Caborn, D.N.M., Coen, M., Neef, R., Hamilton, D., Nyland, J., Johnson, D.L., 1998. Quadrupled semitendinosus-gracilis autograft fixation in the femoral tunnel: A comparison between a metal and a bioabsorbable interference screw. Arthroscopy: The Journal of Arthroscopic and Related Surgery 14 (3), 241–245. Chen, E.H., Black, J., 1980. Materials design analysis of the prosthetic anterior cruciate ligament. Journal of Biomedical Materials Research 14 (5), 567–586. Choi, J., Kung, H.J., Macias, C.E., Muratoglu, O.K., 2012. Highly lubricious poly(vinyl alcohol)-poly(acrylic acid) hydrogels. Journal of Biomedical Materials Research Part B: Applied Biomaterials 100 (2), 524–532. Cook, J.L., Luther, J.K., Beetem, J., Karnes, J., Cook, C.R., 2010. Clinical comparison of a novel extracapsular stabilization procedure and tibial plateau leveling
1469
osteotomy for treatment of cranial cruciate ligament deficiency in dogs. Veterinary Surgery 39 (3), 315–323. Cooper, J.A., Lu, H.H., Ko, F.K., Freeman, J.W., Laurencin, C.T., 2005. Fiber-based tissue-engineered scaffold for ligament replacement: design considerations and in vitro evaluation. Biomaterials 26 (13), 1523–1532. Drury, J.L., Mooney, D.J., 2003. Hydrogels for tissue engineering: Scaffold design variables and applications. Biomaterials 24 (24), 4337–4351. Duthon, V.B., Barea, C., Abrassart, S., Fasel, J.H., Fritschy, D., Me´ne´trey, J., 2006. Anatomy of the anterior cruciate ligament. Knee Surgery, Sports Traumatology, Arthroscopy 14 (3), 204–213. Freeman, J.W., Kwansa, A.L., 2008. Recent advancements in ligament tissue engineering: The use of various techniques and materials for ACL repair. Recent Patents on Biomedical Engineering 1, 18–23. Freeman, J.W., Woods, M.D., Cromer, D.A., Ekwueme, E.C., Andric, T., Atiemo, E.A., Bijoux, C.H., Laurencin, C.T., 2011. Evaluation of a hydrogel-fiber composite for ACL tissue engineering. Journal of Biomechanics 44 (4), 694–699. Freeman, J.W., Woods, M.D., Laurencin, C.T., 2007. Tissue engineering of the anterior cruciate ligament using a braid-twist scaffold design. Journal of Biomechanics 40 (9), 2029–2036. Gentleman, E., Lay, A.N., Dickerson, D.A., Nauman, E.A., Livesay, G.A., Dee, K.C., 2003. Mechanical characterization of collagen fibers and scaffolds for tissue engineering. Biomaterials 24 (21), 3805–3813. Gong, J.P., 2006. Friction and lubrication of hydrogels—its richness and complexity. Soft Matter 2 (7), 544–552. Guidoin, M.F., Marios, Y., Bejui, J., Poddevin, N., King, M.W., Guidoin, R., 2000. Analysis of retrieved polymer fiber based replacements for the ACL. Biomaterials 21 (23), 2461–2474. Hassan, C.M., Peppas, N.A., 2000a. Structure and applications of poly(vinyl alcohol) hydrogels produced by conventional crosslinking or by freezing/thawing methods. Advances in Polymer Science 153, 37–66. Hassan, C.M., Peppas, N.A., 2000b. Structure and morphology of freeze/thawed PVA hydrogels. Macromolecules 33 (7), 2472–2479. Hamner, D.L., Brown, C.H., Steiner, M.E., Hecker, A.T., Hayes, W.C., 1999. Hamstring tendon grafts for reconstruction of the anterior cruciate ligament: biomechanical evaluation of the use of multiple strands and tensioning techniques. Journal of Bone and Joint Surgery, American Volume 81 (4), 549–557. Hopper, R.H., Grant, J.W., Popper, P., 1995. Mechanics of a hybrid circular braid with an elastic core. Textile Research Journal 65 (12), 709–722. Horan, R.L., Collette, A.L., Lee, C., Antle, K., Chen, J., Altman, G.H., 2006. Yarn design for functional tissue engineering. Journal of Biomechanics 39 (12), 2232–2240. Kartus, J., Movin, T., Karlsson, J., 2001. Donor-site morbidity and anterior knee problems after anterior cruciate ligament reconstruction using autografts. Arthroscopy: The Journal of Arthroscopic and Related Surgery 17 (9), 971–980. Kim, D.H., Wilson, D.R., Hecker, A.T., Jung, T.M., Brown, C.H., 2003. Twisting and braiding reduces the tensile strength and stiffness of human hamstring tendon grafts used for anterior cruciate ligament reconstruction. The American Journal of Sports Medicine 31 (6), 861. Kurtz, S.M., 2009. UHMWPE Biomaterials Handbook: Ultra High Molecular Weight Polyethylene in Total Joint Replacement and Medical Devices. Academic Press, London. Laurencin, C.T., Freeman, J.W., 2005. Ligament tissue engineering: An evolutionary materials science approach. Biomaterials 26 (36), 7530–7536. Mamada, K., Fridrici, V., Kosukegawa, H., Kapsa, P., Ohta, M., 2011. Friction properties of poly(vinyl alcohol) hydrogel: Effects of degree of polymerization and saponification value. Tribology Letters 42 (2), 241–251. Mascarenhas, R., MacDonald, P.B., 2008. Anterior cruciate ligament reconstruction: A look at prosthetics—past, present and possible future. McGill Journal of Medicine 11 (1), 29–37. Moyen, B., Lerat, J.-L., 1994. Artificial ligaments for anterior cruciate ligament replacement. Journal of Bone and Joint Surgery 76-B (2), 173–175. Noguchi, T., Yamamuro, T., Oka, M., Kumar, P., Kotoura, Y., Hyon, S.-H., Ikada, Y., 1991. Poly(vinyl alcohol) hydrogel as an artificial articular cartilage: Evaluation of biocompatibility. Journal of Applied Biomaterials 2 (2), 101–107. Noyes, F.R., Butler, D.L., Grood, E.S., Zernicke, R.F., Hefzy, M.S., 1984. Biomechanical analysis of human ligament grafts used in knee-ligament repairs and reconstructions. Journal of Bone and Joint Surgery, American 66 (3), 344–352. Noyes, F.R., DeLucas, J.L., Torvik, P.J., 1974. Biomechanics of anterior cruciate ligament failure: An analysis of strain-rate sensitivity and mechanisms of failure in primates. Journal of Bone and Joint Surgery 56-A (2), 236–253. Noyes, F.R., Grood, E.S., 1976. The strength of the anterior cruciate ligament in humans and rhesus monkeys. Journal of Bone and Joint Surgery, American Volume 58 (8), 1074–1082. Petrigliano, F.A., McAllister, D.R., Wu, B.M., 2006. Tissue engineering for anterior cruciate ligament reconstruction: A review of current strategies. Arthroscopy: The Journal of Arthroscopic and Related Surgery 22 (4), 441–451. Pioletti, D.P., Rakotomanana, L.R., Leyvraz, P.-F., 1999. Strain rate effect on the mechanical behavior of the anterior cruciate ligament-bone complex. Medical Engineering & Physics 21 (2), 95–100. Sahoo, S., Ouyang, H., Goh, J.C.-H., Tay, T.E., Toh, S.L., 2006. Characterization of a novel polymeric scaffold for potential application in tendon/ligament. Tissue Engineering. Tissue Engineering 12 (1), 1–9. Sathe, R.D., Ku, D.N., 2007. Flexible prosthetic vein valve. Journal of Medical Devices 1 (2), 105–112. Suggs, J., Wang, C., Li, G., 2003. The effect of graft stiffness on knee joint biomechanics after ACL reconstruction—A 3D computational simulation. Clinical Biomechanics 18 (1), 35–43.
1470
J.S. Bach et al. / Journal of Biomechanics 46 (2013) 1463–1470
Vunjak-Novakovic, G., Altman, G., Horan, R., Kaplan, D.L., 2004. Tissue engineering of ligaments. Annual Review of Biomedical Engineering 6, 131–156. Woo, S.L.-Y.,Adams, D.J. 1990. The tensile properties of the human anterior cruciate ligament (ACL) and ACL graft tissues. Knee Ligaments: Structure, Funtion, Injury, and Repair. In: D. Daniel, W. H. Akeson, J. J. O’Connor. (Eds.), New York, Raven Place, pp. 279–289. Woo, S.L.-Y., Hollis, J.M., Adams, D.J., Lyon, R.M., Takai, S., 1991. Tensile properties of the human femur-anterior cruciate ligament-tibia complex: The effects of specimen age and orientation. American Journal of Sports Medicine 19 (3), 217–225. Woo, S.L., Debski, R.E., Withrow, J.D., Janaushek, M.A., 1999. Biomechanics of knee ligaments. American Journal of Sports Medicine 27 (4), 533–543. Wu, Q., Chen, N., Li, L., Wang, Q., 2012. Structure evolution of melt spun poly(vinyl alcohol) fibers during hot drawing. Journal of Applied Polymer Science 124 (1), 421–428.
Yagi, M., Wong, E.K., Kanamori, A., Debski, R.E., Fu, F.H., Woo, S.L.-Y., 2002. Biomechanical analysis of an anatomic anterior cruciate ligament reconstruction. The American Journal of Sports Medicine 30 (5), 660–666. Yahia, L.H., Hagemeister, N., Drouin, G., Rivard, C.H., Rhalmi, S.,Newman, N. 1996. Evaluation of crutiate ligament prostheses—Critique of current concepts and Procedures, Ligaments and Ligamentoplasties. In: L. H. Yahia (Ed.), Berlin, Springer-Verlag. 1, pp. 1–16. Yamaura, K., Kamakura, R., 2000. Gel-spinning of partially saponificated poly(vinyl alcohol). Journal of Applied Polymer Science 77 (13), 2872–2876. Yao, L., Haas, T.W., Guiseppi-Elie, A., Bowlin, G.L., Simpson, D.G., Wnek, G.E., 2003. Electrospinning and stabilization of fully hydrolyzed poly(vinyl alcohol) fibers. Chemistry of Materials 15 (9), 1860–1864.