TMS QUICK LINKS:
  • TMS ONLINE
  • MEMBERS ONLY
  • MEMBERSHIP INFO
  • MEETINGS CALENDAR
  • PUBLICATIONS

JOM QUICK LINKS:
  • JOM HOME PAGE
  • MORE HTML ARTICLES
  • COMPLETE ISSUES
  • BOOKS REVIEWS
  • SUBSCRIBE

This paper aims to give a broad overview of the challenges that are faced in load-bearing medical devices and focuses specifically on the challenges faced in utilizing polymeric materials in such applications. Three specific cases are given in the field of polymeric biomaterials. These cases build in complexity and initiate with examination of the evolution of intravascular catheter design in which the materials, properties, and processing have been optimized to develop a system that can be used in an angioplasty procedure with little concern of clinical failure.

INTRODUCTION

Medical polymers are used in a broad range of applications including tissue repair and replacement, drug delivery, and wound healing.¹ Polymers are capable of a wide range of structural properties that depend on backbone structure, molecular weight, entanglement density, degree of crystallinity, and degree of crosslinking.⁹ In general, polymers exhibit time-dependent mechanical behavior and are known to be viscoelastic. For example, the elastic modulus and yield strength of a polymer generally increases with increasing strain rate while the strain to failure typically decreases with increased loading rates. Similarly, sustained loads can result in time-dependent strain or creep in polymers. Time-dependent material properties render the prediction of invivo performance challenging, particularly when the load conditions become complex. In fact, load-bearing medical devices often subject the polymer components to their limits of yield, fracture, wear, and fatigue resistance. Table I presents several applications of polymers in load-bearing implants.

Understanding structure-property design relationships is essential for the successful performance of a medical implant. Yet, implants often undergo iterative changes in design, materials selection, and processing, resulting from the study of their overall clinical performance. Sometimes challenges can be adequately addressed in the laboratory, but often feedback from the clinical use of the device is key to understanding the factors at play. In this section we detail three cases where the design, material, process, or functional properties have evolved in a systematic way in the medical device industry. These cases include an intravascular balloon catheter in which the materials, properties, and processing have been optimized to develop a system that can be used in an angioplasty procedure with little concern of clinical failure; silicone breast implants, which have utilized a shift in design and materials to develop more robust and leak-resistant implants; and total hip replacements, where a shift in material properties without a change in design enabled catastrophic fracture of the polymer bearing component.

Intravascular Balloon Catheters
Intravascular catheters are widely used in both diagnostic and interventional procedures. Balloon catheters are probably best known for their clinical success in coronary angioplasty.¹⁰ In such applications the balloon at the distal end of a catheter is inflated to open an occluded blood vessel afflicted with heart disease. Such systems are also used in stent deployment and Figure 3 shows a rendition of a catheter balloon used in the deployment of a coronary stent for the repair of an occluded artery.¹¹ There are a few important functional requirements for the polymeric balloon: its profile must be small enough to be navigated through the coronary arteries; it must provide sufficient radial force to open an occluded vessel, deploy a stent, or in some instances it may need to exert a radial force on a highly calcified plaque; and it must withstand the inflation pressures necessary for the clinical procedure without rupturing. For these reasons most polymeric catheter balloons are typically made of polyester or nylon due to their tensile strengths and ease of processing.

An intravascular balloon catheter system used in coronary angioplasty is typically inserted through the femoral artery and then it is navigated through the tortuous vasculature to its final destination in the heart. Due to the anatomical requirements there are a number of unique functional requirements for intravascular catheter systems. Intravascular catheters are generally constructed of a long tube with an inner opening or lumen that accommodates the guide wire that delivers the balloon to its final destination and that facilitates an inflation mechanism. Figure 4a shows a schematic illustration of a balloon catheter system and its cross section and Figure 4b shows a specific cross section in a modern catheter comprised of a high-density polyethylene (HDPE) inner layer, an outer layer of polyester (Pebax®), and a functionalized lowdensity polyethylene tie layer (Plexar®) that facilitates bonding between the two layers.

In designing the catheter shaft it is ideal to make the system in such a way that it can readily navigate the tortuosity inherent to the vascular system without kinking or penetration of the tissue. In order to maneuver through the anatomy, the catheter tube needs to provide flexibility to follow the desired path of the surgeon. This functional property is referred to as “trackability.” The catheter tube must also have sufficient axial stiffness to travel along the winding path of the vasculature. Additionally the catheter tube should offer resistance to twisting or in transmitting torque from the proximal to distal (balloon) end; this property is termed “torqueability.”¹¹

Initial catheter tubes were made of a single polymeric material such as nylon, polyethylene, or polyethylene terephthalate, but these polymers were limited by their relatively high coefficients of friction. In an iteration of design, these polymers were coated with silicone to achieve the desired coefficient of friction, however, these systems were limited in their ability to deliver sufficient trackability and torqueability. Subsequent designs moved toward a co-extruded system using two different polymers: HDPE for the inner tube that rides over the guide wire and a nylon or polyester outer tube that can be chemically or thermally bonded to the distal balloon. In its basic form HDPE does not form a chemical bond with a nylon or polyester material, and consequently these systems were also prone to delamination. In order to solve this material-processing-performance problem, the use of a chemical functional group that could be copolymerized with the HDPE was used to achieve a chemical bond between the two distinct polymers. The use of functional groups or tie layers evolved the intravascular catheter tube into a system that could offer the required functional properties without delamination and could be readily manufactured using a co-extrusion process (Figure 5).

Modern vascular catheters either use functional groups or tie layers to facilitate bonding between the inner HDPE and the outer polyester or nylon material, as shown in Figure 4b. Such systems offer good structural integrity, deliver the required functional properties, and minimize complications due to delamination between the materials. Thus, the design evolution of the modern catheter system involved several iterations of materials selection, processing, and design to achieve the functional requirements of the intravascular balloon catheter in a clinical setting. This is an example of where design iterations that have transpired over several decades have resulted in the evolution of a very reliable biomedical device.

Silicone Breast Implants
Silicone has been utilized in breast implants since 1962; however, early designs were prone to rupture and leaking.¹² The first implants utilized low viscosity forms of silicone gel encased in a solid silicone elastomer shell but were prone to failure through tissue contracture around the implant or rupture.^12–14 Rupture of the implant shell enabled the silicone to leak into the surrounding tissue, which could then elicit a chronic inflammatory response. Over the next 20 years, second and third generations of breast implants experimented with different silicone gels and encasement designs to minimize or prevent leaks and inflammatory complications. While these implants were a vast improvement over previous design iterations, leakage and rupture were still major problems.

Many of the early designs and materials used were prior to the 1976 Medical Device Act that provided the Federal Drug Administration (FDA) the authority to review and regulate medical devices.⁴ Without any formal regulation, companies could switch materials used in their devices without the need of FDA approval. Thus, the earlier silicone breast implant designs were not subjected by the FDA to the scrutiny of structural assessment and biocompatibility testing that newer devices must undergo prior to approval for clinical use. In the early 1990s, extensive litigation and research into the realm of silicone breast implant safety ensued with great controversy. The premise of these lawsuits was that leaking of silicone would lead to connective tissue disease, immune reactions, and ultimately autoimmune disorders.¹⁵

The FDA put together a scientific committee to assess the integrity and medical concerns surrounding breast implants in 1991, while mandating the withdrawal of silicone implants for cosmetic use until the investigation of the scientific committee was complete.¹⁴ In 1999, the scientific committee concluded that silicone breast implants were not responsible for the immunological diseases that had been rampant in many patients who had the silicone implants.¹⁵ In 2006, the FDA approved the re-release of silicone breast implants utilizing a crosslinked form of the polymer, under the conditions that patients must be at least 22 years of age and would require an MRI in the first 3 years and then every 2 years thereafter. The current view is that these implants will not likely last a “lifetime” as initially promised in the early release of silicone implants, and women should plan to have multiple surgeries.¹⁶

Modern silicone breast implants have evolved in design to ensure safety against rupture and leakage. The primary mechanical design requirement of a breast implant is resistance to rupture. This is typically modeled as a thin-walled pressure-vessel to address the stress resulting from peak compressive forces. The primary design change from the first implants has been in the utilization of a crosslinked form of silicone that does not leak if the surrounding shell is ruptured or torn. Figure 6 shows the design evolution of silicone implants and provides an image showing the consequences of rupture when the silicone has low viscosity. In the modern silicone implant the implant retains its structural integrity and does not leak into adjacent tissue.

However, several concerns linger in the wake of the recent FDA approval of silicone implants. One concern is founded upon prior clinical complications of leaking and association with autoimmune disease. While scientific studies have shown no link between silicone and autoimmune disease, much of the public at large remains cautious. Also, while a large number of clinical studies have been undertaken to demonstrate the safety of silicone implants, the long-term performance of these implants has not been established.

This is an example of a design iteration process where the clinical performance has driven the research for optimization of materials selection and design of the implant to ensure safety. As with all devices, the clinical performance drives a continuous feedback loop, and it will be years before the long-term performance of crosslinked silicone breast implants is understood.


Ultra-High-Molecular-Weight Polyethylene in Total Hip Replacements
The designs and materials used in total hip replacements have been under steady improvement for nearly 50 years, and currently enjoy a high degree of success with an estimated 90% survival rate after 10 years in vivo as the result of this effort.¹⁷ In total hip arthroplasty, the bearing system typically employs an ultra-high-molecular-weight polyethylene (UHMWPE) insert that articulates against a cobalt-chromium alloy or ceramic in order to restore function to a damaged or diseased joint. The majority of total hip replacement systems in use today utilize a modular design, where the UHMWPE bearing is assembled to a metal shell that integrates with the bone of the acetabulum of the pelvis (Figure 7). The UHMWPE component must be held in place by a combination of locking mechanisms and interference fitting. Locking mechanisms often take the form of notches or grooves that cause a stress concentration during loading of the implant, and are located where the component can experience substantial tensile stress. Such design features are a potential structural concern, particularly for a relatively flaw-intolerant material such as UHWMPE.

One of the primary clinical concerns in total hip replacements is wear-mediated osteolysis, in which inert microscopic wear debris from the bearing cause an acute immune response that results in bone lesions that can compromise the implant.^18–20 In the last decade, the mitigation of wear volume has been the main focus of technical development, and the principal breakthrough in that area has been the use of ionizing radiation to crosslink the UHMWPE bearing for improved wear resistance. Crosslinking of UHMWPE has been shown to reduce the volume of evolved wear particles in pin-on-disc and in vitro implant simulator studies, with a dose-dependent relationship, as shown in Figure 8a.²¹ This effect saturates around 100 kGy of radiation, when the polymer is termed highly crosslinked. Combined with a post-irradiation annealing or melting step to eliminate free radicals, crosslinking substantially reduces the strength, ductility, toughness, and fatigue crack propagation resistance of UHMWPE.²² Thus, mitigating wear via radiation crosslinking results in a tradeoff against other material performance characteristics, such as fatigue crack propagation resistance (Figure 8b).²²

Performance Tradeoffs in Total Hip Replacements
There are performance tradeoffs in total hip replacements owing to the benefit of improved wear resistance at the expense of fatigue fracture in crosslinked UHMWPE. In fact, recent failure analyses of highly crosslinked UHMWPE hip replacement components have indicated that these systems are susceptible to fracture in a clinical environment.^23,24 The authors analyzed the clinical failure of four catastrophically fractured, crosslinked acetabular liners to elucidate this performance tradeoff.²⁴ Each implant was designed and manufactured by a different device manufacturer, but shared similar design features: an unsupported rim outside the main weight-bearing region containing notches or interfaced with a notch in the metal shell. Another critical aspect these implants shared was that they were all developed originally for uncrosslinked UHMWPE, and subsequently deployed with crosslinked UHMWPE without updating the design to reflect the consequent reduction in defect tolerance.

The failure analysis sought to clarify whether the mechanical compromise resulting from crosslinking might have been sufficient to enable the observed fractures. Fractography results demonstrated that the fractures in each case initiated in a microscopically similar manner, at the root of a stress-concentrating feature, despite their different designs and clinical case histories. The fracture surfaces exhibited faint lines parallel to the advancing crack front, originating at a point on the outside surface of the component near the focus of a stress concentration. Near the initiation site, these surface features were prominent and resembled clamshell markings, propagating in a roughly radial or thumbnail morphology. A representative initiation site is shown in Figure 9, where the procession of clam shell markings is distinctly visible.²⁴ The apparent fractographic similarity of the crack initiation sites was taken as strong evidence in support of a common failure mechanism among the four components, and thus the failures were likely derivative of their common material and design attributes.

A finite element analysis was then conducted of each liner to predict the stress generated during a 500 N direct loading event of the exposed rim. One representative result is given in Figure 10.²⁴ This analysis showed that the resultant maximum principal stress peaked at the root of the notches, near where the cracks initiated. This principal stress exceeded that necessary for incipient propagation of a 2 mm deep notch with an incipient crack in each case, using a stress intensity inception value to depict the conditions for the onset of crack growth. This finding was interpreted to indicate that the peak principal stress was sufficient to propagate initiated cracks at the observed initiation sites. Thus the authors concluded that direct rim loading is a sufficient condition to propagate cracks beyond a rim notch and lead to the observed catastrophic fractures.

The finite element analysis also predicted that the stress rapidly decayed with depth from the surface of the rim notch in each case (Figure 10). Thus, while the stress was severe enough for incipient propagation near the surface, a short distance of growth could put a crack outside the notch-affected zone and lead to crack arrest. As the majority of acetabular liners experience substantial rim loading events,²⁵ we therefore hypothesized that a substantial fraction of intact crosslinked acetabular liners should harbor at least one initiated fatigue crack near an elevated rim notch. A subsequent investigation of intact retrieved crosslinked liners reported that six of nine inspected liners harbored initiated cracks.²⁶ These results motivate the need to understand crack initiation in existing and future designs of UHMWPE acetabular liners.

Design for Crack Initiation Resistance
The crack initiation resistance of a UHMWPE component is governed by intrinsic material behavior, extrinsic design, and clinical factors. The material and design characteristics of importance depend on the physical model used to describe crack initiation. The viscous flow of the highly stressed material at a notch or crack tip has been proposed as the dominant deformation fracture mechanism in UHMWPE.^27,28 A crack initiation framework, based on viscoplastic behavior,²⁹ can be used to evaluate tradeoffs in performance related to material behavior and design features. For materials that obey a power-law creep relation, a constant load yields the same time dependence of energy release rate (J-integral) at a crack tip or notch root. This equation, shown below, implies that the J-integral monotonically increases with time under load, such that a sub-critical value will eventually overcome a threshold for crack initiation, J_c. Thus, one can find the time under a constant load required to surpass the crack initiation criterion, called the initiation time, t_i:

Thus, this model gives the initiation time as a function of intrinsic material parameters (through J_c, t₀, and d) and the extrinsic loading and geometry dependent contributions (contained in J₀). This closed-form prediction of the initiation time provides a means to evaluate how material and design characteristics can directly interact. For instance, a reduction in material toughness could be offset with either alterations to the design or creep resistance. Figure 11 depicts an example of how crosslinking, which affects toughness, elastic modulus, and creep resistance, can result in a substantial reduction in the initiation time with modest changes in individual material parameters.²⁷

Specifying a minimum value of initiation time as a design requirement would provide an industry standard for safety against crack initiation without undue restriction of flexibility in the development of new components. The above model also suggests future directions of research for material optimization against crack initiation. For instance, it is desirable for UHMWPE to exhibit both a high creep resistance and fracture toughness, while an increased elastic modulus beneficially depresses J₀ for a given applied load. As crosslinking generally both increases creep resistance and depresses toughness, its overall influence on crack initiation could be difficult to predict. Melting crosslinked UHMWPE reduces its crystallinity and stiffness, thus elevating J₀. Some highly crosslinked UHMWPE formulations are oxidatively stabilized without remelting, and these newer formulations exhibit improved fracture toughness and elastic modulus; however, their relative crack initiation performance is yet unknown. The lesson implied by this analysis is that notch fatigue in UHMWPE is likely not only governed by fracture toughness or crack propagation resistance, but could be dominated by viscous and elastic effects, and that improved UHMWPE formulations could exploit this phenomenon.

Fractures have been observed in crosslinked UHMWPE acetabular liners in total hip replacements, and are likely attributable to the adoption of a more flaw-intolerant material in a design containing notches in highly stressed locations of the component. Investigation of similar intact components has revealed that a majority of these types of device designs using the crosslinked formulation of UHMWPE harbored initiated cracks at the same locations where fractures were previously observed to initiate. The prevalence of initiated cracks in these case series recommends the prevention of crack initiation as a means to control fatigue failure in total hip replacements. The time-dependent analytical crack tip model presented here provides a simple framework for evaluating the inherent impact of material or design alterations on crack initiation performance.

The clinical performance of a medical device depends on many factors and an understanding of the structure-property-design relationships is essential for the clinical success of the implant. The clinical evolution of the three systems presented in this work were chosen to illustrate the sophisticated interplay between design, material selection, structural properties, processing and clinical demands; and to illustrate these effects on the performance of medical device implants utilizing polymeric materials.

ACKNOWLEDGEMENTS

The authors would like to thank the NSF and NIH for partial funding of this research. We would like to acknowledge Ms. Huayin Wu for assistance with technical illustration of the first seven figures and Prof. Clare Rimnac for her technical insight and discussions with the authors on this topic.

1. J.I. Kroschwitz, editor, Polymers: Biomaterials and Medical Applications (New York: Wiley, 1989).
2. J. Black, Biological Performance of Materials: Fundamentals of Biocompatibility (New York: Marcel Dekker, 1999).
3. J.B. Park and J.D. Bronzino, Biomaterials: Principals and Applications (Boca Raton, FL: CRC Press, 2003).
4. B.D. Ratner et al., Biomaterials Science: An Introduction to Materials in Medicine (New York: Academic Press, 1996).
5. J.B. Park and R.S. Lakes, Biomaterials: An Introduction (New York: Plenum Press, 1992).
6. D.W. Howie et al., Clin. Orthop. (1993), N. AM. 24:4 571-581.
7. H.-G. Willert and M. Semlitsch, J. Biomed. Materials Res., 11 (1977), p. 157.
8. J.L. Gilbert, C.A. Buckley, and J.J. Jacobs, J. Biomed. Res., 27 (12) (1994), pp. 1533–1544.
9. R.J. Young and P.A. Lovell, Introduction to Polymers (London: CRC Press, 1990).
10. A.H. Matsumoto et al., Cardiovasc. Intervent. Radiol., 16 (1993), p. 135.
11. C.A. Fontirroche and S. Querns, U.S. patent 5, 820,594 (1998).
12. S. Gad, Safety Evaluation of Medical Devices (New York: CRC, 2001), pp. 505–536.
13. S. Bondurant, V.L. Ernster, and R. Herdman, Safety of Silicone Breast Implants (Washington, D.C.: Institute of Medicine National Academy Press, 1999), pp. 253–254.
14. D.A. Kessler, New Engl. J. Med., 326 (1992), pp. 1713–1715.
15. L.R. Holmich et al., Plastic and Reconstructive Surgery, 120 (7) (2007), pp. 62S–69S.
16. “Mentor MemoryGel™ Silicone Gel-Filled Breast Implants Product Insert Data Sheet” Physician Labeling Document (Silver Spring, MD: Food and Drug Administration, 2006)
17. “NIH Consensus Development Conference on Total Knee Replacement” (Bethesda, MD: National Institutes of Health), http://consensus.nih.gov/2003/2003TotalKneeReplacement117html.htm
18. Total Hip Replacement, NIH Consensus Statement, 12 (5) (1994), pp. 1–31.
19. H.G. Willert, J. Biomed. Mater. Res., 11 (2) (1977), pp. 157–164.
20. D.W. Howie, J. Arthroplasty, 5 (4) (1990), pp. 337–348.
21. D.O. O’Connor et al., Transactions of 44th Annual Meeting of the Orthopaedic Research Society (Rosemont, IL: Orthopaedic Research Society, 1999), p. 816.
22. D.A. Baker, A. Bellare, and L. Pruitt, J. Biomedical Materials Research, 66A (2003), pp. 146–154.
23. S.S. Tower et al., J. Bone Joint Surg. Am., 10 (2007), pp. 2212–2217.
24. J. Furmanski et al., “Catastrophic Rim Fracture of Four Highly Cross-linked Acetabular Liners” (Paper presented at the Annual Meeting of the Am. Acad. Ortho. Surg., 2008).
25. W.Y. Shon et al., J. Arthroplasty, 4 (2005), pp. 427–435.
26. J. Furmanski et al., “In vivo Crack Initiation in Retrieved Cross-linked UHMWPE Acetabular Liners,” Transactions of the 55th Annual Meeting of the Orthopaedic Research Society (Rosemont, IL: Orthopaedic Research Society, 2009).
27. J. Furmanski, C.M. Rimnac, and L.A. Pruitt, “Brittle Fatigue Crack Propagation of UHMWPE and Its Implications for Total Joint Replacements” (Paper presented at the 14th International Conference on Deformation, Yield and Fracture of Polymers, Kerkrade, The Netherlands, 2009).
28. J. Furmanski, E. Feest, and L.A. Pruitt, “Static Mode Fatigue of UHMWPE” (Paper presented at the 2nd International Congress on the Mechanics of Biomaterials and Tissues, Kauai, HI, 2007).
29. J.G. Williams, Fracture Mechanics of Polymers (Chichester, U.K.: Ellis Horwood Ltd., 1984).

Lisa Pruitt is a professor in the Mechanical Engineering Department, University of California at Berkeley, Berkeley, CA, and Jevan Furmanski is a post-doctoral fellow at Case Western Reserve University, Cleveland, OH. Prof. Pruitt can be reached at lpruitt@me.berkeley.edu.