Carbon Fiber Reinforced PEEK Optima—A Composite Material

journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]]

Available online at www.sciencedirect.com

www.elsevier.com/locate/jmbbm

Research Paper Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants

Ely L. Steinbergn, Ehud Rath, Amir Shlaifer, Oﬁr Chechik, Eran Maman, Moshe Salai

Orthopaedic Division, Tel-Aviv Sourasky Medical Center, 6 Weizmann Street, Tel-Aviv 64239, Israel article info abstract

Article history: Background: The advantageous properties of carbon ﬁber reinforced polyetheretherketone Received 15 July 2012 (CF-PEEK) composites for use as orthopedic implants include similar modulus to bone and Received in revised form ability to withstand prolonged fatigue strain. 3 September 2012 Methods: The CF-PEEK tibial nail, dynamic compression plate, proximal humeral plate and Accepted 10 September 2012 distal radius volar plate were compared biomechanically (by four-point bending, static torsion of the nail, and bending fatigue) and for wear/debris (by amount of the debris Keywords: generated at the connection between the CF-PEEK plate and titanium alloy screws) to Carbon ﬁber reinforced PEEK commercially available devices. Wear/debris test Results: Four-point bending stress of the tibial nail and dynamic and distal radius plates Biomechanical tests yielded characteristics similar to other commercially available devices. The distal volar 2 Plates plate bending structural stiffness of the CF-PEEK distal volar plate was 0.542 Nm versus 2 Nail 0.376 Nm for the DePuy’s DVR anatomic volar plate. The PHILOS proximal humeral internal locking system stainless steel plate was much stronger (6.48 Nm2) than the CF-PEEK proximal humeral plate (1.1 Nm2). Tibial nail static torsion testing showed similar properties to other tested nails (Fixion, Zimmer and Synthes). All tested CF-PEEK devices underwent one million fatigue cycles without failure. Wear test showed a lower volume of generated particles in comparison to the common implants in use today. Interpretation: Thus, these tested implants were similar to commercially used devices and can be recommended for use as implants in orthopedic surgery. & 2012 Elsevier Ltd. All rights reserved.

1. Introduction one or more discontinuous phases embedded within a continuous phase. The discontinuous phase is usually harder and A composite material consists of two or more constituent stronger than the continuous phase and is called the reinforce- materials of different physical or chemical properties in the ment material, whereas the continuous phase is termed the microtomacrosizerange(104–102). This pattern encompasses matrix (Migliaresi and Alexander, 2004). One of the composites

nCorresponding author. Tel.: þ972 3 6974727; fax: þ972 3 6974546. E-mail address: [email protected] (E.L. Steinberg).

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 2 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]]

currently available in orthopedic use is the carbon fiber reinforced polyetheretherketone (CF-PEEK). Its mechanical, stiffness, strength, toughness and long-term properties have been char- acterized elsewhere (Kurtz and Devine, 2008; Rae et al., 2007; Vakiparta et al (2004); Zhao et al., 2009). CF-PEEK composites show a high utilization of the fiber in stiffness and strength, and a similarly high utilization of the matrix ductility in toughness. The effect of temperature on strength was examined by Carlile et al. (1989) for fiber and matrix dominated lay-ups. Those authors demonstrated that a high proportion of those properties were retained at different temperatures up to 143 1C(theglass transition temperature), whereas the crystalline melt transition temperature (Tm) occurred at around 343 1C. CF-PEEK exhibited high levels of toughness in a wide range of assessment techni- ques, such as fracture mechanics, instrument falling weight impact and post-impact compression. The fatigue test studied Fig. 1 – The four tested Picollo composite carbon fiber 6 lifetimesinexcessofoneyear(10 cycles), enabling cyclic loads reinforced polyetheretherketone (CF-PEEK) items: TN—10 to be accounted for in structural applications (Carlile et al., 1989). mm tibial nail, DCP—dynamic compression plate, PRP— After polymerization, CF-PEEK was shown to be chemically inert proximal humeral plate and DVRP—distal volar radial plate. and insoluble in all conventional solvents at room temperature, with the exception of 98% sulfuric acid (Ha et al., 1997). The radiation stability of crystalline and amorphous CF-PEEK has been studied extensively. Total exposure of high doses of gamma irradiation between and above 10–50 MGy showed some degradation and cross-linking of CF-PEEK. This value is far higher than the irradiation needed for repeated sterilization, i.e., up to four times the 25–40 KGy doses of gamma irradiation in air, or 40–70 KGy in cases of treatment for bone metastasis (Kim et al., 2007; Kwarteng and Stark, 1990; Sakurai et al., 2002; Vaughan and Stevens, 1995). Composite materials started to interest orthopedic sur- geons in the late 1980s, who initially used them for joint replacement and for fracture fixation in cases of trauma (Bradley et al., 1980; Brown et al., 1990; Kurtz and Devine, 2008; Pemberton et al., 1992; Skinner, 1988; Tayton et al., 1982). Composite materials were introduced in the form of spinal cages as adjuvants to spinal fusion for degenerative discs (Brantigan et al., 2000; Brantigan and Steffee, 1993). A composite material was found to facilitate radiographic assessment by being radiolucent irregardless of the used modality, i.e., X-ray, computerized tomography or magnetic Fig. 2 – Clinical case: A 33 years old male was operated on resonance (Schulte et al., 2000). The use of the composites for for a proximal humeral fracture (A). The fracture was total joint replacement, however, currently has more draw- stabilized by a humeral Picollo CR-PEEK plate (B). Fracture backs than advantages due to a high failure rate (Adam et al., healing progress follow-up was facilitated due to the hard- 2002; Akhavan et al., 2006; Jakim et al., 1983; Morscher and ware transparency property. Dick, 1983). We can assume that the high rate of joint replacement failure can be explained by a fibrous capsule (Baker et al., 2004; Pemberton et al., 1992; Rohner et al., formation around FRC implants that prevents their integra- 2005; Tayton et al., 1982). The aim of this study is to evaluate tion to bone (Ballo et al 2010). The first composite plates were the biomechanical properties of the nail and the plates made of thermosetting polymers, such as plastic or epoxy manufactured from CF-PEEK Optima, the Piccolo system that had the disadvantage of the inability to be contoured (Figs. 1 and 2), as well as the resultant wear/debris during a according to bony anatomy, a property that was preserved by fatigue test in a cell test compound. the use of thermoplastic polymeric plates. Additional biomechanical studies to verify which design can be best fitted for orthopedic purposes had been per- 2. Materials and methods formed on different fiber orientations of CF-PEEK bone plates (Fujihara et al., 2003, 2004). Although there are some clinical 2.1. Nail description and animal studies that used carbon fiber plates for trauma, no intramedullary nails or plates were clinically tested The 10 mm composite tibial nail (CTN) was designed according or used with the CF-PEEK Optima composite thus far to the specifications of a classic intramedullary interlocking

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]] 3

nail. It has a slight 10% curve with three screw holes in the proximal part and three screw holes with a 10 mm distance between their centers in the distal part. The proximal screw hole is located 20 mm from the cranial end of the cranial, and the last hole is located 10 mm proximal to the distal end of the nail. The carbon fiber HexTows IM7 (Hexcel corporation, Stanford, CT, USA) used in this setting is a continuous, high- performance, intermediate modulus with a filament count tow of 6 K (6000). Each tow’s cross-sectional area is 0.13 mm2 and its filament diameter is 5.2 mm. The fiber volume is 60% in the composite matrix. Three samples of each new nail were evaluated according to the American Society of Testing and Materials (ASTM) (ASTM F 1264, 2000). The nail’s characteristics were compared to three other commercially available nails: Fixion 8.5/13.5 mm, Synthes URTN 10 mm and Zimmer 10 mm (Fig. 1).

2.2. Plates description Fig. 3 – Single cycle bending test setup for 8-hole Piccolo composite diaphyseal plate. Three types of plates were studied: dynamic compression (DCP), proximal humeral (PHP) and distal volar radial (DVRP). bridge span settings at 125 mm in the middle portion for the 2 lower bars and at 25 mm in the middle portion for the 2 upper 2.2.1. Dynamic compression plate (DCP) bars (Fig. 3). Four-point bending bridge distances were set Five samples of the Piccolo composite DCP that were 4.5 mm following the requirements of ASTM F382, based on plate in thickness and 190 mm long, and one sample of a predicate design and hole locations. The test was performed using the device, the Synthes 4.5 mm locking compression plates (LCP), single axis compression machine (Testometric M350-10K, were evaluated. The Synthes LCP was chosen for the test Rochdale UK). because its holes are smaller than those of the Piccolo composite DCP, and, as such, the latter plate represents the 2.3.2. Static torsion of the nail worst case for the bending test. In addition, the results of According to the ASTM 1264 Annex A2, the test was per- preliminary tests on the Piccolo composite LCP (with the formed using a custom-made torsion adapter mounted on different types of holes characteristic) indicated that the the single axis tension/compression machine (Testometric failure point following static four-point bending tests (accord- M350-10K, Rochdale UK), which allows transformation of ing to ASTM F382 recommendations) occurred at the com- axial load into torsion by the pulling of a cable wrapped on pression portion of the combination hole, further supporting a wheel. The tension and cable displacement versus the load, the decision to test the Piccolo composite DCP. The plate with as measured by the testing machine load cell, were recorded. a thickness of 4.5 mm was evaluated since it was the thinnest The pull test machine speed is set to a constant speed rate of of the diaphyseal plates and therefore represented the worst 5 mm/min, and the maximal torque is the product of the case for testing the bending properties of diaphyseal plates. tension load multiplied by the wheel’s radius. In order to simulate the weakest condition, 5 mm pins were placed in 2.2.2. Proximal humerus plate (PHP) the distal and proximal screw holes closest to the nail’s Five samples of the Piccolo composite PHP (3.7 mm in thick- center and connected to the testing machine’s gripping ness and 190 mm long), and one sample of a predicate device, devices. The expandable classic Fixion tibial nail was con- the Synthes 3.5 mm PHP (Philos) were evaluated. nected directly to the gripping device using the same gauge length. 2.2.3. Distal volar radius plate (DVRP) Five samples of the Piccolo Composite DVRP (2.4 mm in 2.3.3. Bending fatigue test thickness and 90 mm long), and one sample of a predicate At least three implants were used for cyclic loading for each device, the DePuy 2.4 mm DVR anatomic volar plate, were Piccolo composite nail and plate type. Load was determined evaluated. These samples were chosen in accordance with so that the minimal alternating bending load would be at ASTM F 1264 requirements, although the tests were per- least 60% of the static yield bending strength as determined formed under the requirements listed in ASTM F382, which in the four-point bending static test. All implants had to describe test methods for only three specimens. sustain the load for at least 1,000,000 cycles, based on ASTM F1264 requirements, which is equivalent to a one-year period 2.3. Biomechanical characteristics of healing. A load value of 60% or more of the bending strength was adapted to titanium, a common material of 2.3.1. Four-point bending test which predicate devices are made and one that has an

The bending strength and stiffness of the nails and plates endurance limit of 0.5–0.6 su (Titanium Information Group). were evaluated by the four-point bending stress test. The Piccolo composite DCPs and the DVRPs were tested as tests were performed with the nail and plates being held with described above (Sections 2.2.1 and 2.2.3). The tests were

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 4 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]]

performed using a servo hydraulic fatigue testing machine dried with compressed air. All components were immersed in (Instron 8871, Instron UK) (Fig. 3) with a dedicated testing jig. ethanol (70% or 100%) for 5 min, dried under a laminar flow hood and assembled. The assembly was washed with filtered 2.4. Wear/debris test PBS solution that was filtrated using clean 1.0 mm and 0.2 mm filters before use, taking care that the filters were intact and The purpose of this part of the study was to evaluate the had no observable defects, such as tears, holes or flashes possible amount and characteristics of wear/debris which along their perimeters. The filters were cleaned using filtered could be generated at the connection area between the compressed air or high purity Argon (99.99–99.9999). Each CR-PEEK Piccolo composite plate and the titanium alloy filter was weighed three times with an analytical scale. The (Ti–6Al–4V) screws provided with the Piccolo Composite testing device was cleaned using 100% ethanol and dried with System, and to compare the results to those of similar metal all components under laminar flow hood, using lint-free (titanium) devices. Plates were evaluated under the following fabric. After one million cycles, the device was removed from conditions. A broken cross-section femur bone was simulated the testing machine bed. The fluid was filtered under laminar by a polyacetyl (Derlin) elliptic hollow rod with cross- flow, using the pre-test weight 1.0 and 0.2 mm filters. The dimensions of 30 26 mm2. The rod was cut and a 4 mm filters were dried for at least 12 h under a laminar flow hood gap was set with a four-hole 4.5 mm DCP (both a Piccolo or for at least 2 h inside an oven at 60–65 1C. Each dried filter composite plate and titanium plate after fixation). This was weighed three times with an analytical scale, the results relatively large fracture gap may be regarded as a worst case were recorded, and the average was calculated. The results of for mechanical testing as well as when referring to fracture the PBS residues were calculated on the filters and recorded healing, given that large gaps delay the healing process as well. Filters were kept inside the covered Petri dishes for (Fig. 4)(Yamaji et al., 2001). This assembly was placed within further examination. Following fatigue testing, the liquid was a special jig that was connected to a fatigue machine (Instron collected and its contents were analyzed. For each specimen 8871, Instron UK). The jig enables specimen testing while tested, the weight of each empty filter was subtracted from being immersed within a liquid (here, buffered saline solu- the weight of the same filter following test solution filtration, tion [PBS]). The implant jig assembly allows combined alter- and the filters were evaluated using an optical microscope nate loads of both axial and bending loads under a constant and filter photography. static torsion of 2 Nm. The applied load on the plate was calculated according to Schneider et al. (2001) under partial load bearing and was set to (300) N with R¼0.1 for 1,000,000 3. Results cycles following ASTM F1264 Annex A3 at 2 Hz (ASTM F 1264, 2000). Cleaning and drying was performed as follows: the 3.1. Nails parts were placed in a bath with a soap solution for 10 min and then washed with flushing water for one minute. The 3.1.1. Four-point bending test items were transferred to an ultrasonic bath for 5–10 min and As noted above, the nails were tested using a tension/ compression machine at a speed of 5 mm/min, and the results are presented in Table 1. These results complied with the acceptance criteria by exhibiting characteristics similar to other commercially used nails.

3.1.2. Static torsion The tested nails were compared to three other commercially available nails (Fixion expandable, Synthes and Zimmer) as shown in Table 2. Torsion stiffness test results for the composite product were similar to those of the other three nails.

3.1.3. Fatigue test

The nails were tested under a load of 0.7 Fy (Fy evaluated as the static four-point bending test, and 0.7 was chosen as the highest safety margin). The three tested nails had an average F value of 3213.3723.1 N. Therefore, the tested load was set to 2240 N. No nail failure was observed following one million cycles.

3.2. Plates

3.2.1. DCP four-point bending test The average bending strength of the ﬁve Piccolo plates was Fig. 4 – Implant connection to the broken bone construct 19.172.30 M [Nm] in comparison to 16.7 M [Nm] for the model within the testing jig. Synthes 4.5 mm LCP. The bending stiffness (K) and the

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]] 5

Table 1 – Static four-point bending test for the carbon ﬁber reinforced polyetheretherketone (CR-PEEK) Piccolo composite nail in comparison to other commercially available nails.

Nails’ types Structural stiffness [Nm] Bending stiffness [Nm2]

Composite (10 mm) 80.370.6 33.972.8 Fixion IL (8.5–13.5 mm) 6472.9 41.474.4 Synthes (8 mm) (tibia) 43 20.7 Synthes (11 mm) (tibia) 43.8 33.1 Synthes (10 mm) (femur) 107.5 40.7 Zimmer (10 mm) (femur) 75 45

Table 2 – Torsion test results of three nails were compared to the tibial carbon ﬁber reinforced polyetheretherketone (CR-PEEK) nail.

Nails’ types Structural stiffness [Nm] Torsional stiffness [Nm/1]

Composite (10 mm) 15.372.2 1.1 Fixion IL (8.5–13.5 mm) 14.3 0.83 Zimmer (10 mm) (femur) 13.6 1.2 Synthes (10 mm) (femur) 16 1.8

bending structural stiffness (EI) was 210.8711.01 K [N/mm] and 4.170.21 EI [Nm2], respectively, for the Piccolo plates and 172 K [N/mm] and 3.4 EI [Nm2], respectively, for the Synthes plate.

3.2.2. DCP fatigue test Three plates were evaluated for cyclic fatigue under a load corresponding to at least 60% of the bending strength determined in the single cycle bend test. A load value of 60% or more of the bending strength was chosen for the fatigue test based on a comparison to titanium, a common material out of which predicted devices are made. Based on published data by Beardmore (2010), the endurance limit of titanium is about 0.5–0.7 s (ultimate strength). The tests were com- u Fig. 5 – Picollo composite carbon fiber reinforced polyether- pleted with no failure following at least one million fatigue etherketone (CR-PEEK) debris after filtration through a cycles. 0.2 lm filter using an optical magnifying microscope.

3.2.3. PHP four-point bending test The average bending strength of the 5 Piccolo PHPs was 15.271.13 [Nm] in comparison to15.12 [Nm] for the Synthes tests were completed with no failure following at least one Philos plate. The bending structural stiffness was 1.170.10 EI million fatigue cycles. [Nm2] for the Piccolo plates and 6.48 EI [Nm2] for the Synthes Philos plate. 3.3. Wear debris test

3.2.4. DVRP four-point bending test The tested plates and screws, particularly the interconnected The average bending strength of the 5 Piccolo DVRPs was area and debris, were visually inspected and assessed with 7 3.88 0.41 M [Nm] in comparison to 3.24 [Nm] for the DePuy the aid of an optical microscope (Figs. 5 and 6). The test jig DVR anatomic volar plate. The bending structural stiffness container contents were collected and weighed after the 7 2 was 0.542 0.082 EI [Nm ] for the Piccolo composite DVRP and fatigue test was completed. Each weighing procedure was 2 0.376 EI [Nm ] for the DePuy DVR anatomic volar plate. repeated three times, the average and standard deviation were calculated, and they are presented in Table 3 for 1 mm 3.2.5. DVRP fatigue test filters and Table 4 for 0.2 mm filters. In order to derive the net The same protocol was applied as for the diaphyseal plate weight (E) of the tested sample debris, the average weight of (Section 2.3.3). Three plates were evaluated for cyclic fatigue the empty filter (B) was subtracted from each total filter under a load corresponding to at least 60% of the bending weight (A) following filtration. The average weight of the strength as determined in the single cycle bend test. The PBS residues (D) was also subtracted from the total weight.

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 6 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]]

The results showed a significant difference between both generated at the connection between the CF-PEEK plate and collected debris filters, thus indicating that the CF-PEEK had a the titanium alloy screws. The biochemical and biological lower debris weight than the titanium control. properties of various composite materials had been extensively investigated and reported earlier (Brown et al., 1990; Carlile et al., 1989; Ha et al., 1997; Kwarteng and Stark, 1990; 4. Discussion Kurtz and Devine, 2008; Migliaresi and Alexander, 2004; Rae et al., 2007; Skinner, 1988). The biomechanical properties of The introduction of new technology or devices requires the composite materials were evaluated as well (Fujihara transparent reporting of biochemical, biological and biome- et al., 2003, 2004), but for other composite combinations and chanical properties. In this study, we assessed the biomecha- for other plate designs. nical properties of devices produced from carbon fiber In the present study, we evaluated the biomechanical reinforced polyetheretherketone (CF-PEEK). We also con- properties of a tibial 10 mm nail, a 4.5 mm thick and 190 mm ducted tests for wear by evaluating the amount of the debris long DCP, a proximal humeral 3.7 mm thick and 190 mm long plate, and a volar radial 2.4 mm thick and 90 mm long plate, all made of CF-PEEK Optima. Our results we compared to those of three commercially available similarly designed devices in current use in the clinical orthopedic setting. The 10 mm tibial nail passed the four-point bending test successfully by yielding similar results to those of the commercially available nails. The same diameter and material composition is the base for the other nails’ line, as humeral or femoral nails. The results of the static torsion test that was performed for the studied and the control nails were found to comply with the acceptance criteria of the ASTM standards. The studied nails passed the fatigue test that included the one million cycles without any visual signs of failure under a 2240 N load at a frequency of 5 Hz. The Fig. 6 – Titanium debris in Fig. 4 after filtration through a selected force was derived from the static four-point bending 0.2 lm filter using an optical magnifying microscope. test multiplied by 0.7 as being the highest safety margin.

Table 3 – Wear/debris weight following ﬁltration by 1 lm ﬁlters.

(A)1 mm filters average (B)1 mm clean (C)Wear debris (D)PBS residues average (E)Wear debris Sample weight after test solution filters average weight weight after 1 mm Weight filtration [mg] weight [mg] (C¼AB) [mg] filtration [mg] (E¼ABD) [mg]

#1 84 78.4 5.6 2.35 #2 151.3 148.3 3.0 –0.25a #3 128.1 126.1 2.0 –1.25a Average – – 3.53 3.25 0.78b Titanium 137.1 128.5 8.6 5.35

a The negative weight is probably the result of the analytical weight resolution and accuracy. The calibrated analytical weight used in the test is accurate within 0.1 mg, therefore absolute uncertainty in the observed weight is70.1 mg. b The negative values were calculated as 0 for the calculation of the average result.

Table 4 – Wear/debris weight following 0.2 lm ﬁlters ﬁltration.

(A)0.2 mm filters average (B)0.2 mm clean (C)Wear debris (D)PBS residues average (E)Wear debris Sample weight after test solution filters average weight weight after 0.2 mm weight filtration [mg] weight [mg] (C¼AB) [mg] filtration [mg] (E¼ABD) [mg]

#1 112.7 112.5 0.2 1.2a #2 103.0 101.2 1.8 0.4 #3 114.7 111.5 3.2 1.8 Average – – 1.73 1.4 0.73b Titanium 115.5 110.5 5.0 3.6

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]] 7

The CF-PEEK composite materials have no plastic deforma- titanium debris, and it can be reasonably assumed that its tion but rather only elastic ones, therefore failure is defined use may potentially provide a lesser likelihood of inducing when the bending test is terminated by a resultant plate local inflammation. The wear values we obtained for the breakage, which is expressed as the yield load value. The composite material are comparable with those reported in four-point bending and bending fatigue tests were evaluated the literature, especially by Kinbrum (2009a) who derived for the DCS, the PHP and the DVRP. The bending test results them from experiments that used CF-PEEK in orthopedic obtained for the DCP and the DVRP were almost equivalent to implants (hip and knee replacement). The wear test for those obtained for the controls. Specifically, LCP 4.5 mm the CF-PEEK acetabular cup recorded an average rate of stainless steel plate for the DCP, and Titanium 2.4 mm plate 1.16 mm3/million cycles compared to reported values of for the Volar radial plate. The results of the bending strength 48.2 mm3/million cycles for conventional metal on ultra- for the PHP were inferior to the 3.5 mm Synthes Philos high-molecular-weight-polyethylene (UHMWPE) joints and stainless steel plate by 20%, but nevertheless sufficient for 4.5 mm3/million cycles for cross-linked polyethylene. Also, the purpose of humeral fracture fixation. Specifically, accord- additional testing of CF-PEEK as a knee-bearing material ing to published data, the expected load applied on a produced medial and lateral wear rates of 1.70 and humerus bone of an individual who is 170 cm tall and weighs 1.02 mm3/million cycles, respectively. These results are com- 80 kg is 52 N, which is substantially smaller than the load parable with the published values of 6.69 and 2.98 mm3/million that can withstand a bending test with an average value of cycles obtained for a unicondylar knee test manufactured of 667 N (Ullian et al., 2008). UHMWPE against CoCrMo (Cobalt–Chrome–Molibden), and All the tested plates passed the bending fatigue test 6mm3/year obtained in a study investigating medial Oxford successfully, completing at least one million cycles at a load Partial Knee bearings retrieved from patients (Kinbrum, 2009b). as high as 60% or more of the plates’ bending strength. As Attempts to modify implant microbiology were studied indicated in the ASTM F 382 Appendix X3 (Rationale for because periprosthetic infection related to biofilm formation Annex A2), 106 cycles for estimating plate fatigue strength on implants is of considerable concern. One option was are considered conservative since no bone plate in clinical chitlac coating of fiber-reinforced composites that was found service would normally be expected to withstand 106 high to have antimicrobial activity and to be stable for highly stress loading cycles. There were no observable signs of concentrated lysozyme or H2O2 (Nganga et al., 2012). failure after careful visual inspection of the plates following Implants exposed to long-term moist conditions can lose the completion of the fatigue test. Therefore, it was con- some of their biomechanical properties, as demonstrated cluded that fatigue testing of the weakest (the distal volar in a study on fiber-reinforced composites (Vallittu, 2007). radius) and the strongest (the diaphyseal) Piccolo composite The need for flawless trauma implants is limited to the plates is sufficient, and that further testing of the PHP is not healing period. After fracture healing, the bone no longer necessary. needs support and the implant can be removed. Further The results of the bending and bending fatigue evaluations studies are needed for the implants that carry constant load, demonstrated that the composite material Piccolo CF-PEEK such as vertebral cages or joint prosthesis, and those that are Optima passed these tests successfully, thus complying with exposed to long-term moist conditions. the acceptance criteria and behaving similarly to other commercially available nails and plates in these capabilities. The accumulated debris on the 1 mm filters weighed very little, i.e., an average of 0.78 mg of the CF-PEEK material, in 5. Conclusion comparison to 5.35 mg of the titanium sample. In order to evaluate a worst case scenario, we assumed that all the Based on the results of this study, the Piccolo CF-PEEK Optima particles of debris originated from the CF-PEEK plate and devices provide the recommended requirements for strength not from the titanium screws (although microscopic exam- and wear for new devices and are safe and effective for ination of the filter revealed titanium particles): the specific intended use in humans. gravity of CF-PEEK is smaller than that of titanium, resulting in higher volumetric wear. Taking 1.55 mg/mm3 as the average specific CF-PEEK gravity, the obtained composite debris weight that resulted from the wear test was 0.503 mm3 and Acknowledgment the volumetric wear rate was 0.503 mm3/million cycles. In comparison, the weight of the titanium debris particles that Esther Eshkol is thanked for editorial assistance. originated from the control sample and had been accumu- references lated on the 1 mm filter was 5.35 mg, i.e., about 6 times heavier than the Piccolo composite plate-screw samples. Converting this to volume, it emerges that 1.189 mm3 of titanium debris were generated from testing the predicate Adam, F., Hammer, D.S., Pfautsch, S., Westermann, K., 2002. Early device sample (titanium specific gravity is 4.5 mg/mm3), failure of a press-fit carbon fiber hip prosthesis with a smooth surface. Journal of Arthroplasty 17, 217–223. resulting in a volumetric wear rate of 1.189 mm3/million Akhavan, S., Matthiesen, M.M., Schulte, L., Penoyar, T., Kraay, M.J., cycles. The same results were observed for debris obtained Rimnac, C.M., et al., 2006. Clinical and histological results from the 0.2 mm filters. The volumes generated for the related to a low-modulus composite total hip replacement composite material in this study are far lower than the stem. Journal of Bone and Joint Surgery 88, 1308–1314.

Please cite this article as: Steinberg, E.L., et al., Carbon ﬁber reinforced PEEK Optima—A composite material biomechanical properties and wear/debris characteristics of CF-PEEK composites for orthopedic trauma implants. Journal of the Mechanical Behavior of Biomedical Materials (2012), http://dx.doi.org/10.1016/j.jmbbm.2012.09.013 8 journal of the mechanical behavior of biomedical materials ] ( ]]]]) ]]]– ]]]

ASTM F 1264, 2000. Standard specifications and test methods for Introduction to Materials in Medicine 2nd edition Elsevier metallic angled orthopedic fracture fixation devices. Annual Academic Press, San Diego, pp. 181–195. Book of ASTM Standards 13, 48–54. Morscher, E.W., Dick, W., 1983. Cementless fixation of isoelastic Baker, D., Kadambande, S.S., Alderman, P.M., 2004. Carbon fiber hip endoprosthesis manufactured from plastic materials. plates in the treatment of periprosthetic fractures. Injury 35, Clinical Orthopaedics and Related Research 176, 77–87. 596–598. Nganga,S.,Travan,A.,Donati,I.,Crosera,M.,Paoletti,S.,Vallittu,P.K., Ballo, A.M., Na¨rhi, T.O., Akca, E.A., Ozen, T., Syrja¨nen, S.M., 2012. Degradation of silver-polysaccharide nanocomposite in Lassila, L.V.J., Vallittu, P.K., 2010. Prepolymerized vs. in situ- solution and as coatings on fibre reinforced composites by polymerized fiber-reinforced composite implants—a pilot lysozyme and hydrogen peroxide. Biomacromolecules 13, study. Journal of Dental Research 90, 263–267. 2605–2608. Beardmore, R., 2010. Tables, schedules and formula related to Pemberton, D.J., Mckibbin, B., Savage, R., Tayton, K., Stuart, D., mechanical engineering and engineering materials, Roymech 1992. Carbon–fiber reinforced plates for problems fractures. website: /http://www.roymech.co.uk/Useful_Tables/Fatigue/ Journal of Bone & Joint Surgery 74, 88–92. Fatigue.htmlS. Rae, P.J., Brown, E.N., Orler, E.B., 2007. The mechanical properties Bradley, J.S., Hasting, G.W., Johnson-Nurse, C., 1980. Carbon fibre of poly (ether-ether-ketone) (PEEK) with emphasis on the large reinforced epoxy as high as a high strength, low modulus compressive strain response. Polymer 48, 598–615. material for internal fixation plates. Biomaterials 1, 38–40. Rohner, B., Wieling, R., Mager, F., Schneider, E., Steiner, A., 2005. Brantigan,J.W.,Steffee,A.D.,Lewis,M.L.,Quinn,L.M.,Persenaire,J.M., Performance of a composite flow moulded carbon fibre rein- 2000. Lumbar interbody fusion using the Brantigan I/F cage for forced osteosynthesis plate. Veterinary and Comparative posteriorlumbarinterbodyfusion and the variable pedicle screw Orthopaedics and Traumatology 18, 175–182. placement system. Spine 25, 1437–1446. Sakurai, H., Hayakawa, K., Mitsuhashi, N., Tamaki, Y., Nakayama, Y., Brantigan, J.W., Steffee, A.D., 1993. A carbon fiber implant to aid Kurosaki, H., et al., 2002. Effect of hyperthermia combined with interbody lumbar fusion: two-year clinical results in the first external radiation therapy in primary non-small cell lung cancer 26 patients. Spine 18, 2106–2117. with direct bony invasion. International Journal of Hyperthermia Brown, S.A., Hasting, R.S., Mason, J.J., Moet, A., 1990. Character- 18, 472–483. ization of short-fiber reinforced thermoplastic for fracture Schneider, E., Michel, M.C., Genge, M., Zuber, K., Ganz, R., Perren, fixation devices. Biomaterials 11, 541–547. S.M., 2001. Lads acting in an intramedullary nail during fracture Carlile, D.R., Leach, D.C., Moore, R., Zahlan, N., 1989. Mechanical healing in the human femur. Journal of Biomechanics 34, properties of the carbon fiber/PEEK composite APC-2/AS-4 for 849–857. structural application. In: Newaz, G.M. (Ed.), Advances in Schulte, M., Schulthiess, M., Hartwig, E., Wilke, H.J., Wolf, S., Thermoplastic Matrix Composite Materials. ASTM STP 1004. Sokiranski, R., et al., 2000. Vertebral body replacement with American Society for Testing and Materials, Philadelphia a bioglass–polyurethane composite in spine metastases— pp. 199–222. clinical, radiological and biomechanical results. European Fujihara, K., Huang, Z.M., Ramakrishna, S., Satknanantham, K., Spine Journal 9, 437–444. Hamada, H., 2003. Performance study of braided carbon/ Skinner, H.B., 1988. Composite technology for total hip arthro- PEEK composite compression bone plates. Biomaterials 24, plasty. Clinical Orthopaedics and Related Research 235, 2661–2667. 224–236. Fujihara, K., Huang, Z.M., Ramakrishna, S., Satknanantham, K., Tayton, K., Johnson-Nurse, C., Mckibbin, B., Bradly, J., Hasting, G., Hamada, H., 2004. Feasibility of knitted carbon/PEEK compo- 1982. The use of semi-rigid carbon–fiber-reinforced plastic sites for orthopedic bone plates. Biomaterials 25, 3877–3885. plates for fixation of human fractures. Results of preliminary Ha, S.W., Kiurch, M., Birchler, F., Eckertb, K.L., Mayer, J., Winter- trials. Journal of Bone & Joint Surgery 64, 105–111.

mantel, E., et al., 1997. Surface activation of polyetherether- Titanium Information Group. Titanium Alloys—Ti6AI4V Grade 5. ketone (PEEK) and formation of calcium phosphate coatings An AZoNetwork Website (materials, applications, /http:// by precipitation. Journal of Materials Science: Materials in www.azom.com/article.aspx?ArticleID=1547S). Medicine 8, 683–690. Ullian, V., Mazzer, N., Barbieri, C.H., Moro, C.A., Oliveira, L.A.A., Jakim, I., Berlin, C., Sweet, M.B., 1983. RM isoelastic total hip 2008. Primary stabilization of human shaft fractures: an arthroplasty. A review of 34 cases. Journal of Arthroplasty 3, experimental study of different osteosynthesis methods. Acta 191–199. Ortope´dica Brasileira 16, 8–12. Kim, S., Chun, M., Wang, H., Cho, S., Oh, Y.T., Kang, S.H., et al., Va¨kiparta, M., Yli-Urpo, A., Vallittu, P.K., 2004. Flexural properties 2007. Bone metastasis from primary hepatocellular carci- of glass fibre reinforced composite with partially biodegrad- noma: characteristic of soft tissue formation. Cancer Research able polymer matrix. Journal of Materials Science: Materials in and Treatment 39, 104–108. Medicine 15, 7–11. Kinbrum, A., 2009a. Taking a PEEK at material options for Vallittu, P.K., 2007. Effect of ten years of in vitro aging on the orthopedics. Medical Design Technology, January, 26–29. flexural properties of fiber-reinforced resin composites. Inter- Kinbrum, A., 2009b. The PEEK of large joint performance? New national Journal of Prosthodontics 20, 43–45. PEEK yields lower wear and reduced immunological reaction Vaughan, A.S., Stevens, G.C., 1995. On crystallization, morphology for knee bearings. Orthopedic Design & Technology 03, March, and radiation effects in poly(ether ether ketone). Polymer 36, 1–3. 1531–1540. Kwarteng, K.B., Stark, C., 1990. Carbon fiber reinforced PEEK (APC- Yamaji, T., Ando, K., Wolf, S., Augat, P., Claes, L., 2001. The effect 2/AS-4) composites for orthopedic implants. SAMPE Quarterly of micromovement on callus formation. Journal of Orthopedic 122, 10–14. Science 6, 571–575. Kurtz, S.M., Devine, J.N., 2008. PEEK biomaterials in trauma, Zhao, D.S., Moritz, N., Laurila, P., Mattila, R., Lassila, L.V.J., orthopedic, and spinal implants. Biomaterials 28, 4845–4869. Strandberg, N., Ma¨ntyla¨, T., Vallittu, P.K., Aro, H.T., 2009. Migliaresi, C., Alexander, H., 2004. Classes of Materials Used in Development of a biomechanically optimized multi-compo- Medicine—Composites 2004. In: Ratner, B.D., Hoffman, A.S., nent fiber-reinforced composite implant for load-sharing Schoen, F.J., Lemons, J.E. (Eds.), Biomaterials Science—An conditions. Medical Engineering & Physics 31, 461–469.