**Musculoskeletal Multibody Simulation Analysis on the Impact of Patellar Component Design and Positioning on Joint Dynamics after Unconstrained Total Knee Arthroplasty**

#### **Maeruan Kebbach 1,\* , Martin Darowski 1, Sven Krueger 2, Christoph Schilling 2, Thomas M. Grupp 2,3, Rainer Bader <sup>1</sup> and Andreas Geier 1,4**


Received: 28 February 2020; Accepted: 16 May 2020; Published: 21 May 2020

**Abstract:** Patellofemoral (PF) disorders are considered a major clinical complication after total knee replacement (TKR). Malpositioning and design of the patellar component impacts knee joint dynamics, implant fixation and wear propagation. However, only a limited number of studies have addressed the biomechanical impact of the patellar component on PF dynamics and their results have been discussed controversially. To address these issues, we implemented a musculoskeletal multibody simulation (MMBS) study for the systematical analysis of the patellar component's thickness and positioning on PF contact forces and kinematics during dynamic squat motion with virtually implanted unconstrained cruciate-retaining (CR)-TKR. The patellar button thickness clearly increased the contact forces in the PF joint (up to 27%). Similarly, the PF contact forces were affected by superior–inferior positioning (up to 16%) and mediolateral positioning (up to 8%) of the patellar button. PF kinematics was mostly affected by the mediolateral positioning and the thickness of the patellar component. A medialization of 3 mm caused a lateral patellar shift by up to 2.7 mm and lateral patellar tilt by up to 1.6◦. However, deviations in the rotational positioning of the patellar button had minor effects on PF dynamics. Aiming at an optimal intraoperative patellar component alignment, the orthopedic surgeon should pay close attention to the patellar component thickness in combination with its mediolateral and superior–inferior positioning on the retropatellar surface. Our generated MMBS model provides systematic and reproducible insight into the effects of patellar component positioning and design on PF dynamics and has the potential to serve as a preoperative analysis tool.

**Keywords:** joint replacement; knee joint; total knee arthroplasty; patellar component; musculoskeletal multibody simulation; patellofemoral joint

#### **1. Introduction**

Total knee replacement (TKR) is an established and effective surgical procedure for progressive osteoarthritis. TKR is currently performed over 700,000 times a year in the USA and this number is

expected to grow exponentially worldwide [1,2]. Nevertheless, the rate of satisfied patients is 80%, which is rather low [2–4]. The patellofemoral (PF) joint represents a crucial part after total knee arthroplasty, and persistent PF pain remains a common postoperative complication with or without patellar resurfacing [5–7]. Complications include anterior knee pain, patellar maltracking, fracture, and patellar component loosening [8–10]. In this context, patella resurfacing is an important intraoperative factor: in the USA, more than 80% of primary TKRs are performed with this technique [11]. During surgery, the accurate positioning of the patellar component remains challenging because intraoperative alignment involves considerable inaccuracies [5,9,12,13]. The incidence of PF disorders ranges from 7% to 30% after a minimum of two postoperative years [14]. Patellar component malalignment is related to increased retropatellar loading and abnormal patellar kinematics [15,16]. Regarding the high number of intraoperative parameters, the dynamic interplay of the articulating joint partners of PF and tibiofemoral joint need to be analyzed in a systematic and reproducible manner [15,17,18].

Clinical observations alone cannot entirely explain unsatisfactory patient outcomes as they are often limited to the retrospective analysis of the influence of intraoperative parameters on pain and functional outcome [5,7,11–13,19,20]. However, clinical observations could be correlated with the mechanical loading of the joint to identify the underlying biomechanical causative chain. Therefore, knowledge of the biomechanical influence of intraoperative positioning of the patellar component during typical movements of daily living is essential to understand the underlying mechanical causes of implant failure and to improve the postoperative outcome. In this regard, the squat motion is known as one of the most dissatisfying motions after TKR [19].

Despite improvements in surgical instruments and techniques, many of the causes of revisions and patient dissatisfaction are directly related to implant component malalignment [6,18,21,22]. The morphology of the knee joint differs between both genders and different ethnicities [23–26]. These differences cannot only affect the implant component size selection but also the trochlear groove geometry, as well as the positioning of the patellar component between patients. Likewise, interindividual morphological joint variation is known to influence knee joint dynamics [27] and functionality of TKRs [26,28]. In this regard, Chen et al. [29] showed that the internal rotation of the femoral component and the varus malpositioning of the tibial component led to unfavorable postoperative loading conditions. Moreover, the malpositioning of the implant components often results in the overloading of the articulating joint compartments [18,30,31]. Different designs (dome-shaped, modified dome or anatomic design) and positions of the patellar button have been shown to affect knee joint dynamics [6,28,32–37]. An increased patellar component size was reported to affect the patellofemoral kinematics and postoperative outcome [38,39]. For instance, variations in the mediolateral and superior–inferior position of the patellar component resulted in different PF contact forces and kinematics [6,28,30,32,33,40,41]. Furthermore, the PF dynamics is influenced by the patellar component design [6,28] and the considerable impact of the patellar component thickness on PF kinematics and knee flexion has been widely reported [6,21,35,42–45]. For example, Abolghasemian et al. [43] reported a flexion loss of 1.28◦ with every millimeter of increased patellar component thickness. Bracey et al. [42] performed a similar study on 10 cadaveric knees and showed a flexion loss of 1.2◦ for each 2-mm increase of the patellar component thickness which resulted in a lateral patellar shift of more than 2 mm and a lateral patellar tilt of more than 4◦ whereas the patellar rotation remained nearly unchanged [43]. Bengs et al. [35] showed for 31 CR-TKRs with four different patellar thicknesses that on average the flexion was decreased by 3◦ for every 2-mm increase of the patellar component thickness but had no major impact on patella subluxation or tilt. Additionally, PF overstuffing has been reported with increased polyethylene wear, PF maltracking, and aseptic loosening of the patellar component [6,35,42,46,47]. Biomechanical studies have analyzed various mechanical aspects of patellar maltracking [15,32,33], e.g., medialization of the patellar component decreased the PF contact force [32,36,48,49] and altered the PF kinematics [36,41,48,50]. In a biomechanical study, Anglin et al. [48] found for a 2.5 mm component medialization a mean change in

lateral patellar shift of 1.9 mm and a mean change in lateral tilt of 3.2◦. These studies have contributed to understanding the positioning and design of the patellar component from a biomechanical perspective.

However, most experimental studies have only investigated passive knee flexion without active muscle forces, simulated rather uncommon motion patterns, commonly the knee rig configuration [28,33,42,51], or assumed a static quadriceps force, usually some predefined maximum value [28]. Furthermore, only a few studies addressing the biomechanical impact of different patellar component positions and designs on PF dynamics could be identified [15,18,28,33]. Instead, studies have revealed that the optimum position of the patellar component remains controversial [5,28,33,43,49], although evidently important for the postoperative outcome [28]. Some research groups have recommended medialization [8,41,48,50,52], while others have suggested centralization [32]. Moreover, biomechanical studies analyzing the superior–inferior position of the patellar component have reported contradictory findings [5,32,33,48], indicating a lack of understanding of this issue [5,49]. Hence, the PF joint mechanics have not yet been sufficiently quantified and understood so far [5,33]. The effects of various surgical parameters remain unclear and somewhat controversial. Using computational models, the influence of surgical and implant design parameters on knee joint dynamics can be investigated more comprehensibly [29,33,53–56].

Therefore, the current computational study aimed to systematically analyze and determine the biomechanical impact of patellar component malpositioning and design on PF dynamics during a dynamic squat motion using musculoskeletal multibody simulation (MMBS) in which a detailed knee joint model resembled the loading of a virtually implanted unconstrained cruciate-retaining (CR)-TKR including a dome patellar button. The findings could contribute to improving surgical techniques, preventing postoperative complications, and reducing wear propagation.

#### **2. Materials and Methods**

Our MMBS model is based on the experimental dataset of the *4th Grand Challenge Competition to Predict In Vivo Knee Loads* [57], which is a standardized dataset used by the research community to validate musculoskeletal models. This dataset comprises the CT scans (pre- and post-op) of a male subject (age = 88 years, height = 168 cm, and weight = 66.7 kg) who underwent TKR due to primary osteoarthritis and received an instrumented cruciate-retaining (CR) TKR (P.F.C. Sigma, DePuy Synthes, Raynham, MA, USA). This implant design was imported without changing the original standard size. The implanted TKR has a first-generation tray design eKnee and allows measuring the in vivo tibiofemoral contact forces during activities of daily living using an integrated telemetric force sensor. Concerning the knee implant, it represents a fixed-bearing, unconstrained bicondylar design which was implanted into the right knee of the patient. The femoral component had an asymmetrical dual radius design and was composed of cobalt–chromium alloy. The tibial component was composed of titanium alloy. Regarding the patellar component, an all-polyethylene dome-shaped component with three fixation pegs was used. The dataset enables the computational reconstruction of the 3D bone segments of the lower right extremity (pelvis, femur, patella, tibia, fibula, and pes), as well as the implant component geometries (femoral component, patellar button, tibial insert, and tibial tray). Moreover, motion capturing marker trajectories for activities of daily living are included, resulting in a comprehensive database for kinematic and kinetic model validation. A detailed description of this dataset can be found in Fregly et al. [57].

#### *2.1. Overview of the Deployed Workflow of Musculoskeletal Multibody Simulation*

The workflow for developing our subject-specific MMBS model for the computational analysis of the PF joint after TKR is depicted in Figure 1. Briefly, the relevant bone geometries were reconstructed from preoperative CT scans [57,58], which allowed the virtual implantation of TKR implant geometries. The implementation of the musculoskeletal geometry was based on the data reported in the TLEM 2.0 anthropometric database [42]. Additionally, the origin and insertion of muscles and ligaments were verified by an experienced orthopedic surgeon. Contacting surfaces were modeled by means of

the polygonal contact model [59] to enable physiological-like joint dynamics. In this regard, relevant ligament structures were modeled as sets of nonlinear force elements [60,61] to resemble the respective anatomy of the ligaments. The inverse kinematics analysis was performed on a realistic squatting motion as reported in the SimTK data set [57], in which recorded marker trajectories applied to a patient allowed the calculation of the relative joint coordinates in the MMBS model deploying a global optimization procedure [62,63]. Finally, an inverse dynamics analysis coupled with a static optimization [55,62,64–66] allowed the computation of the individual muscle forces for the forward dynamic prediction of knee joint dynamics, e.g., tibiofemoral and PF contact forces, tibiofemoral and PF kinematics, as well as muscle and ligament forces under the presence of surgical parameter variations related to TKR.

**Figure 1.** Workflow for generating the musculoskeletal multibody simulation model of the lower extremity with a total knee replacement. The illustration marked with \* was taken from [57]. Permission to publish is granted under a CC BY open access license.

#### *2.2. Musculoskeletal Multibody Simulation Model with a Cruciate-Retaining Total Knee Replacement*

The MMBS model for the detailed analysis of the PF joint was generated based on a previously presented and validated MMBS model of the lower right extremity [65], Figure 2A,B. Implant and bone geometries, as well as relevant soft tissue structures, were modeled in the multibody software SIMPACK (V9.7, Dassault Systèmes Deutschland GmbH, Gilching, Germany). A variant of the computed muscle controller (CMC) with static optimization for individual muscle force prediction was implemented in MATLAB/Simulink® (v8.1, 2013a, The MathWorks Inc., Natick, MA, USA) and interfaced the MMBS model via TCP/IP-communication for forward dynamic co-simulation. All simulations were performed on an off-the-shelf computer (Intel® Xeon E5-1650 v4 CPU @3.60 GHz, 32 GB RAM).

The reconstructed bone segments were mutually connected by ideal joints as described in [62,65]. The respective centers of rotation of the ideal joints were determined by fitting cylinders or spheres into the cartilage surfaces of the articulating joint compartments. To ensure physiological-like roll-glide dynamics in the knee joint, both tibio- and patellofemoral joint compartments were modeled with six degrees of freedom (DoF) by implementing a polygon contact model [59], thereby resembling the complex articulation of the freeform implant surfaces. The lower left extremity was modeled as a symmetry condition in the sagittal plane. More precisely, the movement of the pelvis perpendicular to the sagittal plane was restrained by a spring-damper force element connecting the symmetry plane of the pelvis to the sagittal plane, thereby representing the lower left extremity during a symmetrical squat motion [65]. Regarding the squat motion, the patient started from a standing position and the range of knee flexion was about 0◦–90◦. The mass properties of the bone segments and soft tissue structures were calculated using regression equations as a function of the patient body weight [67].

**Figure 2.** The developed musculoskeletal multibody simulation model with a cruciate-retaining total knee replacement in the lower right extremity during a dynamic squat motion combining musculoskeletal motion dynamics, knee implants with articular contact definitions, muscles, and ligaments (**A**). Detailed representation of the knee joint with implant components and muscle structures, including muscle wrapping. Note that ligaments are not shown for the sake of clarity (**B**). Investigated implant components with ligament structures of the tibio- and patellofemoral joint (**C**).

For the description and comparison of the joint dynamics, standardized coordinate systems were established [68–70] which further allowed the identification of attachment points of relevant muscle and ligament structures as described in *Twente Lower Extremity Model 2.0* anthropometric database [71]. Additionally, the attachment sites of muscles and ligaments were verified by an experienced orthopedic surgeon. Muscles were implemented in the form of unidimensional Hill force elements [55,71,72]. The muscles were further subdivided into several structural bundles based on the attachment area to account for the wide attachment surfaces [71]. Muscle deflection phenomena around the bones were incorporated either as segment-fixed via points or using wrapping surfaces [73] where required.

For a physiological representation of the knee joint dynamics, the MMBS model comprised, next to the explicit contact surface modeling of both the joint compartments, the implementation of all relevant ligamentous soft tissue structures (Figure 2C) with nonlinear force-strain relation (as nonlinear springs) [60,61]. Precisely, we implemented the posterior cruciate ligament (PCL), medial collateral ligament (MCL), lateral collateral ligament (LCL), oblique popliteal ligament (OPL), arcuate popliteal ligament (APL), posterior capsule (pCAP), medial patellofemoral ligament (MPFL), lateral patellofemoral ligament (LPFL), and the patellar ligament (PL) according to their anatomic descriptions, i.e., all ligaments were modeled as bundles of strands extending between the origin and insertion as described in Smith et al. [74]. The force elements for the representation of the ligaments followed a nonlinear elastic characteristic with a slack region as reported by Blankevoort et al. [61]. As it concerns the ligament parameterization, we first generated a MMBS model resembling the passive knee flexion, since the passive knee joint dynamics majorly depend on the ligaments and their parameterization (initial parameter sets were taken from the literature [61,74]). Then, by repeatedly simulating the knee flexion motion, the ligament parameters were adjusted with respect to the joint quantities for validation of the passive knee flexion. Once the appropriate ligament parameters were identified, the very same, parameterized ligament apparatus was transferred to the MMBS model resembling the squat motion and model validation was performed as described below (Section 2.5. *Validation of the musculoskeletal multibody simulation model*). Due to its high stiffness, the patellar ligament (PL) was modeled as a rigid coupling element between *Apex patellae* and *Tuberositas tibiae* with a fixed length [54,75]. The anterior cruciate ligament was virtually resected as it is sacrificed during CR-TKR surgery. A complete summary of the mechanical material properties for the ligaments [60,61,66,74] is provided in the Supplementary Information (Appendix A, Table A1). The PF joint was characterized by a patellar length of 43.6 mm, a patellar tendon length of 59.7 mm, and a tibial tuberosity-trochlear groove distance of 9.61 mm. The anthropomorphic details of the joint were in the normal range for Caucasians with native patellae [76,77]. Moreover, the retropatellar surface area after resection was in the normal range [78].

Hence, the developed MMBS model of the knee joint, allows for the systematic and reproducible analysis of the TF and PF joint dynamics with its emphasis on the detailed representation of the articulating endoprosthesis components and ligamentous structures during dynamic, muscle-induced full-body motion.

#### *2.3. Kinematic Analysis*

The motion capture data comprises the trajectories of reflective skin markers applied to the patient's body [57] as depicted in Figure 1. These marker trajectories were used as input for an optimization algorithm [79] to derive the generalized joint coordinates *q*, . *q*, .. *q* that describe the timely trajectories of the MMBS model's DoF. More precisely, the skin markers were modeled as moving reference points and coupled to the respective anatomical landmark, i.e., related segment-fixed points on the bone surface using spring-damper force elements. In this manner, the inverse kinematics problem could be resolved by the minimization of the spring-damper potential, thereby minimizing the error between motion capture data and the MMBS model motion. The desired joint trajectories *qd*, . *qd*, .. *qd* were then used as input for the forward dynamic simulation of the muscle-induced squat motion for numerous MMBS model variations as described in the Section 2.6. *In silico study on the e*ff*ect of the patellar component design and positioning on patellofemoral joint dynamics after TKR*. In this manner, we generated a nominal configuration of the MMBS model based on the optimal surgical technique [57] which was further used to predict different postoperative situations of the patellar component design and position. Finally, the implant configurations were verified by an experienced orthopedic surgeon.

#### *2.4. Forward Dynamic Musculoskeletal Multibody Simulation of a Squat Motion*

A variant of the CMC algorithm [66] with static optimization for individual muscle force prediction was implemented to track the desired joint coordinates, as derived from the experimental motion capture data, by generating coordinated muscle forces. This controller has been recently described and verified to enable in vivo knee kinematics and kinetics in patient-specific musculoskeletal models [55,65,66,74,80]. Within CMC, the inverse dynamics model serves for input-output linearization of the neuro-musculoskeletal system, which is further superposed with generic feedback control to accurately track the experimentally obtained motion maneuvers *qd*, . *qd*, .. *qd*. Subsequently, the computed joint torques τ*<sup>m</sup>* necessary to drive the DoFs *q*, . *q*, .. *q* are distributed over the available muscle actuators *f m*,*i* (·) by means of static optimization [64]:

$$\min\_{\mathfrak{a}} l(\mathfrak{a}) \equiv \mathfrak{a}^T V \mathfrak{a}, \text{ subject to } D\mathfrak{a} = \ \tau\_{\mathfrak{m}} \text{ and } 0 \le a\_l \le 1. \tag{1}$$

The aforementioned muscle distribution problem was solved for optimal muscle activation levels *a*∗ *<sup>i</sup>* by minimizing an energy-optimal quadratic cost function *J*(*a*) in which the diagonal weight matrix *V* = *diag*(*V*1, ... , *Vn*) includes the muscle volumes *Vi* of each muscle unit and *ai* is bounded by its physiological limits 0 ≤ *ai* ≤ 1. Moreover, the optimization problem is constrained by the set of linear equations *Da* = τ*<sup>m</sup>* where *D* represents the contribution of the respective muscle to the respective joint as described in a previous study [65]. We deployed a Hill-type muscle of the form:

$$f\_{m,i}(q\_i \dot{q}\_i \ a\_i) = (f\_{l \upsilon i}(s\_i \dot{s}) \ C\_i \ a\_i) \ \text{ } \mathfrak{u}\_i \text{ with } \mathbb{C}\_i = A\_i \ \sigma\_i \text{ and } i = 1, \dots, n,\tag{2}$$

where *n* is the number of muscle units, *flv*,*i*(·) is the force-length-velocity relation with muscle length *s* and muscle contraction velocity . *s*, *Ai* describes the physiological cross-sectional area, *u<sup>i</sup>* is the muscle's unit direction vector, and σ*<sup>i</sup>* is the maximal isometric muscle stress which was set to σ*<sup>i</sup>* = 1 MPa [72]. Note that within this work, the force-length-velocity factor was assumed to *flv*,*i*(·) = 1 as properties of the activated muscle structures and the activation dynamics have little influence on the prediction of

muscle forces [64]. Therefore, the defined muscle force element depends on its theoretical maximum force *Ai* σ*i*, its activation level *ai*, and on the resulting moment arm only.

#### *2.5. Validation of the Musculoskeletal Multibody Simulation Model*

The predicted TF contact forces were validated using in vivo measured knee forces measured during squat motion exercises [57]. Furthermore, the TF contact force as well as the quadriceps force and PF contact force was compared with literature data [81–84] which is in Supplementary Information (Appendix B, Figure A1).

Since the in vivo measurements provided by the *4th Grand Challenge Competition to Predict In Vivo Knee Loads* [57] are limited to the contact forces acting in the TF joint, TF and PF kinematics were validated based on the in vitro and in silico results reported by Woiczinski et al. [85]. They measured TF and PF dynamics of 15 fresh frozen specimens with implanted CR-TKR during a squat motion using a knee rig [85].

#### *2.6. In Silico Study on the E*ff*ect of the Patellar Component Design and Positioning on Patellofemoral Joint Dynamics after TKR*

The validated MMBS model, as described previously, has been used as the nominal configuration corresponding to the optimal surgical technique [57] and served as a reference for subsequent model variations. Accordingly, with respect to the nominal model, we analyzed six patellar component configurations: spin of the patellar component ±5◦, tilt of the patellar component ±5◦, flexion-extension of patellar component ±5◦, superior–inferior positioning of patellar component ±3 mm, and mediolateral positioning of patellar component ±3 mm. The position of the patellar component was changed based on a standardized coordinate system [70]. For instance, the shift was varied by medial or lateral movement of the center of the patellar component along a mediolateral axis defined as fixed to the patella. Moreover, we investigated the most important patellar button design parameter by increasing/decreasing the thickness of the patellar component ±2 mm.

The variations for each parameter configuration (Figure 3) were systematically chosen according to the reported findings from clinical [36,48,49], in vitro [15,16,32], and in silico [28,33] studies. The effect of each configuration was statistically evaluated with respect to the nominal simulation model as a function of the knee flexion angle over the full range of motion of the dynamic squat motion. Note that the calculated forces from the TF contact models were expressed with reference to the tibial component system to allow a direct comparison with the in vivo measured knee forces as reported by Fregly et al. [57].

**Figure 3.** In silico study on the effect of the patellar component designs and positionings. Mediolateral positioning of patellar component ±3 mm, superior–inferior positioning of patellar component ±3 mm, and patellar button design parameter by increasing/decreasing the thickness of the patellar component ±2 mm. The rotational positioning was analyzed by the spin of the patellar component ±5◦, tilt of the patellar component ±5◦, and flexion-extension of patellar component ±5◦.

#### *2.7. Statistical Metrics*

The predicted TF contact force of the MMBS model was compared with the in vivo measured knee force [57]. The prediction accuracy was quantified by mean absolute deviation (MAD) as a measure of magnitude differences, root-mean-square error (RMSE) as a measure for the difference between values predicted by the numerical model and the values actually observed from the experimental setup, Pearson correlation coefficient (*r*2) as a measure of shape differences, and coefficient of determination (*R*2) as a measure of magnitude and shape differences. For clarity, the mathematical definitions of the statistic metrics are provided in the Supplementary Information (Appendix C).

#### **3. Results**

#### *3.1. Validation of the Musculoskeletal Multibody Simulation Model*

The validation of the TF contact force, as well as the quadriceps force and PF contact force using literature data [81–84], is presented in Supplementary Information (Appendix B, Figure A1).

The predicted and in vivo measured medial, lateral, and total TF contact forces during two-leg squat motion are depicted in Figure 4. Their quantification in terms of statistic metrics (Table 1) for the predicted contact forces of the MMBS model was in good agreement with a satisfactory level of accuracy (RMSE = 0.39 body weight (BW), *R*<sup>2</sup> = 0.94, *r*<sup>2</sup> = 0.97) in terms of the magnitude and the general trend. The medial contact force was slightly overestimated (RMSE = 0.35 BW). The predicted TF contact forces in the lateral compartment exhibited an excellent agreement (RMSE = 0.10 BW). Therefore, the MMBS model closely captured the overall pattern and timing (*R*<sup>2</sup> = 0.94, *r*<sup>2</sup> = 0.97) of the in vivo measured TF contact force, indicating sufficient model fidelity for contact force prediction.

**Figure 4.** Musculoskeletal multibody simulation of the patellofemoral joint during the dynamic squat motion (**A**). Model validity was confirmed by comparing the reported lateral (**B**), medial (**C**), and total (**D**) tibiofemoral contact forces (in unit of body weight BW) of the in vivo measurements (blue, [57]) to our predictions (red).



Moreover, the TF and PF kinematics were validated based on the in vitro and in silico results reported by Woiczinski et al. [85]. Specifically, anterior-posterior translation, tibial rotation, patellar tilt, patellar rotation, and patellar shift were compared (Figure 5). The PF and TF kinematics were reproduced with good agreement to the experimental and simulation data [85]. In general, the MMBS model predicted TF and PF kinematic patterns with a reasonable level of accuracy [85]. Overall, good agreement was observed between our MMBS model and reported data.

**Figure 5.** Tibio- and patellofemoral kinematics during dynamic squat motion. In silico (red dotted line) and in vitro (green area) [85] comparison of tibiofemoral and patellofemoral kinematics with kinematics obtained from musculoskeletal multibody simulation (MMBS) model (blue line). Comparison of anterior-posterior tibial translation with reference to the femur (**A**). Tibial internal/external rotation with reference to the femur (**B**). Patellar shift (**C**). Patellar rotation (**D**). Patellar tilt (**E**).

#### *3.2. E*ff*ect of Patellar Component Design and Positioning on Patellofemoral Joint Dynamics after Unconstrained Total Knee Arthroplasty*

PF dynamics was strongly affected by the position and design of the patellar component (Figures 6–10). Generally, the PF contact forces increased with increasing knee flexion angle due to the progressive involvement of the *M. quadriceps femoris* that enables such deep flexion angles. However, depending on the positioning and design of the patellar component, it is possible to reduce or increase the PF contact force. In this context, the analysis showed that the PF contact force was strongly affected by the patellar component thickness (RMSE = 440 N), considerably affected by superior–inferior positioning (RMSE = 199 N), and only moderately affected by mediolateral (RMSE = 98 N) positioning (Figure 10).

The variation in the superior–inferior positioning of the patellar component is depicted in Figure 6. An inferior position of the patellar component led to a reduction of the PF contact force (RMSE = 199 N) by up to 16% (Figure 6A). The maximum PF contact forces of the inferior, reference, and superior positions were 1626 N, 1947 N, and 2048 N, respectively. Superior positioning tended to cause a more medial tilting of the patella during flexion. An inferior position of the patellar component slightly increased the medial patellar shift, whereas a superior position of the patellar component decreased the lateral patellar tilt and shift (Figure 6B,C).

**Figure 6.** Effect of the superior–inferior positioning of the patellar component on patellofemoral dynamics. Effect of superior–inferior positioning of the patellar component on patellofemoral contact force (**A**), patellar shift (**B**), patellar tilt (**C**), and patellar rotation (**D**).

Similarly, the mediolateral positioning of the patellar component clearly affected the PF dynamics. Concerning the PF contact force, medialization and lateralization of the patellar component affected the PF contact forces (Figure 7A), i.e., a medialized patellar component decreased the maximum PF contact force during knee flexion by up to 8%. For instance, at 90◦ knee flexion, the PF contact force of the medial, reference, and lateral positions were 1798 N, 1947 N, and 2082 N, respectively. Furthermore, the mediolateral position of the patellar component clearly influenced PF kinematics regarding patellar shift, patellar tilt, and rotation. Mediolateral positioning of the patellar component shifted and tilted the patella path in the opposite direction, e.g., a medialized patella resulted in a more lateral shift and tilt of the patella and vice versa. Similarly, the patellar tilt was most sensitive to the mediolateral position of the patellar component. The more the patellar component was medialized, the more the patella tilted laterally with respect to the femur (Figure 7C).

**Figure 7.** Effect of the mediolateral positioning of the patellar component on patellofemoral dynamics. Effect of mediolateral positioning of the patellar component on patellofemoral contact force (**A**), patellar shift (**B**), patellar tilt (**C**), and patellar rotation (**D**).

The variation in the thickness of the patellar component dramatically changed the PF dynamics (Figure 8). Generally, lower PF contact force values were seen in thinner components, while with an increase in component thickness, the contact force increased by up to 27% (Figure 8A). For the nominal configuration, the peak PF contact force was 1947 N, and the corresponding value obtained from the other variations were 2480 N (+2 mm), and 1344 N (−2 mm). The contact force was 176% of that of the thinner component. There was a direct relationship between the maximum contact force and patellar component thickness (Figure 8E): an increasing component thickness leads to higher maximum contact forces following a linear trend (R2 = 0.9976). Furthermore, we found that for every 1 mm of increased patellar thickness, PF contact force increased by 13.7% (0.41 BW). There was also an effect of the patellar component thickness on patellar tracking, e.g., an increased patellar component thickness led to an increased lateral patellar tilt (Figure 8C). Contrarily, the effect on the patellar shift was smaller (Figure 8B).

The rotational positioning of the patellar component influenced PF dynamics only slightly in our TKR design (Figure 9). Tilting the patellar component, however, corresponded to a nonsymmetrical rotation of the patellar dome and therefore affected patellar tilt and shift. A medial tilt of the patellar component slightly increased the PF contact force, the patella shifted, and tilted more laterally. Flexion-extension of the patellar component slightly affected the PF contact force, while the spin of the patellar component left the PF dynamics nearly unchanged (Figure 10).

**Figure 8.** Effect of the patellar component thickness on patellofemoral dynamics. Effect of patellar component thickness on patellofemoral contact force (**A**), patellar shift (**B**), patellar tilt (**C**), and patellar rotation (**D**). Relationship of the maximum patellofemoral contact force for different patellar component thickness (**E**).

**Figure 9.** Effect of the tilting of the patellar component on patellofemoral dynamics. Effect of mediolateral positioning of the patellar component on patellofemoral contact force (**A**), patellar shift (**B**), patellar tilt (**C**), and patellar rotation (**D**).

**Figure 10.** Radar chart representing the root-mean-square errors (RMSE) of patellofemoral contact force (**A**), patellar shift (**B**), patellar tilt (**C**), and patellar rotation (**D**) for different patellar component designs and positions: superior–inferior position, mediolateral position, patellar component thickness, tilt, flexion-extension, and spin.

#### *3.3. E*ff*ect of Coupling Patellar Component Design and Positioning Parameters on Patellofemoral Joint Dynamics after Unconstrained Total Knee Arthroplasty*

Additionally to the presented findings, we analyzed the coupling of the most critical patellar component design parameter with all other patellar component positioning parameters, i.e., we combined the increase of the thickness of 2 mm with all other positioning parameters (Figure 11). Similarly, the increase in the thickness combined with the rotational positioning of the patellar component influenced PF dynamics only slightly in our TKR design. However, an increase in the thickness combined with changing the mediolateral position of the patellar component mainly affected the PF kinematics. For instance, a solely increase in the thickness slightly changed the patellar shift (Figure 8B), whereas combined with a medialized patellar component the patellar shifted more laterally (Figure 11E).

**Figure 11.** Effect of the increase of the patellar component thickness by +2 mm with all positioning parameters on patellofemoral dynamics. Effect of patellar component thickness combined with the positioning parameters on patellofemoral contact force (**A**,**B**), patellar tilt (**C**,**D**), patellar shift (**E**,**F**), and patellar rotation (**G**,**H**).

#### **4. Discussion**

In our computational study, we evaluated the effects of different intraoperative positionings and designs of the patellar component on both PF kinematics and kinetics during a dynamic squat motion of a subject with implanted CR-TKR with a dome patellar button. We developed a musculoskeletal knee joint model within a multibody simulation framework capable of analyzing the PF kinematics and kinetics. The results presented imply that even minor variations in the position and design of the patellar component can have a crucial effect on PF dynamics (Figures 6–10). PF contact forces were mostly affected by patellar component thickness (up to 27%), patellar component superior–inferior positioning (up to 16%), and mediolateral patellar component positioning (up to 8%). Concerning PF kinematics, the patellar component thickness as well as the mediolateral position of the patellar component revealed the greatest impact. Based on our findings, we can conclude that the malalignment in mediolateral as well as superior–inferior direction and the thickness of the resurfaced patella are the most important intraoperative parameters affecting PF dynamics. Furthermore, we could show that the translational positioning is more critical than rotational positioning regarding the resulting PF contact force. Generally, our findings are in good agreement with those of previous experimental and clinical studies [5,6,28,32,35,36,48,49,86].

However, similar to other in vitro and in silico studies, our approach has some limitations. Our MMBS does not reflect a diverse population of different patients. Moreover, only the two-leg squat motion was considered. Additional simulations which cover more activities (e.g., rising from a chair, normal gait) will be helpful in the future for a more robust investigation and to support our presented findings as total knee kinematics are task-dependent [87]. However, the investigations were performed for a two-leg squat motion because this motion covers a wider range of knee flexion compared to gait and it is reported to be one of the most dissatisfying motions after total knee arthroplasty [19]. One limitation which should be considered when transferring our in silico findings to clinical practice is associated with ligament modeling as accurate representation of ligament parameters is still challenging, and ligaments' mechanical properties might change during surgery. In the present study, the ligament parameters utilized were obtained from the literature. These assumptions might

affect the prediction of joint dynamics. Likewise, the musculoskeletal geometry was based on the TLEM 2.0 anthropometric database [71]. It can be assumed that changes in the musculoskeletal geometry, e.g., muscle cross-sectional areas or muscle moment arms, of the muscles acting on the knee joint would have an effect on the forces and in turn kinematics of the PF joint. In addition, up to date, MMBS models hardly account for effects related to age or underlying conditions that might affect ligament or muscle structures. However, we believe that our MMBS model captures the physical condition of the examined subject well enough, as was made sure by rigorous model validation and diligent result analysis in cooperation with orthopaedic surgeons. The deployed methods are state-of-the-art and widely used in in silico models [29,33,53,75,80]. As explained above, our findings are based on one patient with one implant design and cannot be applied to all types of knee implants commercially available. Anatomical variability and further implant designs should be analyzed in future studies to support the reported findings. In future studies, different surgical alignment techniques (mechanical vs. kinematic), implant designs, different activities, and more subjects should be considered [37,38].

Orthopedic surgeons attempt to optimize the dynamic behavior of the knee joint by a correct intraoperative alignment of the patellar component. Interindividual morphological joint variation [23–26] is known to influence knee joint dynamics [27] and functionality of TKRs [26,28]. Furthermore, the difference in patellar component designs and the size of the patellar component are factors that influence the PF dynamics [28,37,38]. Regarding patellar resurfacing, the all polyethylene dome-shaped patellar component is currently the most common used design [8,28]. Therefore, our study covers the clinically most relevant TKR type. Our data indicate that the thickness of the analyzed dome-shaped patellar component affects PF contact forces, patellar tilt, and patellar rotation which has also been reported by other research groups [6,8,43]. In this regard, it was found that a higher retropatellar loading due to an increase in the thickness of the patellar component might contribute to anterior knee pain after TKR [6,44]. These higher forces can be explained by the increasing strain of the PF ligaments and can have a similar effect to overstuffing the PF joint which leads to anterior knee pain in some cases [8]. Although reduced patellar component thickness was beneficial due to a decreased PF contact force, it is important to avoid an excessively thin patellar component due to the risk of patellar maltracking and decreased PF ligament forces [46]. Our findings showed that, if the implant thickness increased, the lateral patellar tilt also increased which is in line with previous studies [21,42]. The influence of the thickness was not as high as reported by Bengs et al. [35]. Hsu et al. [44] reported that the PF contact force of the thicker patella was 174% higher compared to the thinner patella; in our study it was 176%. Some biomechanical studies [21,42,44,45] reported that overstuffing the PF joint induced higher lateral tilt of the patella which was also observed in our results. Additionally, a higher thickness might lead to increased polyethylene wear and aseptic loosening of the patellar component [35,42,46,47].

It has been reported that mediolateral positioning of the patellar component affects the PF dynamics [6,41]. We demonstrated the same in our study, i.e., a medialized patellar component reduced the resulting PF contact force during knee flexion in agreement with clinical observations [6,32,36,52,86]. Furthermore, medialization led to a more lateral path and tilt of the patella which is in line with previous findings [36,41,48]. A medialized patellar button decreases the Q-angle [8,41,86] and consequently lateralizes the patella bone, thereby leading to a higher lateral tilt [86]. Our results also indicate that the PF contact force decreased above 60◦ flexion due to medialization of the patella which might decrease postoperative pain [32,36,48]. In clinical studies, the incidence of lateral release was less for the medially placed patellar component [8,20,48,52]. However, we also showed that medialization caused a higher tendency for lateral tilting. The reported changes in shift, tilt, and PF forces are in good agreement with previous studies [20,48,52,86]. Concerning the patellar tilt, our data are comparable to a clinical study investigating the effects of medialization [50]. The reduction in the PF contact force due to medialization is consistent with a decrease in the Q-angle due to lateral shift of the patella [8,48].

To date, it is not clear whether the superior–inferior position of the patellar component influences knee dynamics after total knee arthroplasty [28,33,49]. We showed that the superior–inferior position of the patellar component affected the PF dynamics which is in line with previous studies [28,32,33]. In another study, Fitzpatrick et al. [28] identified the superior–inferior positioning as the most sensitive parameter using a dynamic finite element model supporting our findings. In this regard, our results, however, favor an inferior placement of the patellar component due to a decreased PF contact force. Nakamura et al. [33] recommended a superior placement of the patellar component, whereas other studies recommended an inferior placement [5,32]. As it concerns the PF contact forces, our results are contradictory to findings presented by Nakamura et al. [33] but consistent with other studies [5,30,32]. One explanation for the divergent data could be differences in the considered load cases and numerical simulation models used. The contradicting consequences of inferior position of the patellar component [33] occur only at high flexion angles (around 130◦ flexion) which were not reached in our analyzed load case. Furthermore, a higher number of PF ligaments were considered in our study. In general, inferior position of the patellar button might be favorable for daily living activities such as squatting but less desirable in patients who demand to reach higher flexion angles [88]. Based on our findings, the physiological reconstruction of the patellar position is of high importance. The most common conventional surgical techniques for patellar resection, i.e., freehand with a saw, using a saw guide, and a reamer reveal a high variance regarding the accuracy of the patellar cut. Although computer-navigation is used in TKR to improve the accuracy of the bony cut and component placement, currently it is focused on the positioning of the tibial and femoral component. However, studies with navigated patellar resurfacing systems showed an equal or higher accuracy for the patellar cut and the reconstructing of the physiological patellar position in comparison to conventional techniques with good short- and mid-term outcomes [12,13]. Our results emphasize the importance to focus on precise placement of the patellar component.

Another advantage of our study is the use of a detailed MMBS model incorporating relevant muscles of the lower extremity as active force elements compared to previous studies that assumed the muscles to apply only constant forces [15,16,30,51,85], considered only few muscles [28,33], or analyzed passive load cases without active muscle forces [34]. Our study showed the potential of MMBS in very comprehensively investigating the effects of surgical parameters on knee joint dynamics. The presented non-invasive method to simultaneously predict muscle, ligament, and knee joint forces can be used to improve the preclinical testing of TKRs as described by Affatato et al. [47] on the example of knee wear simulators. Computational data can provide additional insight into the influence of patellar component malpositioning on PF dynamics during active knee motions that commonly occur during daily living. Most orthopedic surgeons perform patellar bone preparation and component positioning without a navigation system [12,13]; however, small changes in the positioning leads to crucial changes in PF dynamics as demonstrated in this study. Furthermore, we showed that translational positioning of the patellar component is a relevant factor supporting the need for a navigation-based surgical procedure [12,13]. We pointed out that the patellar component is basically robust to rotational malalignment due to a consistent contact region between the articulating surfaces, which is in agreement with the findings of Fitzpatrick et al. [28]. Therefore, intraoperative placement of patellar components should first focus on translational position rather than rotational orientation.

#### **5. Conclusions**

In conclusion, the presented effects of patellar component design and positioning on PF kinematics and kinetics are in good agreement with previous experimental and computational studies. Aiming at an optimal intraoperative patellar component alignment, the most important parameters are component thickness, mediolateral and superior–inferior positioning. In both manual and navigation-based surgical techniques, the patellar thickness, patellar tracking, patellar position relative to the joint line, the orientation of the trochlear, and positioning of the anterior femoral component should be considered in order to prevent PF complications. Our findings will support orthopedic surgeons in intraoperative patellar component positioning from a biomechanical perspective. Regarding the different positioning

of the patellar component, close attention should be paid to translational positioning as this might result in poor patient outcome.

**Author Contributions:** Conceptualization: M.K., T.M.G., R.B. and A.G.; methodology: M.K., M.D., S.K., C.S., R.B. and A.G.; software: M.K. and A.G.; validation: M.K. and A.G.; data curation: M.K., R.B. and A.G.; formal analysis: M.K. and A.G.; investigation: M.K., M.D., S.K., C.S., T.M.G., R.B. and A.G.; resources: R.B.; writing—original draft preparation: M.K.; writing—review and editing: M.K., M.D., S.K., C.S., T.M.G., R.B. and A.G.; visualization: M.K. and A.G.; supervision: M.K., R.B. and A.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** Three of the authors (S.K., C.S. and T.M.G.) are employees of Aesculap AG Tuttlingen, a manufacturer of orthopedic implants. The role of the three authors (S.K., C.S. and T.M.G.) had no effect on the collection and processing of the data. There are no personal circumstances or interest that influenced the representation or interpretation of reported results. The other authors (M.K., M.D., R.B. and A.G.) declare no conflict of interest.

#### **Appendix A**

#### *Mechanical Material Properties of the Ligaments*

**Table A1.** Mechanical properties (stiffness and reference strain) of the implemented ligaments were adapted from the literature [60,61,74,80]. The tibiofemoral ligament structures included: two bundles posterior cruciate ligament (PCL), two bundles lateral collateral ligament (LCL), three bundle medial collateral ligament (MCL), two bundles oblique popliteal ligament (OPL), one bundle arcuate popliteal ligament (APL), two bundles posterior capsule (pCAP), three bundles lateral patellofemoral ligament (LPFL), three bundles medial patellofemoral ligament (MPFL), and the patellar ligament (PL). Reference strains are the respective ligament strain values for the reference position upright standing. Stiffness is expressed in Newton per unit strain.


#### **Appendix B**

Investigated joint quantities during active two-leg squat motion for the validation of the musculoskeletal multibody simulation model (Figure A1). The tibiofemoral contact forces (Figure A1A) agreed with the measured contact forces in patients with telemetric TKR [84]. Likewise, the quadriceps force was in good agreement with previously published studies [82,83] (Figure A1B). Patellofemoral

contact forces (Figure A1C) calculated at 30◦, 60◦, and 90◦ knee flexion were similar to findings by Wallace et al. [81] and Innocenti et al. [82].

**Figure A1.** Validation of the musculoskeletal multibody simulation (MMBS) model. The MMBS model has been validated in terms of the tibiofemoral contact force (**A**), quadriceps force (**B**) and the patellofemoral contact force (**C**) during a dynamic squat motion. The tibiofemoral contact force is compared against the in vivo measurements of instrumented total knee replacements for three subjects (K1L, K2L, K3R, and K5R) [84]. The quadriceps force is compared against two simulation studies [82,83]. The resultant patellofemoral contact force is an important validation parameter for patellofemoral joint dynamics: it is compared against studies described in [81,82].

#### **Appendix C**

#### *Equations Used for Statistical Evaluation*

The computational formulas for mean absolute deviation (*MAD*)

$$MAD = \frac{\sum\_{i=1}^{n} |m\_i - c\_i|}{n},\tag{A1}$$

root mean square error (*RMSE*),

$$RMSE = \sqrt{\frac{\sum\_{i=1}^{n} \left(m\_i - c\_i\right)^2}{n}},\tag{A2}$$

Pearson correlation coefficient (*r*2)

$$\sigma^2 = \frac{n\sum m\_i c\_i - \sum m\_i \sum c\_i}{\sqrt{\left[n\sum m\_i^2 - \left(\sum m\_i\right)^2\right]} \sqrt{\left[n\sum c\_i^2 - \left(\sum c\_i\right)^2\right]}} \tag{A3}$$

and coefficient of determination (*R*2)

$$R^2 = \left(r^2\right)^2,\tag{A4}$$

presented here, where *mi* and *ci* represent the measured and calculated values at each time; step i, n is the total number of steps.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Corrosion Behavior of Surface-Treated Metallic Implant Materials**

**Therese Bormann 1,\*, Phuong Thao Mai 2, Jens Gibmeier 2, Robert Sonntag <sup>1</sup> , Ulrike Müller <sup>1</sup> and J. Philippe Kretzer <sup>1</sup>**


Received: 28 February 2020; Accepted: 23 April 2020; Published: 25 April 2020

**Abstract:** Corrosion of taper connections in total hip arthroplasty remains of concern, as particles and ions generated by corrosive processes can cause clinical problems such as periprosthetic osteolysis or adverse reaction to metallic debris. Mechanical surface treatments that introduce compressive residual stresses (RSs) in metallic materials can lead to a better performance in terms of fretting and fatigue and may lower the susceptibility to corrosion. The study investigates the impact of mechanical surface treatments on the corrosion behavior of metallic biomaterials. Compressive RSs were introduced by deep rolling and microblasting in Ti6Al4V and CoCrMo samples. Polished samples served as reference. Corrosion behavior was characterized by repeated anodic polarization. Residual stresses of up to about −900 MPa were introduced by deep rolling with a reach in depth of approximately 500 μm. Microblasting led to compressive RSs up to approximately −800 and −600 MPa for Ti6Al4V and CoCrMo, respectively, in the immediate vicinity of the surface. For Ti6Al4V, microblasting resulted in decreased corrosion resistance with lower breakdown potentials and/or increased passive current densities in comparison to the polished and deep-rolled samples. The corrosion behavior of CoCrMo on the other hand was not affected by the mechanical surface treatments.

**Keywords:** implant; biomaterial; corrosion; residual stress; total hip replacement; taper connection; anodic polarization; surface treatment

#### **1. Introduction**

Corrosion of metallic parts of artificial joints remains of concern as ions and particles that are generated by the corrosive processes can lead to clinical problems [1,2]. Alloys based on titanium and cobalt-chromium or stainless steels are usually applied as implant materials. The alloys themselves are highly corrosion resistant as they are passive metals (i.e., a protective oxide layer insulates the metal from the environment preventing corrosion). However, if the passive layer is continuously damaged, corrosion can—and does—occur.

In endoprosthetic implants, modularity enables intraoperative flexibility, which is especially beneficial in case of revision operations. In total hip arthroplasty (THA), the most prominent modular example is the taper connection between hip stem and femoral head. The material combination Ti6Al4V–CoCrMo is commonly used in these taper connections, because hip stems are often manufactured from Ti6Al4V due to its relatively low elastic modulus and the good osseointegrative properties, while metallic femoral heads are usually made of CoCrMo because of its high abrasion resistance. Corrosive processes at this interface are often referred to as *mechanically assisted crevice corrosion*

(MACC) [3]. This means that micromotion caused by cyclic implant loading leads to passive layer damage, which, in turn, promotes fretting wear as well as galvanic and crevice corrosion. In extreme cases, almost the complete trunnion of the taper connection was found to be worn off (trunnionosis) [4–6]. However, even in less dramatic cases, a considerable amount of metal loss can be generated at the head–neck interface, which can cause clinical problems, such as periprosthetic osteolysis and adverse reaction to metallic debris [7]. Furthermore, elevated blood metal ion levels can be provoked by corroding taper connections [7]. Using ceramic heads instead of heads from CoCrMo is a possibility to reduce corrosion at this particular interface. However, also for ceramic heads on metallic bone stems, taper corrosion has been reported, albeit to a lower extent [8]. Independent of head material, state-of-the-art to prevent taper corrosion is a proper head impaction during surgery. It has been shown that an impact power of at least 4 kN should be applied to ensure good interlocking of head and stem and reduce micromotion as far as possible [9–12].

Another approach in increasing fatigue and corrosion resistance is to alter the surface properties of the metals. The induction of compressive residual stresses in the surface and subsurface region reduces the susceptibility to crack initiation, for example by nanocrystal formation and strain hardening, which, in turn, enhances the resistance towards fatigue and fretting fatigue [13–15]. For Ti6Al4V, a superior fatigue and fretting fatigue performance after shot peening processes has been shown, for example, by Sonntag et al., Altenberger et al. and Liu et al. [14,16,17]. Aside from the mechanical behavior, corrosion characteristics of metals are influenced by residual stresses. It is well known that tensile RSs increase the susceptibility of stress corrosion cracking. Compressive RSs, on the other hand, can stop crack propagation and decrease the susceptibility of pitting corrosion [18–20]. The beneficial effects of compressive RSs on the corrosion resistance were shown for example for AA2024-T3—an aluminum alloy with poor corrosion properties—and a magnesium alloy for biomedical applications [19,20]. Lee at al. investigated the effect of shot peening on Ti6Al4V, which is a well-known technique to induce compressive RSs, and found increased high-cycle fatigue resistance even under seawater environment [21]. Both by shot peening and deep rolling, compressive RSs can be induced for depth ranges of some 100 μm depending on the process parameters used. However, techniques like deep rolling or laser shock peening have likewise been applied to induce compressive RSs to depths of up to a millimeter and even more depending on the intensity of the process [16].

This study aims at investigating the corrosion properties of Ti6Al4V and CoCrMo after inducing compressive RSs by the techniques of deep rolling and microblasting. The corrosion behavior of the mechanically surface-treated materials is investigated by means of anodic polarization. The two investigated alloys are commonly used as pairing in head–neck taper connections in total hip arthroplasty. As corrosion at this interface is complex, for separation of the individual influencing factors in this first approach, only the material–electrolyte interaction was investigated. We are aiming to assess the potential of mechanical surface treatments with regard to improved corrosion resistance of modular taper junctions in THA.

#### **2. Materials and Methods**

Surface treatments and electrochemical corrosion experiments were performed on discs made of Ti6Al4V ELI (Ti) as defined in DIN ISO 5832-3 and the low carbon CoCrMo-alloy (CoCr) defined in DIN ISO 5832-12, respectively. Ti6Al4V was received as rod material with a diameter of 15 mm. To provide a defined microstructure and to eliminate all residual stresses, the rod material was heat treated at 900 ◦C for 10 min under vacuum followed by oil quenching. Finally, the material was annealed at 500 ◦C for 4 h followed by a slow furnace cooling. CoCrMo was received as rod material with a diameter of 28 mm. This material was used in the as-received state. For surface treatments and electrochemical tests, discs with a diameter of 15 mm (Ti) and 28 mm (CoCr) and a thickness of 3 mm were manufactured.

#### *2.1. Surface Treatments*

The corrosion behavior of the two materials was investigated for the three different surface conditions of (i) metallographical polishing (PO), (ii) deep rolling (DR) and (iii) microblasting (MB).

A metallographic preparation (i.e., grinding, fine grinding and polishing by diamond and/or oxide suspension) removes the residuals from previous processing. Therefore, the metallographically-prepared samples served as reference. The mechanical surface treatments of deep rolling and microblasting on the other hand induced significant compressive residual stresses. Microblasting is a shot peening process with grit sizes in the micrometer range and a commonly used technique for surface treatments of metallic implant materials [22,23].

All samples were at first on one face ground and fine ground using SiC-paper from a grit of P320 to finally P2500. For the polished samples, the final surface treatment comprised diamond polishing applying diamond suspension down to a grain size of 3 μm.

Deep rolling and microblasting were carried out according to the process parameters listed in Table 1.


**Table 1.** Processing parameters applied for deep rolling and microblasting.

For the deep rolling, the rolling tool is directed using a meandering pattern, as schematically shown in Figure 1. Here, the compressive residual stresses at the material's surface are generated through continuous plastic deformation induced through the rolling contact (Hertzian pressure) between the tool and the workpiece material. The overlapping of adjacent rolling tracks induces characteristic non-axisymmetric residual stress distributions in lateral direction, i.e., maximum compressive residual stresses occur in feed direction, while in rolling direction, typically, much lower compressive residual stresses develop.

Microblasting is carried out by moving the blast nozzle also following a meandering pattern. Each blasting grain induces a residual imprint due to plastic deformation of the interacting material. Due to the high coverage the resulting lateral compressive residual stress distribution in the near-surface region is generally independent of direction (i.e., an axisymmetric residual stress state is generated).

The results of the surface treatments were assessed by X-ray residual stress analysis.

Scanning electron microscopy (LEO EVO 50, Zeiss, Oberkochen, Germany) in secondary electron contrast mode was applied in order to visualize the sample surfaces after the final surface treatment.

**Figure 1.** Scheme of the meandering pattern used for deep rolling and microblasting with indication of the feed and the rolling direction.

#### *2.2. X-ray Residual Stress Analysis*

Residual stresses in the two principal directions were determined using a custom-made 3-axis X-ray diffractometer in ψ-configuration according to the well-known sin2ψ-technique [24]. X-ray diffraction analysis was performed with Ni-filtered Cu Kα radiation for the {213}-lattice planes in the 2Θ-range <sup>136</sup>◦ <sup>≤</sup> <sup>2</sup><sup>Θ</sup> <sup>≤</sup> 147.4◦. Fifteen sample tilts were considered in the range <sup>−</sup>60◦ <sup>≤</sup> <sup>ψ</sup> <sup>≤</sup> <sup>60</sup>◦. As primary aperture, a ø 1 mm pin hole collimator was applied. On the secondary side, a 4 mm symmetrizing slit was used in front of the scintillation counter. Data post-processing was carried out using a Pearson VII fit after background subtraction. The diffraction elastic constants (DEC) E{213} = 113 GPa and ν{213} = 0.32 for Ti6Al4V and E{220} = 227 GPa and ν{220} = 0.3 for the CoCr samples were applied. Depth distributions of residual stresses were determined by means of electrochemical sublayer removal and reapplication of the sin2ψ-measurement at the newly created surface.

#### *2.3. Electrochemical Testing*

Electrochemical tests were accomplished using a potentiostat (Wenking MLab 200, Bank Elektronik – Intelligent Controls, Pohlheim, Germany) in combination with an electrochemical cell. The standard three-electrode system was applied for electrochemical testing with the Ti or CoCr discs serving as working electrode. Discs were covered by a silicone sealing mask which exposed the metal to the electrolyte through an orifice of 11.3 mm diameter, resulting in a working electrode area of 1 cm2. The KCl saturated Ag/AgCl reference electrode (Sensortechnik Meinsberg, Waldheim, Germany) was placed in a Luggin capillary. The distance between capillary tip and working electrode was set to 1 mm. A platinized titanium bolt of 6 mm diameter served as counter electrode. Anodic polarization curves were recorded between −800 and 2000 mV for Ti and between −800 and 1000 mV for CoCr samples. Anodic polarization was repeated four times for each sample. The scan rate was set to 1 mV/s as proposed in the ASTM standard F2129-15 [25]. Open-circuit potential (OCP) measurements were conducted before, in between and after the polarization scans for one hour in each case. Dulbecco's phosphate buffered saline (PBS) (Biochrom, Berlin, Germany) with a pH of 7.4 was used as electrolyte for electrochemical tests of the three generated surface conditions for Ti and CoCr samples. In addition, a second electrolyte containing 10 g/l FeCl3 in Ringer's solution (FeCl3) (B. Braun AG, Melsungen, Germany) was used for electrochemical tests of polished, deep-rolled and microblasted Ti discs. The FeCl3-containing electrolyte has a low pH of 1.8 and a higher concentration of Cl<sup>−</sup> ions and was used as corrosion accelerating environment [26,27]. Electrochemical testing was done at 37 ◦C and by using aerated electrolytes. Before electrochemical testing, all samples were ultrasonically cleaned in ethanol for 10 min. The current density of the passive plateau (Ip) was determined from the anodic polarization curves as qualitative measurement of material degradation processes. The breakdown potential (Eb), which characterizes the onset of pitting corrosion, was determined from the intersection between passive current density and a linear fit of the polarization curve in the transpassive regime (i.e., the part of the curve leaving the passive plateau abruptly towards higher current density values). All curves were median filtered prior to the determination of Ip and Eb.

#### *2.4. Surface Roughness Measurements*

Surface roughness measurements were carried out after electrochemical testing using a perthometer (M2, Mahr, Göttingen, Germany). For each sample, four measurements were done in the area exposed to the electrolyte as well as in sample regions that were not exposed to the electrolyte during electrochemical testing but were covered by a silicone sealing sheet. The direction of the surface roughness profiles was changed about 90◦ for each measurement. The first roughness profile was placed arbitrarily on the sample surface. We consider the unexposed surface region to reflect the initial surface roughness of the samples before electrochemical testing. The length of the measuring line was 1.75 mm.

All results are given as mean ± standard deviation from three samples (n = 3).

#### **3. Results**

#### *3.1. Mechanical Surface Treatment*

The scanning electron microscope (SEM) images presented in Figure 2 emphasize the different characteristics of the mechanical surface treatments. Compared to the polishing and deep rolling process, microblasting results in a more rugged surface due to the stochastic impact of individual grits.

**Figure 2.** Scanning electron microscope images of the polished, deep-rolled and microblasted Ti6Al4V and CoCrMo (CoCr) samples.

The residual stress depth distributions for the Ti6Al4V samples are presented in Figure 3 together with the depth distributions of the mean integral widths of the recorded diffraction lines. In comparative studies, using the same measuring parameters, the change in the integral widths values corresponds to the degree of cold working (i.e., higher values indicate a larger degree of cold working). As expected, the polished samples can be treated as stress-free. The integral widths of the diffraction lines also indicate that almost no plastic deformation was induced by the polishing process, since the values are similar to the ones of the bulk material, which is unaffected by the surface treatment. Deep-rolled and microblasted samples show maximum RSs of about −800 MPa. While the microblasted sample shows its maximum directly at the surface, deep rolling led to maximum compressive RSs at a depth of about 200 μm. It should be noted that the compressive RSs resulting from deep rolling are directional

(i.e., the RSs are significantly higher in the feed direction compared to the rolling direction (Figure 1)). Compressive RSs due to the deep rolling process were introduced up to a depth of approximately 600 μm, whereas for the microblasted sample the range of influence extents into approximately 50 μm depth. Regarding the mean integral width, for the microblasted sample the value dropped from 3.7◦ to 1.8◦ within the first 50 μm below the surface, indicating a rather steep cold working gradient. For the deep-rolled sample, the value directly at the surface referred to about 2.4◦ and decreased continuously with increasing distance to the surface.

**Figure 3.** Residual stress depth distributions (**a**) and depth distributions of the average integral widths of the diffraction lines (**b**) for the three differently mechanically surface-treated Ti6Al4V (Ti) samples.

Figure 4 shows the depth distributions of RSs and mean integral widths for the CoCrMo samples. The courses show similarities to the ones of the Ti6Al4V samples. The polished sample is almost unaffected by the surface treatment, it shows only a small amount of tensile residual stresses close to the surface. Deep rolling and microblasting on the other hand induced characteristic compressive residual stress distributions. Microblasting resulted in compressive RSs of approximately −600 MPa directly at the surface. In addition, a steep gradient with a zero crossing and a change of sign towards tensile RSs in a depth of about 30 μm was found. As for the Ti6Al4V samples, the RSs are higher in the feed direction than in the rolling direction. Directly at the surface, compressive RSs in the feed direction of about −470 MPa and in the rolling direction of about −120 MPa were determined. At a depth of approximately 100 μm, the maximum of the compressive RSs reached about −900 MPa in the feed direction and about –580 MPa in the rolling direction. Zero crossing is observed at a depth of about 500 μm. At depths of more than 500 μm, tensile RSs were determined.

Regarding mean integral widths, the polished sample exhibited values in the range of the unaffected bulk material (see Figure 4). The microblasted sample showed significantly higher values of about 1.9◦ mean integral width directly at the surface. The value dropped within the first 30 μm below the surface to the values of the unaffected bulk material. The deep-rolled CoCrMo samples exhibited mean integral widths of about 1.2◦ at the very surface, which is slightly higher than for the unaffected bulk material. The depth profile shows a maximum of about 1.5◦ in approximately 100 μm depth, which is related to the location of maximal Hertzian stresses. With increasing depth, the integral widths continuously decreased to the value of the unaffected bulk material (1.05◦).

**Figure 4.** Residual stress depth distributions (**a**) and depths distributions of the average integral widths of the diffraction lines (**b**) for the three differently surface-treated CoCrMo (CoCr) samples.

#### *3.2. Electrochemical Tests*

Figure 5 represents the breakdown potentials (Eb) obtained from anodic polarization curves of the two materials in different surface conditions. For the polished and deep-rolled Ti discs, the breakdown potential could not be reached within the measuring range of up to 2000 mV if the used electrolyte was PBS. The microblasted samples; however, showed a distinct passive layer breakdown potential of 1536 ± 105 mV. If tested in the FeCl3-containing electrolyte, the breakdown potential of all three investigated surface conditions was below 2000 mV. Polished samples had the lowest breakdown potential of 1504 ± 17 mV, while Eb of deep-rolled and microblasted samples referred to 1688 ± 115 and 1704 ± 100 mV, respectively. The surface treatments did not influence the breakdown potentials of CoCr discs, Eb accounted to 589 ± 10, 591 ± 10 and 593 ± 9 mV for polished, deep-rolled and microblasted discs, respectively. The breakdown potentials were hardly influenced by the scan number of the cyclic repeated polarization.

**Figure 5.** Breakdown potential (Eb) of polished (PO), deep-rolled (DR) and microblasted (MB) Ti6Al4V samples tested in PBS and in FeCl3-containing electrolyte (FeCl3) and of polished, deep-rolled and microblasted CoCrMo samples tested in PBS. Values comprise all four anodic polarization scans of three tested samples. Values marked with \* refer to breakdown potentials above 2000 mV.

Figure 6 presents the passive current densities from the anodic polarization curves. Passive current densities of the first and the following scans, respectively, are displayed separately because Ti shows clearly increased passive current densities during the first anodic polarization cycle compared to all following scans (cp. Table 2). The difference amounts to about one order of magnitude. Passive current densities of the microblasted Ti samples are about two times higher than those of polished and deep-rolled samples in both tested electrolytes. In PBS, polished and deep-rolled surfaces exhibited comparable passive current densities. If tested in FeCl3-containing electrolyte, the polished and deep-rolled samples showed comparable passive current densities in the first polarization cycle, while in the following cycles the deep-rolled surface show a tendency towards lower Ip. Passive current density of MB Ti samples was raised during the first anodization cycle in the FeCl3-containing electrolyte in comparison to PBS. This tendency diminished during repeated polarization.

Passive current densities of CoCr are less affected by cycle number and surface condition, respectively (cp. Table 2). However, there are slight tendencies towards higher passive current densities during the first anodization and for the microblasted surface, respectively.

**Figure 6.** Passive current densities of polished, deep-rolled and microblasted Ti6Al4V samples tested in PBS and in FeCl3 and of polished, deep-rolled and microblasted CoCrMo samples tested in PBS. (**a**) Passive currents from the first anodic polarization cycle. (**b**) Passive currents from polarization cycles no. 2–4.


**Table 2.** Passive current densities (mean (SD)) during first anodic polarization cycles and anodic polarization cycles no. 2–4 of polished, deep-rolled and microblasted Ti6Al4V and CoCrMo samples.

The anodic current density–potential plots shown in Figure 7 illustrate the distinct first polarization cycle of Ti samples. The passive region subdivides in two parts of a less distinct plateau at passive current densities in the range of several hundred nA, and a second, more pronounced plateau in the μA range. In scans no. 2–4, a stable plateau region at the lower passive current densities developed. The FeCl3-containing electrolyte resulted further in a displacement of the corrosion potential (i.e., the potential at zero crossing of the current density, which refers to the minimal value in logarithmic representation), from about −600 to about 500 mV (cp. Table 3), which means that the cathodic reaction is enhanced if FeCl3-containing electrolyte is used.

**Figure 7.** Anodic polarization curves of Ti6Al4V samples. (**a**) Deep-rolled sample tested in PBS, (**b**) deep-rolled sample tested in FeCl3, (**c**) microblasted sample tested in PBS and (**d**) polished sample tested in FeCl3.

**Table 3.** Corrosion potential (mean (SD)) of polished, deep-rolled and microblasted Ti6Al4V specimens tested in PBS and FeCl3-containing electrolyte, respectively.


For CoCr discs, again the curve progression in the first polarization cycle differs from the following cycles, as illustrated for a polished and a microblasted CoCr sample, respectively, in Figure 8. For polished and deep-rolled samples, during the first cycle, a passivation peak at approximately −700 mV can be noticed (cp. Figure 8a). After reaching the local maximum, the current density lowers to the passive current densities displayed in Table 2. With further increasing potential starting at approximately 0 mV, the passive current rised again slowly to a shoulder before the breakdown potential—characterized by the abruptly rising current density—is reached. This progression was observed for all mechanical surface treatments.

**Figure 8.** Anodic polarization curves of (**a**) a polished and (**b**) a microblasted CoCrMo sample tested in PBS.

#### *3.3. Surface Roughness*

Figure 9 presents the surface roughness value Ra of all tested samples in the as-mechanicallytreated condition as well as after electrochemical testing. In Table 4, the Ra and Rz values are displayed. Generally, Ti samples comprise a higher surface roughness than CoCr samples. The roughness values of polished and deep-rolled samples are comparable, as shown in Table 4. For Ti samples only, deep rolling leads to slightly increased Rz values in comparison to polished samples. Microblasting results in a surfaces roughness of about one order of magnitude higher than in case of polishing or deep rolling. These findings are reflected by the SEM images in Figure 2.

During electrochemical testing, the samples were exposed to the electrolyte in combination with cyclic anodic polarization. Electrochemical testing of Ti samples in PBS did not result in significant changes of the surface roughness of the exposed areas. Polished Ti samples showed; however, a tendency towards decreasing surface roughness values in the regions that have been exposed to the FeCl3-containing electrolyte.

CoCr samples were solely tested in PBS. After electrochemical testing, a slightly increased surface roughness of the polished and deep-rolled surfaces was detected. The microblasted surfaces; however, were not altered by the cyclic anodic polarization, regardless of the sample material or electrolyte.

**Figure 9.** Ra values (nm) for Ti6Al4V and CoCrMo in polished, deep-rolled and microblasted surface condition. "Original surface" (OS) refers to the region which had not been in contact with the electrolyte during electrochemical testing; PBS refers to the surface after being electrochemically tested in PBS; FeCl3 refers to the surface after being electrochemically tested in FeCl3-containing electrolyte.


**Table 4.** Ra and Rz values in nm for all tested materials, surface conditions and electrolytes.

#### **4. Discussion**

Generally, Ti6Al4V showed higher breakdown potentials than CoCrMo, meaning that Ti has a higher resistance to pitting corrosion. At polished and deep-rolled samples, breakdown of the passive layer could not be accomplished in PBS up to a potential of 2000 mV. This is consistent with the literature reporting breakdown potentials between 2500 and 3500 mV for Ti6Al4V in chloride-containing electrolytes [28,29]. The microblasted sample, on the other hand, exhibited a clearly lower breakdown potential of 1536 mV in PBS, which indicates a greater susceptibility to pitting corrosion of microblasted surfaces in comparison to polished or deep-rolled surfaces. Polarization of Ti samples in the FeCl3-containing electrolyte led to breakdown of the passive layer for all tested surface conditions well below 2000 mV, while in this corrosion-accelerating electrolyte, the polished samples exhibited passive layer breakdown at lower potential than microblasted and deep-rolled samples. Interestingly, the addition of FeCl3 only resulted in decreased breakdown potentials for PO and DR surfaces, the breakdown potential of the microblasted surface was not further lowered by the FeCl3-containing environment.

We consider passive current densities as qualitative measurement for material degradation processes, as these involve exchange and transport of ions and charged molecules. Higher passive current densities would therefore represent increased material degradation rates, while lower passive current densities would refer to smaller corrosion rates. Polished and deep-rolled Ti6Al4V samples showed substantially lower passive current densities than samples with microblasted surfaces, which confirms that the corrosion resistance of Ti6Al4V is higher after metallographical polishing and deep rolling than after microblasting of the surface.

However, during the first anodic polarization cycle, passive current densities of Ti are substantially higher than passive current densities of CoCr. In addition, we observed two distinct regions in the passive potential range. It is well known that Ti6Al4V spontaneously forms a stable oxide layer when exposed to air. The oxide layer has a thickness of approximately 5 nm and consists mainly of TiO2 and a small quantity of TiO and Ti2O3 at the metal–oxide interface [28]. Milosev et al. showed that upon polarization of Ti6Al4V, the thickness of the TiO2 layer increases and Al2O3 is introduced in the passive layer already at low potentials [28]. We; therefore, attribute the distinctive passive region during the first polarization to growth and chemical transformation of the oxide layer. In all following polarization cycles, the oxide layer stabilized and does not undergo further changes, as can be seen from the constant current density within the passive region and the almost identical current density–potential plots.

As illustrated in Figure 8, the first polarization cycle of CoCrMo samples differed from all following cycles. For polished and deep-rolled samples, a passivation peak (i.e., an active–passive transition) was observed during the first polarization scan. Hence, we suppose a thickening or healing of the existing oxide layer at low potentials. All surface treatments exhibited increased passive current densities over a potential range of approximately 300 mV just before passive–transpassive transition. This might be attributed to chemical changes, as the composition of the oxide layer is highly dependent on the applied potential [30,31]. The spontaneously formed oxide layer mostly consists of Cr2O3 and small quantities of Co- and Mo-oxides and—according to Milosev et al.—has a thickness of about 1.8 nm [31]. Chemical modifications caused by polarization in intermediate potential ranges include the incorporation of Co- and Mo-oxides in the passive layer, which is accompanied by an increase in thickness [31]. Milosev et al. found the increase in thickness of the passive layer to account for about a factor 3, if CoCrMo is polarized up to 800 mV (against a saturated calomel electrode) in simulated physiological solution [31]. However, after the first polarization cycle, the passive layer of CoCrMo stabilized, so that, in all following cycles, neither passivation peaks nor increasing passive current densities were observed before passive–transpassive transition. Among the different surface treatments, the passive current densities of CoCrMo hardly varied. The changes within the oxide layer on CoCrMo caused by polarization can; therefore, be considered to be independent of the mechanical surface treatments.

The decreased corrosion behavior of microblasted Ti6Al4V samples can be attributed to the higher roughness of the surface. A higher corrosion resistance related to lower surface roughness has been reported for Ti6Al4V but also for other metals and alloys, such as cp-Ti and NiTi [29,32–34]. The higher passive current densities in microblasted Ti samples can be attributed to the true surface area, which increases with roughness. The irregular grooves that are formed upon microblasting might, in addition, lead to small, enclosed areas with reduced diffusion in the electrolyte. This could lead to a locally-restricted oxygen depletion in the electrolyte, which may result in localized corrosion similar to the process of crevice corrosion [29]. Additionally, the oxide layer of the rough surface may possess a higher degree of imperfections, which would make the surface more susceptible to pitting corrosion [29]. The lower breakdown potential of the microblasted Ti6Al4V samples tested in PBS can be attributed to these effects. In a more aggressive corrosive environment; however, these effects seem to play a minor role, as in the FeCl3-containing electrolyte all three Ti6Al4V surface modifications exhibited breakdown potentials in the same range.

For CoCrMo; however, neither influence of the near-surface RS state nor influence of roughness on corrosion behavior could be observed. The roughness values of microblasted CoCrMo are about two times lower than the roughness values of microblasted Ti. There might be a critical roughness, above which the corrosion behavior is affected. It is possible; however, that the influence of roughness on breakdown potential and passive current densities is alloy dependent. Evidence can be found in literature of metals and alloys that do not show a clear relation of corrosion resistance and surface roughness (e.g., β-Ti alloy and stainless steel) [33,35].

In addition to the analysis of the potentiodynamic polarization curves, the roughness of the materials before and after polarization might be an indicator for the extent of the corrosional attack during cyclic polarization. The deep-rolled and polished CoCrMo samples showed a tendency towards increased surface roughness after the four conducted polarization cycles. This indicates material loss due to pitting corrosion. Deep-rolled and microblasted Ti samples did not show altered roughness values after cyclic polarization in PBS and FeCl3-containing electrolyte. The polished samples polarized in FeCl3-containing electrolyte showed; however, a slight tendency towards decreasing surface roughness after the electrochemical tests. Roughness decrease due to electrochemical material degradation can be explained by the difference in Volta potential between valley and peaks, as electrons in peak areas are more likely to escape the material than electrons in valley areas [36]. This would mean that roughness progression due to galvanic corrosion would pass a minimum before a roughness increase due to pitting becomes evident. Furthermore, in the FeCl3-containing environment, polished Ti samples exhibited lower breakdown potential than deep-rolled Ti samples. This indicates that the corrosion resistance of deep-rolled Ti6Al4V might be superior to that of polished Ti6Al4V, which might be correlated to the high compressive RSs in the near-surface region. Thus, for the three investigated surface conditions of Ti6Al4V, deep rolling resulted in relatively higher corrosion resistance.

This work investigated the potential of surface treatments for enhancing the corrosion resistance of passive metals, which are applied in the field of surgical implants. It has been shown that the surface treatments did impact the corrosion behavior of the metals, even though the observed differences on the breakdown potentials and passive current densities between polished and deep-rolled samples were rather small. The corrosion of taper connections is; however, mainly initiated by occurring micromotion, which can damage the passive layer leading, in turn, to fretting and galvanic corrosion [3,37,38]. The mechanical resistance of the oxide layer is; therefore, crucial for the corrosion resistance of the applied alloys. Assessment of corrosion in this study was somehow simplified, as the mechanical component of taper corrosion was not taken into account. This was done in order to separate the effects of occurring micromotion from the effects caused by galvanic corrosion. In this setup, the corrosion behavior is; therefore, rather governed by sample roughness than by the induced compressive RSs, especially since the investigated alloys are passive metals with a naturally high corrosion resistance. However, apart from the slightly improved corrosion resistance for the deep-rolled surface of Ti6Al4V observed in this study, the induction of compressive RSs by means of deep rolling have moreover the potential of combining a mirror smooth sample surface with the beneficial effects in terms of fretting and fatigue resistance.

Further studies are needed that involve actually occurring forces and resulting micromotion in order to evaluate the impact of compressive RSs onto the corrosion resistance of modular taper connections.

#### **5. Conclusions**

Compressive RSs were introduced in Ti6Al4V and CoCrMo by means of deep rolling and microblasting, while metallographically-prepared samples served as reference. For CoCrMo, the surface treatments did not alter the corrosion behavior. For Ti6Al4V, microblasting led to a lowered breakdown potential and/or increased passive current densities, which indicates decreased corrosion resistance. Further studies comprising mechanical loads are needed in order to assess the impact of mechanical surface treatments with respect to corrosion of modular taper connections in total joint arthroplasty.

**Author Contributions:** Conceptualization, J.P.K., J.G. and R.S.; methodology, J.P.K., U.M. and J.G.; validation, all authors; formal analysis, T.B. and P.T.M.; investigation, T.B. and P.T.M.; resources, J.P.K. and J.G.; data curation, T.B. and P.T.M.; writing—original draft preparation, T.B. and P.T.M.; writing—review and editing, J.P.K., R.S., U.M. and J.G.; visualization, T.B. and P.T.M.; supervision, J.P.K. and J.G.; project administration, J.P.K and J.G.; funding acquisition, J.P.K., R.S. and J.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the *Deutsche Forschungsgemeinschaft* (DFG) within the framework of the project no. 382919963.

**Conflicts of Interest:** The authors declare no conflict of interest related to this study.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Review* **What the Surgeon Can Do to Reduce the Risk of Trunnionosis in Hip Arthroplasty: Recommendations from the Literature**

#### **Claude B. Rieker 1,\* and Peter Wahl <sup>2</sup>**


Received: 29 February 2020; Accepted: 17 April 2020; Published: 21 April 2020

**Abstract:** Trunnionosis, defined as wear and corrosion at the head–neck taper connection, is a cause of failure in hip arthroplasty. Trunnionosis is linked to a synergistic combination of factors related to the prosthesis, the patient, and the surgeon. This review presents analytical models that allow for the quantification of the impact of these factors, with the aim of providing practical recommendations to help surgeons minimize the occurrence of this failure mode. A tighter fit reduces micromotion and, consequently, fretting of the taper connection. The paramount parameters controlling the fixation force are the coefficient of friction and the impaction force. The influence of the head diameter, as well as of the diameter and angle of the taper, is comparatively small, but varus alignment of the taper and heads with longer necks are unfavourable under physiologic loads. The trunnion should be rinsed, cleaned, and dried carefully, while avoiding any contamination of the bore—the female counterpart within the head—prior to assembly. Biological debris, and even residual water, might critically reduce the fixation of the taper connection between the head and the neck. The impaction force applied to the components should correspond to at least two strong blows with a 500 g hammer, striking the head with an ad hoc impactor aligned with the axis of the taper. These strong blows should correspond to a minimum impaction force of 4000 N.

**Keywords:** hip arthroplasty; trunnionosis; trunnion failure; fretting corrosion; head–neck junction; mechanically assisted crevice corrosion

#### **1. Introduction**

Total hip arthroplasty (THA) is so successful in restoring mobility and relieving pain in patients with degenerated hip joints [1] that it has been nominated as the operation of the 20th century [2]. Failure, however, remains an issue, with between one-third and nearly half of THA procedures requiring postoperative revision within 30 years [3,4]. One of the possible causes of failure is trunnionosis [5]. Trunnionosis is defined as wear and corrosion at the head–neck taper connection [6]. Hence, it is associated with the modularity of the head–stem construct. Modularity gives the surgeon the flexibility to choose femoral heads of varying materials and diameters, with variable neck lengths, so that the joint replacement can be adjusted according to the patient's anatomy [7]. It can also reduce the inventory and consecutive storage costs [8,9]. In THA, the use of modular heads began in the early 1970s and has almost completely supplanted monobloc femoral components [6,10].

Epidemiological data on the incidence of clinically relevant trunnionosis are scarce. Up to 4.7% of revisions are reported to be attributable to taper corrosion [11–13]. Rates of up to 10.5% are even reported for certain subgroups [12]. While the latter number seems rather high, the occurrence of

trunnionosis is certainly underreported, as the taper is seldom analysed. In some revisions, only the femoral head is revised and the stem is retained, making a full analysis impossible. Sometimes, the stem may be revised without disconnecting the head. Additionally, it remains quite difficult to determine the clinical relevance of this issue when other reasons for revision are present concomitantly. While taper corrosion might be observed frequently on retrievals [14–16], the clinical relevance might remain difficult to determine [11], particularly because approximately one-third of the adverse local tissue reaction (ALTR) pseudotumours related to taper corrosion identified on magnetic resonance imaging are asymptomatic [17]. A determination of the levels of cobalt and chromium in the synovial fluid might help identify taper corrosion-related issues [18].

The aetiology of trunnionosis is believed to be a synergistic combination of factors related to the prosthesis, the patient, and the surgeon [19–25]. Trunnionosis involves both fretting corrosion as well as crevice corrosion [26,27]. This process is called mechanically assisted crevice corrosion (MACC) [19]. Taper corrosion can lead to elevated metal ion levels in the synovial fluid of the affected joint as well as in the serum, and may cause ALTR [11,13,18,28,29]. ALTR includes lymphocyte-dominated inflammatory reactions and macrophage infiltrates reacting to particulate corrosion products [11,30]. These can lead to synovitis, local osteolysis, the necrosis of periprosthetic tissues and, finally, component loosening [11,13,31]. Long-term MACC leads to material loss at the taper junction, which can, in rare cases, lead to the frank dissociation of the connection, as well as marked taper deformity [13,20,32–34].

This manuscript aims to quantify the impact of these factors according to published analytical models, as well as clinical and in vitro studies, and provide practical recommendations to help surgeons minimize the occurrence of trunnionosis. We conducted a thorough non-systematic review of the literature using two search engines (PubMed and Google Scholar, using the following keywords: trunnionosis, fretting corrosion, taper connection, taper corrosion, taper failure, MACC, modularity, assembly force, disassembly force, micromotion) and cross-referenced related studies to identify the relevant literature.

#### **2. Technical Aspects of Taper Connections in Hip Arthroplasty**

A taper connection is a means of reliably joining two mechanical components, by tightly fitting a cone into a negative cone-shaped counterpart [35]. The male component is referred to as the trunnion, while the female counterpart is a bore [24,35]. A taper is defined by three parameters: the largest diameter at its base, the smallest diameter at its opening or tip, and its angle [36].

MACC of a Morse taper connection is caused by fretting and crevice corrosion [26,27]. Fretting first disrupts the protective oxide layer on the surfaces of the taper and causes wear. Changes in local chemistry within crevices then lead to the complex interactions of crevice corrosion [26,27]. Although repassivation (i.e., reformation of the protective oxide layers) occurs naturally, fretting alters the repassivation of the exposed metals [24,25,37].

A recently published study indicates that trunnionosis is mainly determined by fretting corrosion, rather than by crevice corrosion [38]. Therefore, minimizing the micromotions at the head–neck taper interface would mitigate the starting conditions of trunnionosis. A strong press-fit fixation of the taper interface will logically lower these micromotions [21,39]. This underlines the importance of a stable fixation between the femoral head and the stem's trunnion [6,20,21]. Fretting corrosion at the taper interface is linked to micromotions of 5 μm to 12 μm [21,40]. Time in vivo (i.e., exposure to repeated loads) is also linked to the degree of corrosion of tapers [12,26]. Thus, a stable taper connection with lower micromotions under physiological loads will produce a reduced risk of trunnionosis. The force required to remove the head from the taper is a measure of taper stability, denoted as the fixation force or the pull-off force. Given the association between fixation force (i.e., the force necessary to dissociate the taper connection) and micromotion at the taper's interface [21,39], the fixation force is a surrogate parameter for the rest of this analysis. Fessler et al. [41,42] and MacLeod et al. [43] have both provided analytical models to estimate the fixation force between the neck and the head. These two analytic models are presented in Figure 1.

$$F\_{f\cos a\pi \text{(eq)}} = F\_{\cos a\pi \text{ etc}} \cdot \frac{\mu \cos(0.5\alpha) - \sin(0.5\alpha)}{\mu \cos(0.5\alpha) + \sin(0.5\alpha)}$$

$$F\_{\text{friction}} = F\_{\text{inject}} \cdot \frac{\mu \cdot \text{cod}(0.5a) - \text{sn }(0.5a)}{\mu \cdot \text{cod}(0.5a) + \text{sn }(0.5a)} \cdot \frac{E\_h \eta}{2\eta^2} \cdot \left(1 - \frac{\eta^2}{\eta^2}\right) \cdot \left[\frac{\eta}{E\_h}\left(\left\{\frac{\eta^2 + \eta^2}{\eta^2 - \eta^2}\right\} + \vartheta\_h\right) + \frac{\eta}{E\_t}(1 - \vartheta\_t)\right]$$

**Figure 1.** Fessler and MacLeod's formulas illustrated. MacLeod's model uses the same fundamental formula as Fessler's model, as illustrated in the figure, but adds three factors that consider the geometry and the material properties of the system. In both models, the fixation force correlates linearly with the force applied to impact the taper. The effects of the other parameters are illustrated in Figures 2–4.

MacLeod's model is an extension of Fessler's model and adds three multiplying factors that consider the geometry and the material properties of the system. In both models, the fixation force correlates linearly with the force applied to impact the taper. The fixation force, therefore, has a major effect on the stability of the taper connection [6,44–52]. Because the taper angle α is determined by the chosen prosthesis design (typically between 5◦30 and 6◦ [36]), this parameter has to be considered as fixed. The effect of the taper angle α on the fixation force is negligible, as illustrated in Figure 2. The taper angle α should not be confused with the slope of the taper, which is α/2 [43]. MacLeod's equation demonstrates that the influence of the diameter of the head (28 mm to 60 mm) is also relatively small, accounting for a maximum variation of 6.4%, as demonstrated in Figure 3. Both models predict that the fixation force is nil for a coefficient of friction of about 0.05, and this fixation force increases as the coefficient of friction increases (Figure 4).

When all other parameters are equal, MacLeod's model estimates a higher fixation force than Fessler's model by about a third (Figure 2). However, when considering the technical aspects of the different experimental setups, nearly all fixation force values described in the literature correspond to the values estimated by Fessler's formula, ranging 40% to 55% of the impaction force [6,21,44,47–52]. Surprisingly, the values measured by MacLeod et al. do not correspond to those estimated by their own model [43].

**Figure 2.** Influence of the angle of the taper on fixation force, as estimated by Fessler and MacLeod's models. With the impaction force having a linear effect in both models, the fixation force is represented as a proportion of the impaction force. The range of taper angles illustrated covers the range of tapers available commercially in hip arthroplasty. For both models, a coefficient of friction μ of 0.2 was considered. For MacLeod's estimate, a 32-mm head was considered.

**Figure 3.** Influence of the diameter of the head on fixation force, as estimated by Fessler and MacLeod's models. As this parameter is not considered in Fessler's model, only MacLeod's model shows a variability depending on this factor. The influence on fixation force, however, remains relatively small and negligible. Notably, MacLeod's formula provides results with a positive influence for an increasing head size, whereas the measurements describe a decrease of approximately 20% in the fixation force when the head size is increased from 28 to 36 mm. With the impaction force having a linear effect in both models, the fixation force is represented as a proportion of the impaction force.

MacLeod's model is not completely trustworthy for two additional reasons. First, it predicts a higher fixation force than the impaction force, starting from a coefficient of friction of 0.372 (for a 32-mm head) and upward. This would imply the creation of energy within the system, which is clearly impossible. Secondly, while the experimental observations indicated a reduction in the fixation force with an increasing head diameter (with the fixation force of 36 mm heads being 20% less than the fixation force for 28 mm heads [43]), the model predicts the contrary (Figure 3). Considering all these elements, MacLeod's equation will be omitted for further analysis in this review, and only Fessler's equation will be considered.

Therefore, the two main parameters controlling the fixation force are the impaction force, which has a linear effect, and the coefficient of friction μ, which should be as high as possible. Under ideal conditions, the value of μ is approximately 0.15 to 0.25 [41,53–55]. However, the relationship between the fixation force and the coefficient of friction is not linear. This force becomes nil when the coefficient of friction approaches 0.05 (Figure 4).

**Figure 4.** Influence of the coefficient of friction μ on fixation force, as estimated by Fessler and MacLeod's models. For MacLeod's model, a head diameter of 32 mm was considered. With the impaction force having a linear effect in both models, the fixation force is represented as a proportion of the impaction force. The effect of the coefficient of friction μ is not linear. With a coefficient of friction < 0.05, stable fixation is not possible. MacLeod's model not only predicts higher fixation forces than Fessler's model, it also predicts a fixation force higher than the impaction force starting from a coefficient of friction of 0.372 (for a 32-mm head) and upward, which is physically impossible.

Symptomatic trunnionosis appears to be associated with increased head offsets and longer neck lever arms [21,22,56,57]. Micro-grooved tapers, designed to improve stress distributions in ceramic heads, may increase the likelihood of corrosion when combined with metal heads [21,58,59]. However, retrieval studies have not been able to confirm the impact of surface topography, because they have shown comparable fretting corrosion in vivo [21,60]. The variability induced by surface topography is obviously far less important than the other relevant factors in vivo. The association of head diameter with trunnionosis is not always consistent in the literature [12,22,26,57,61,62], but confounding parameters such as time in vivo were not always considered in these studies. Based on some of these publications, large-diameter heads induce force transmission on the taper with

a greater lever arm, and this will be mechanically unfavourable when micromotions have to be minimized [43,57,63]. Interestingly, the influence of the head size is not always seen on retrievals [62]. In addition, and perhaps somewhat counterintuitively, the length of the trunnion does not appear to have a significant impact on micromotion or fixation force [47,64]. Increased trunnionosis has also been linked with low flexural rigidity necks (i.e., those with increased elasticity) [65] and small taper angle differences [66]. However, all these parameters are associated with the chosen design of the prosthesis and may not be modifiable intraoperatively. Taper incongruences caused by mixing heads and tapers from different manufacturers might affect the strength of the taper connection and must be avoided, given the variability of the effect associated with cobalt–chromium (CoCr) heads, because taper incongruences may critically reduce the fracture load of ceramic heads [67,68], and because of certification issues. Gross trunnion failures are reportedly above average for tapers made of beta titanium alloy (titanium–molybdenum–zirconium–iron (TMZF)) [34]. This specific beta titanium alloy might have fretting corrosion characteristics that are unfavourable for taper connections.

#### **3. Patient-Related Factors for Trunnionosis**

Patient-related factors that affect corrosion may include excess body weight [69] and high-impact activities with a resultant increase in demand on the prosthesis [70]. Bergmann et al. have shown that prosthetic loads are several times the patient's body weight, and the exact load level greatly depends on the patient's activity [71]. Impact activities greatly increase prosthetic loads; for example, jogging and brisk walking increased the prosthetic load by 3.9 times a patient's body weight, while stumbling increased the prosthetic load by 11 times a patient's body weight [71]. These larger loads obviously amplify micromotions at the taper interface. However, the surgeon cannot influence these parameters, as they depend on the patient's activities in daily life.

The risk of trunnionosis should be considered when planning for THA in patients with larger femoral neck offsets. Varus orientation of the taper and the use of heads with +4 mm necks or longer significantly increase the risk of micromotion and MACC in the taper interface [21]. Given the risk of trunnionosis, implanting the stem in a varus axis to increase the offset might not be mechanically sound. This might well become an issue, particularly with the so-called short stems, which are often recommended for implantation in varus to reconstruct larger femoral offsets [72]. As heads with longer necks might also be detrimental [21], stem designs with larger offsets might have to be favoured. In general, older designs have smaller offsets. When choosing stems with increased offsets, the reduced flexural rigidity of the neck or the taper should be avoided [65,73]. In our opinion, the traditional solution in THA in compensating offset loss with leg-lengthening should be re-evaluated and updated in light of the increasing awareness of trunnionosis.

#### **4. Surgeon-Related Factors Determining Taper Fixation**

Micromotions in the taper interface, and thus the risk of trunnionosis, are determined by factors under the direct control of the surgeon. These factors are the coefficient of friction and the impaction force, which also appear to be the main determinants of the fixation force, once the prosthetic design has been chosen.

The coefficient of friction is significantly affected by the condition of the trunnion at the time of assembly. This has a major influence on the fixation force, as illustrated in Figure 4. According to several in vitro studies, fluid or fat left on the trunnion at the time of head assembly negatively affect fixation force [10,45,74,75]. By cleaning the trunnion with saline solution and drying it with gauze directly before the assembly of the head, the disassembly forces increase to values observed on pristine control trunnions [10,44,75]. The effect of contamination on the coefficient of friction is independent of the head material (CoCr versus ceramic) [44]. The contamination of the female taper should be avoided carefully while manipulating the head before seating it, as adequate cleaning of the bore would be particularly difficult. A lower coefficient of friction reduces the fixation force and causes higher hoop stresses, increasing the fracture risk for ceramic heads [75]. Drying and cleaning should be done with gauze only, as metallic brushes or pads damage the surface of the taper [76].

The head should never be struck directly with the hammer, but instead with an adequate impactor to avoid damaging the bearing surface of metal heads or fracturing ceramic heads due to point loading. The characteristics of the impactor greatly influence the force transmitted from the hammer blow to the taper connection. Hard plastic tips found on commercially available impactors avoid damage to the head, but they reduce the assembly force by approximately 20% compared to a metallic tip [77]. Older impactors (with questionable stiffness) and impactors with rubber tips should be abandoned [77].

Practical recommendations regarding the impaction force require a discussion of the technical aspects of the studies found in the literature. Quasi-static assembly procedures ensure the precise measurement of the impaction force but do not correspond to the technical solutions available intraoperatively. As the rate of impaction has no relevant effect on the fixation force [49,78], impaction with a hammer might be considered equivalent, even if minor mechanical differences might be identified [55]. While force sensors at the tip of the impactor might approximate the assembly force applied to the taper, the energy dissipated by the impactor must be subtracted from the values measured by the sensing hammers [10,43,50,78,79]. Some studies do not even adequately describe the impaction force used or the fixation force measured [45,74].

The number of hammer blows does not play a significant role, as the impaction force is controlled by the impact with the highest energy [6,45,79]. To ensure the optimal impaction of the taper, two hammer blows are recommended [74]. The first blow could be seen as being the alignment blow and the second blow as being the definitive impaction blow. Many lighter impactions are not useful, as the effect is not cumulative. We recommend a good alignment between the impactor and the neck of the femoral stem. Based on a trigonometric model, a misalignment up to 20◦ could be tolerated, as 94% of the impacting force would still be maintained in the correct direction. This is confirmed experimentally, with fixation forces not altered significantly by seating loads applied at 20◦ [52]. Asymmetries of the seating load displacement could, however, be observed at the taper [51,52], and this might explain notable differences in the fixation force observed with off-axis impactions of only 10◦, when combined in different planes [80]. Axis deviations of more than 20◦ are common during THA, at least when using a posterior approach to the hip [81]. Due to the off-axis orientation of physiologic loads, the manual assembly of a taper, relying on later impaction under the patient's weight, is inadequate [49,70,71].

Impaction with a hammer has its advantages, in that a short impulse causes a lower transmission of energy to the tissues distal to the taper than a slower application of the force [44,55,77]. It might be expected that the impaction of the taper with a hammer is associated with a low risk of femoral fracture, even if high forces are applied. An increased impaction force increases the contact surface, and this has a favourable effect on the stability of the taper connection [39]. However, there is a clinically relevant upper limit to the amount of force applicable. However, to the best of our knowledge, no study has established a maximum recommended force.

Impaction with at least 4000 N is recommended. An adequate impaction force can be reached with a strong blow from a 500 g hammer [46,50]. Surgeons should become familiar and proficient with force-measuring instruments, especially those surgeons with low levels of experience in arthroplasty, to avoid the application of insufficient or off-axis impaction force [46]. The settings of the instrumentation should be checked carefully to ensure that the force measured corresponds to the potential assembly force, and not the force applied with the hammer.

#### **5. Discussion**

Contemporary hip arthroplasty includes modular heads with variable neck lengths, as this increases the surgical options to tailor the implant to the patient's individual anatomy and allows for the use of heads made of materials that differ from that of the stem [7]. However, because of the additional interface, modularity may result in additional specific failures. Trunnionosis, defined as fretting and corrosion of the taper connection [6], may result in metal ions and metal particles entering the joint, causing ALTR. Since the first report of a pseudotumour related to taper corrosion was published as early as 1988 [82], corrosion and fretting at the taper junction have become increasingly linked to implant failure [11–13,20,22,83]. Surgeons should be aware of all controllable factors to minimize the risk of trunnionosis. This review highlights the important controllable factors that determine the stability of the head–neck taper connection: impaction force and the coefficient of friction at the taper interface.

Fessler's model quantifies the fixation force of a taper connection [6]. A higher fixation force leads to a tighter fit between components. This will reduce micromotions in the interface between the bore in the head and the stem's trunnion during physiologic loads, thereby preventing fretting corrosion [10,20,21,40]. The head size is not considered in Fessler's analytical model of fixation force (Figure 1). Since trunnionosis came into focus as clinically relevant, around ten years ago, multiple factors with a possible impact on the lever arm to the centre of rotation (e.g., large-diameter heads, high mediolateral offset, large neck length, and bearing type) have been discussed as factors that may exacerbate the fretting corrosion process [23,29,58,70]. However, the available data remain inconclusive as to their relevance, and their actual impact remains unclear even to date. Other factors excluded from the analytical model, but under the direct influence of the surgeon, are the impaction technique [6] and the avoidance of mismatch by combining heads and stems with different taper designs (i.e., from different manufacturers) [25,67,68,84].

Modern hip arthroplasty is prone to fretting and corrosion at the taper junction for three reasons. First, procedures are increasingly being performed through small incisions [85], which may impair the proper cleaning and drying of the taper due to reduced exposure. Secondly, driven by the trend to increase the range of motion, there has been a reduction in the diameter of the neck and of the length of the tapers, which reduces flexural rigidity [65,73]. At the same time, head diameters have increased with the aim of reducing the risk of dislocation, along with an increasing range of motion [24,86,87], a change justified by the improved wear characteristics of highly cross-linked polyethylene and ceramic bearings [88–90]. Thirdly, these developments have coincided with an increase in obesity rates in most patient populations, with the prevalence of adult obesity exceeding 50% in numerous countries [91]. Taken together, these three developments have created challenges for proper taper fixation and have resulted in corrosion issues.

Finally, there is an increasing understanding of the inter-subject variation in terms of the biologic response to wear particles. Macrophages are activated by wear particles from CoCr alloys [92]. This may lead to the cell-induced corrosion of the taper interface [93]. Improperly seated heads may develop micromotions under physiological loads large enough for macrophages to penetrate the interface and contribute to taper corrosion and failure [52]. This may also be the case with heads engaging the taper proximally, which is the rule for ceramic heads [24]. Differences in alleles are strongly associated with the development of pseudotumours after THA with metal-on-metal bearings [94]. Variations in the genetic signal on the seventh chromosome can influence the probability of developing osteolysis after THA [95].

Trunnionosis develops due to the insufficient fixation of the taper connection. Wear and corrosion alter the surface and the geometry of the trunnion. Despite this, well-fixed stems do not necessarily need to be revised when the trunnion shows severe corrosion. The correct application of a new head does not lead to increased revision rates for corrosion-related issues [11,96,97]. Macroscopic material loss on the trunnion obviously does not allow the proper seating of a new head, and the stem should then be revised. Ceramic heads with titanium alloy inner sleeves may reduce the risk of the recurrence of corrosion issues at the level of the taper compared to CoCr heads [11,97]. However, this remains unconfirmed, considering the difficulties in identifying relevant material issues and the small sample sizes in the studies cited.

In conclusion, trunnionosis is a multifactorial phenomenon related to wear and corrosion in the modular links in hip arthroplasty. A tighter fit decreases micromotion and fretting of the taper interface in the long term and thus reduces the risk of trunnionosis. To minimize micromotions leading to wear and corrosion of the taper connection, surgeons should be aware of the factors directly under their control. Analytical models and empirical investigations reflect the critical significance of carefully cleaning, rinsing, and drying the taper before assembly. The absence of biological residues will yield a higher fixation force between the taper and the femoral head. Fixation strength increases linearly with impaction force. Based on the literature, the adequate impaction of the taper connection can be achieved with at least two strong blows from a 500 g hammer. Surgeons are encouraged to undertake training on force-measuring machines to ensure adequate impaction.

**Author Contributions:** Conceptualization, C.B.R. and P.W.; methodology, C.B.R. and P.W.; formal analysis, C.B.R. and P.W.; writing—original draft preparation, C.B.R. and P.W.; writing—review and editing, C.B.R. and P.W.; visualization, C.B.R. and P.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The corresponding author (C.B.R.) is a salaried employee and owns stock of Zimmer Biomet, a manufacturer of orthopaedic implants. P.W. declares no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **The E**ff**ect of N, C, Cr, and Nb Content on Silicon Nitride Coatings for Joint Applications**

**Luimar Correa Filho <sup>1</sup> , Susann Schmidt 2, Cecilia Goyenola 3, Charlotte Skjöldebrand <sup>1</sup> , Håkan Engqvist 1, Hans Högberg <sup>3</sup> , Markus Tobler <sup>2</sup> and Cecilia Persson 1,\***


Received: 29 February 2020; Accepted: 14 April 2020; Published: 17 April 2020

**Abstract:** Ceramic coatings deposited on orthopedic implants are an alternative to achieve and maintain high wear resistance of the metallic device, and simultaneously allow for a reduction in metal ion release. Silicon nitride based (SiNx) coatings deposited by high power impulse magnetron sputtering (HiPIMS) have shown potential for use in joint replacements, as a result of an improved chemical stability in combination with a good adhesion. This study investigated the effect of N, C, Cr, and Nb content on the tribocorrosive performance of 3.7 to 8.8 μm thick SiNx coatings deposited by HiPIMS onto CoCrMo discs. The coating composition was assessed from X-ray photoelectron spectroscopy and the surface roughness by vertical scanning interferometry. Hardness and Young's modulus were measured by nanoindentation and coating adhesion was investigated by scratch tests. Multidirectional wear tests against ultrahigh molecular weight polyethylene pins were performed for 2 million cycles in bovine serum solution (25%) at 37 ◦C, at an estimated contact pressure of 2.1 MPa. Coatings with a relatively low hardness tended to fail earlier in the wear test, due to chemical reactions and eventually dissolution, accelerated by the tribological contact. In fact, while no definite correlation could be observed between coating composition (N: 42.6–55.5 at %, C: 0–25.7 at %, Cr: 0 or 12.8 at %, and Nb: 0–24.5 at %) and wear performance, it was apparent that high-purity and/or -density coatings (i.e., low oxygen content and high nitrogen content) were desirable to prevent coating and/or counter surface wear or failure. Coatings deposited with a higher energy fulfilled the target profile in terms of low surface roughness (Ra < 20 nm), adequate adhesion (Lc2 > 30 N), chemical stability over time in the tribocorrosive environment, as well as low polymer wear, presenting potential for a future application in joint bearings.

**Keywords:** silicon nitride; coating; joint replacement; wear; adhesion

#### **1. Introduction**

Total joint replacements (TJRs) are surgical procedures carried out most frequently on patients suffering from arthritic pain or bone fractures [1–5]. These procedures are largely considered successful, with success rates up to 90% at 10 years follow-up for total hip replacements (THRs) and total knee replacements (TKRs) [6,7]. However, the aging and more active population places higher demands on these implants.

Typically, for THRs the femoral head is replaced by a metal alloy (CoCrMo) or ceramic (zirconia-toughened alumina (ZTA)), and the acetabulum by a ceramic or polymer (ultrahigh molecular weight polyethylene (UHMWPE) or cross-linked polyethylene (XLPE)) [8–12]. TKRs are composed of a metallic femoral component (CoCrMo) as well as a polyethylene XLPE insert attached in the tray [3,13–16]. Ceramic coatings on metallic substrates can be used to reduce wear of structural materials, including manufacturing tools as well as joint implants [17,18]. Different types of ceramic coatings are under investigation for hip joint applications (e.g., TiN, DLC, ZrO2, ZrN, CrN, and Si3N4), while TiN and ZrN coatings are already in clinical use in knee implants [19–23]. These transition metal nitride coatings have the expressed purpose of extending the implant's life time, by either preventing or minimizing the body's immune reaction, which might result in osteolysis, aseptic loosening, and ultimately implant revision [24–35].

According to our previous work, SiNx based coatings have shown potential as an alternative for joint bearings due to their biocompatibility, high wear resistance and hardness, and reduced metal ion release [36–39]. However, it is challenging to achieve an optimal combination of adhesion, coating density, and reactivity in SiNx coatings; a high coating density resulting in a lower reactivity may give an insufficient adhesion to the substrate due to high residual stresses [40]. Alloying with a third element may be an option to improve the chemical stability while maintaining a balance in coating density and adhesion. For silicon nitride, previous studies have shown that the addition of Cr increases oxidation resistance and mechanical properties [41,42], while Nb improves the wear resistance and increases the hardness [43]. In previous studies, we reported that the addition of C altered the surface reactivity of silicon nitride and influenced the coating density and surface morphology [44,45]. In addition, we have shown that an increased N content results in a higher hardness and density [46–48]. In this study, we investigated the effect of N, C, Cr, and Nb content, as well as ion energy, on the properties of silicon nitride (SiNx)-based coatings for joint applications, with a focus on their wear performance in a hard-on-soft contact, since, as mentioned above, the counter surface in a joint implant is usually a polyethylene polymer. The coatings were deposited on top of Cr-based interlayers, and were evaluated in terms of chemical composition, surface roughness, mechanical properties, adhesion, and wear resistance in a hard-on-soft contact.

#### **2. Materials and Methods**

#### *2.1. Coating Deposition*

Coating deposition was conducted in an industrial coating system, with a chamber volume of about 1 m3, equipped with four magnetrons, of which two were operated in unbalanced magnetron sputtering (UBM) and two in high power impulse magnetron sputtering (HiPIMS) mode. The coatings were deposited using 2-fold substrate rotation. The Si targets were operated at average powers of 5 kW and 8 kW in HiPIMS mode, while the Cr and Nb targets were operated in UBM mode at a sputtering power 1 kW for Cr and sputter powers of 1 kW, 2 kW, or 5 kW for Nb. SiNx coatings with thicknesses ranging from 3.7 to 8.8 μm were deposited on mirror polished CoCrMo discs. Ion energies were controlled using three different bias voltages (low, medium, and high) as well as average target power settings. The sputter atmosphere was controlled at a pressure of 600 mPa, with N2-to-Ar ratios ranging between 17% and 40% and the remaining percentage to reach 100% was Ar. Detailed information can be found in Table 1.


**Table 1.** Description of the coatings and deposition processes used in this study. A pressure of 600 mPa was used for all deposition runs.

#### *2.2. Compositional Analysis*

The composition of the SiNx coatings was investigated by X-ray photoelectron spectroscopy (XPS, Axis UltraDLD, Kratos Analytical, Manchester, UK) using monochromatic Al(Kα) X-ray radiation (*h*<sup>ν</sup> <sup>=</sup> 1486.6 eV). The base pressure in the analysis chamber during acquisition was <sup>&</sup>lt; <sup>1</sup> <sup>×</sup> <sup>10</sup>−<sup>7</sup> Pa. The experimental conditions were such that the full width at half maximum (FWHM) of the Ag3d5/<sup>2</sup> peak from the reference Ag sample was 0.45 eV. For all coatings, XPS survey spectra and core levels were recorded on as-received samples and after sputter cleaning. Sputter cleaning consisted of an initial step of 900 s at an Ar<sup>+</sup> beam energy of 2 keV, followed by a second step for 900 s at an Ar<sup>+</sup> beam energy of 4 keV. During sputter cleaning the Ar<sup>+</sup> beam was rastered over an area of 3 <sup>×</sup> 3 mm<sup>2</sup> at an incidence angle of 20◦. Automatic charge compensation was applied throughout the acquisition, using low energy electrons provided by a flood gun. The composition of the coatings was assessed from XPS high-resolution core level spectra recorded from the Si 2p, Ar 2p, N 1s, C 1s, and O 1s regions after sputter cleaning. Core level spectra were analyzed with CasaXPS (v2.3.15, Casa Software Ltd, Teignmouth, UK). A Shirley-type background was subtracted, and the spectra were calibrated using adventitious surface carbon at 284.8 eV as a charge reference. For quantitative analysis of the metal-containing coatings the core levels of the Cr 2p and Nb 3d were applied for determination. The measurement precision for XPS analysis was ±5% for compositions below 10 at % and ±2–3% for compositions above 10 at % [49].

During wear tests and exposure to fetal bovine serum (FBS) solution a reaction occurred, and a white layer was formed on the surface of some coatings. This layer was examined using monochromatic Al (Kα) X-ray photoelectron spectroscopy (XPS, Quantera II, Physical Electronics (PHI), Eden Prairie, MN, USA). Measurements were conducted on the surface after 2 min of sputtering Ar<sup>+</sup> ions at 500 V and after an additional 20 min at 1 kV, to investigate the coating surface and further down in the coating, respectively. The sample was mounted on a glass slide in order to float the sample and automatic charge compensation was used throughout the measurement. Core level spectra were analyzed in CasaXPS, a Shirley-type background was subtracted, and the spectra were calibrated using adventitious surface carbon at 284.8 eV as a charge reference.

#### *2.3. Surface Roughness*

The coating roughness was measured before wear testing using optical profilometry, specifically vertical scanning interferometry (VSI) at 10× and a field of view (FOV) of 1.0. Each measurement corresponded to an area of 451 <sup>×</sup> <sup>594</sup> <sup>μ</sup>m2. Typically, four measurements were performed on each sample to obtain Ra (arithmetic average).

#### *2.4. Nanoindentation*

The hardness and elastic modulus of the coatings were measured in a CSIRO UMIS nanoindenter (Fischer-Cripps Laboratories, New South Wales, Australia) equipped with a three-sided Berkovich tip. All films were tested in the load-controlled mode and for calculations a Poisson's ratio of 0.3 was used. For the tests, at least 30 indents with a load of 20 mN were performed [50].

#### *2.5. Scratch Testing*

In order to investigate coating adhesion, a scratch test was performed at different time points [51] using a scratch tester (CSEM-Revetest (CSEM, Neuchatel, Switzerland)) with a Rockwell C tip (apex 120◦, tip radius 200 μm). A progressive load up to 100 N, at a loading rate of 120 N/min and a horizontal displacement rate of 6 mm/min were applied. This resulted in a scratch length of 5 mm, which was evaluated in a light optical microscope to determine the critical load LC2 indicating where the adhesion failure occurred [52,53]. Each sample was scratched three times at each time point.

#### *2.6. Wear Resistance (2D) in a Hard-on-Soft Contact*

Multidirectional wear tests (MWT) were carried out to evaluate the response of the coatings against polyethylene using cylindrical pins with a nominal length of 19.1 mm and diameter of 9.5 mm. The pins were made of UHMWPE GUR1020 (one of the two most commonly used grades of UHMWPE in orthopaedics, defined as per BS ISO 5834-2 2019 [54]), provided by the collaborating industrial partners Peter Brehm GmbH (Weisendorf, Germany). MWT tests were performed in 0.2 μm filtered bovine serum solution (25%) at 37 ◦C. Prior to testing, the pins were presoaked in serum and cleaned according to standard [55]. The test was carried out with a nominal load of 150 N resulting in an estimated contact pressure of 2.1 MPa within the guidelines of [56], frequency 2 Hz, and sliding velocity 56 mm/s for 2,000,000 cycles (2.0 MC) using a 7 mm × 7 mm square path for a sliding distance of 28 mm/cycle.

#### *2.7. Statistical Analysis*

IBM SPSS Statistics v 26 (New York, NY, USA) was used for all statistical analyses. An analysis of variance (ANOVA) was performed, followed by a Scheffe's post hoc test. When Levene's test for homogeneity of variances was significant, Welch's robust test followed by a Tamhane post hoc test was used instead. The Pearson correlation coefficient was used to evaluate potential correlations. A critical level of α = 0.05 was used to determine significance.

#### **3. Results and Discussion**

#### *3.1. Coating Thickness and Composition*

The growth rate for the SiNx coatings depended on applied target power settings and bias voltages, as well as the N2 and C2H2 gas flows. Increased SiNx and SiMeNx growth rates resulted from more material being removed from the target due to elevated target potentials [57]. Additionally, as Nb or Cr were added to the process, the number of operated targets increased and contributed to the SiMeNx growth. The growth rate for the different bias settings showed a maximum at a medium level, indicating that the flux of film-forming species at low bias voltages was not optimally directed to the substrate table and resputtering occurred at high bias voltage settings. Increasing amounts of N2 led to decreased growth rates due to target poisoning [47] while an increased C2H2 gas flow resulted in an increased growth rate. This was attributed to a reduction in coating mass density and morphological density, specifically a pronounced growth of columns [45].

Following the trend of the coating growth rate, the coating composition changed by increasing N2 and C2H2 gas flows, leading to increased amounts of N and C in the coating. XPS results showed a nitrogen content close to 50 at % for all coatings (Table 2). SiNx coatings with a higher ion energy and 40% of N2 during deposition yielded a nitrogen content in the coatings exceeding 51 at % and a N/Si ratio ≥1. This ratio has previously been shown to be beneficial to a lower dissolution rate, which could be advantageous to the coating lifespan [39]. When C2H2 was added to the process the coating showed higher O contents, which in turn supported the interpretation for the growth rate of SiCNx at elevated C2H2 flows. Here, the reduction in coating density and the pronounced growth of columns led to incorporation of O as the coatings were exposed to air prior to analysis. Further, a reduced morphological density was observed as Cr and Nb were added to the process [40]. This was mirrored in higher O and C contents in the corresponding coatings.


**Table 2.** Deposition settings, coating thicknesses, growth rates, and composition.

The microstructure of similar coatings has been published earlier, for a range of coating parameters [47]. The SiNx coatings that performed well were very dense and, thus, had low O and C contents, but also displayed more residual stresses. Likewise the coatings that did not perform well were less dense and contained more O and C (by adsorption) [48].

#### *3.2. Surface Roughness*

The average surface roughness (Ra) determined for the as-deposited coatings was <50 nm (Table 3), thus fulfilling the standard for biomedical implants (ASTM F2033-12). Coatings *C-low*, *Standard*, *Nb-medium*, and *Nb-high* displayed the lowest values (7.69–12.97 nm), followed by coatings *Nb-low*, *Si Power-high*, *Bias-medium*, *C-high*, and i (14.71–19.97 nm). The highest values of Ra were obtained for coatings *Bias-high*, *N-medium*, *N-low*, and *N-high* (22.2–42.05 nm). As shown in Table 3, the coatings with higher Nb and C content presented relatively low surface roughness values. A lower surface roughness could possibly be attributed to the ionization energy of N being higher than C, which resulted in more carbon atoms being deposited. On the other hand more of the amorphous phase was being created, resulting in a smoother surface [58–62]. No statistically significant correlation could be found between surface roughness and coating thickness, nor between surface roughness and deposition rate.


**Table 3.** Average surface roughness of SiNx coatings, as measured by interferometry. Coatings attributed with the same letters from a–e were not statistically significantly different (i.e., p > 0.05).

#### *3.3. Mechanical Properties (Nanoindentation)*

The coating hardness varied from 13–25.4 GPa, with coatings *Standard*, *N-high,* and *N-low* exhibiting higher values (Figure 1). A similar tendency could be observed for the Young's modulus.

**Figure 1.** Hardness and Young's modulus for SiNx based coatings.

Earlier studies on SiNx coatings have determined similar values for hardness and Young's modulus, although different deposition methods were applied [39,44,47,63,64]. A higher hardness suggests a higher coating density. Hardness values reported for other coatings for joint implants such as ZrN, TiNbN, Ox-Zr, and TiN coatings resided in a similar range, namely 14.0–31.0, 14.0–24.5 and 12.0–14.0 and 33–56 GPa, respectively [65].

#### *3.4. Adhesion*

The scratch test results in terms of Lc2 values are shown in Figure 2. As can be seen, coatings deposited with a higher target power showed lower LC2 values. This was due to higher residual stresses resulting from a higher N content and the increase in Si-N bonds [45,48]. Moreover, these coatings showed a generally denser morphology (data not shown), which in turn contributed to increased residual stresses [66], as demonstrated previously [48]. Furthermore, coatings with higher O and Cr contents displayed higher Lc2 values, which may be related to a lower amount of N-bonds, i.e., an opposite trend to that previously mentioned and/or a decreased coating density and, hence, residual stresses. The following coatings showed statistical differences: *Standard* vs. *C-high*, *Nb-medium*, *Nb-high*, *Bias-medium*, *Bias-high*, and *Si power high* as well as *C-high* vs. *Nb-medium*, *Nb-high*, *Bias-high*, and *Si power high*.

**Figure 2.** On the left axis the results for adhesion (Lc2) are shown for the coatings tested in this study, with bars and standard deviations. On the right axis the O2 content of the coatings is shown, represented by square dots.

#### *3.5. Multidirectional Wear Tests*

#### 3.5.1. Macroscopic Appearance and Surface Analysis

The macroscopic surface structure of the coatings after the wear tests is shown in Figure 3. The formation of an opaque layer on the surface could be observed during testing on some of the coatings (Figure 3). XPS measurements were, therefore, performed on coating *N-medium*, at a region that still displayed a reflective surface (assumed to be unworn) and a region that had formed an opaque layer on the surface. Previous work showed a tribofilm formation in aqueous environments for Si3N4 materials, and in those conditions a SiO2 and Si(OH)2 layer could be found, improving the wear resistance and reducing the coefficient of friction by acting as a self-lubricating layer [67–69]. However, in the XPS measurements herein the use of charge neutralizers and lack of a good charge reference made the positions of the peeks uncertain. To determine whether the Si2p and O1s peaks originated from Si-O bonds, the distance between the peaks, ΔEb, was determined and compared to the distance (ΔEb) from literature according to Briggs et al. [70]. The deconvoluted Si peaks were fitted with the smallest number of curves possible. The spectrum obtained at the surface revealed contributions attributed to Si-C (100.8 eV), Si-N (101.4 eV), and Si-O (102.8 eV), which correlated well with findings from similar materials. After 2 min sputtering at 500 V the Si-O contribution was no longer detected, while there were still contributions attributed to Si-N and Si-C, and after additional sputtering for 20 min at 1 kV only two contributions were identified, Si-N and Si-Si. These results indicated that the outer layer contained more O and C compared to the bulk of the coating, which could be due to the formation of a tribofilm during wear testing.

**Figure 3.** Typical macroscopic appearances of (**a**) a reacted surface (Standard), (**b**) a failed coating (Nb-medium), and (**c**,**d**) coatings with a surface layer: (**c**) Coating Cr and (**d**) coating Si Power-high. In (**e**) a Bias-high coating is shown, which did not present any layer formation or upcoming failure up to 2 MC.

#### 3.5.2. Coefficient of Friction

Low coefficients of friction were observed for the first 0.5 million cycles for *N-high*, *Standard*, *Bias-medium*, and *Bias-high* (0.051–0.067). Coatings *N-low*, *Cr*, and *Si Power-high* showed somewhat higher values, from 0.103–0.108, with little variation. Coefficients of friction did not change markedly throughout the tests, except for coatings that reacted or failed during the test (Figure 4). This work generally showed lower coefficients of friction compared to previous work on Nb-Ti-N coatings (ranging from 0.11 to 0.12) and on TiN (0.14) [66].

**Figure 4.** Coefficient of friction up to 2.0 MC for the tested coatings. During wear tests the following coatings wore through: (\*) N-medium, C-high, and C-low at 0.5 MC; Nb-medium at 1.5 MC; Nb-low and Nb-high at 2.0 MC.

#### 3.5.3. Volumetric Wear Rate

While *N-low* and *N-high* gave the lowest wear rates for the UHMWPE pins (< 0.37 mm3/MC, Figure 4), N-medium failed already in the first 0.5 MC (Figure 3), giving a high wear rate due to the increased surface roughness from the reacted surface. The *Standard* coating also gave a high wear rate, due to a reacted surface (Figure 3). The coatings with a higher C content all failed at 0.5 MC. Nb coatings failed at different time points, for example, *Nb-low* and *Nb-high* had failed at 2.0 MC and *Nb-medium* at 1.5 MC. The remaining coatings did not fail and presented low wear rates (0.74–3.63 mm3/MC). Figures 3 and 5 show that coatings with no apparent reaction or coating failure, and that gave low pin wear rates, were those with an initially high hardness (22.5–28.4 GPa), and, hence, presumably higher density and lower reactivity and/or a high N content (*N-low*, *N-high*, *Bias-medium*, *Bias-high*, and *Si Power-high*), with the exception being the *Standard* (H = 23.4 GPa) coating, which, however, contained more oxygen than the well-performing coatings (Table 2), suggesting a higher reactivity. Coating wear through contact with UHMWPE during the tests was not expected, and coatings failing would rather be associated to a higher reactivity and subsequent dissolution [40].

**Figure 5.** Volumetric wear rate for the UHMWPE pins ran against all coatings in this study. Coatings C-high and C-low showed negative values of <sup>−</sup>2.90 and <sup>−</sup>8.40 mm3/MC, respectively.

#### **4. Conclusions**

Based on the results of the coatings tested in this work, some important conclusions were drawn. First, the low-ion energy coatings generally exhibited a lower hardness and initially higher critical load in scratch testing. High concentrations of impurities (higher O content and lower N content) were associated with early reactions and/or dissolution of the coating, as shown by XPS compositional analysis as well as multidirectional wear tests. During the wear tests coatings with lower or no apparent O content did not fail and showed a low volumetric wear rate of UHMWPE pins. SiNx coatings of high N content, low O content (e.g., *N-high*, *Bias-medium*, *Bias-high*, and *Si Power-high*) are needed for the target–joint implant applications. Promising low wear rates were found for UHWMPE pins sliding against these latter coatings in a multidirectional wear test.

#### **5. Patents**

Ionbond AG, where S.S. and M.T. are employees, owns patents related to similar coatings.

**Author Contributions:** Conceptualization, C.G., M.T., S.S., H.E., H.H., and C.P.; methodology, L.C.F., C.S., M.T., S.S., and C.P.; validation, L.C.F., C.G., C.S., M.T., and S.S.; formal analysis, L.C.F., C.G., C.S., M.T., S.S., and C.P.; investigation, L.C.F., C.G., C.S., M.T., and S.S.; resources, C.P., H.H., and H.E.; writing—original draft preparation, L.C.F., M.T., S.S., and C.P.; writing—review and editing, L.C.F., S.S., C.G., C.S., H.E., M.T., H.H., and C.P.; visualization, L.C.F., S.S., and C.P.; supervision, H.E., H.H., and C.P.; project administration, H.H., H.E., and C.P.; funding acquisition, C.P., H.E., and H.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the European Union, grant number FP7-NMP-2012-310477 (Life Long Joints project); EBW+ Project Erasmus Mundus Programme, Action 2–STRAND 1, Lot 9 (Latin America), Brazil, Grant number 2014-0982. H.H. acknowledges financial support from the Swedish Government Strategic Research Area in Materials Science on Advanced Functional Materials at Linköping University (Faculty Grant SFO Mat LiU No. 2009 00971).

**Acknowledgments:** The authors are grateful for assistance from Mathilde Cogrel and Alejandro López for assistance with MWT measurements.

**Conflicts of Interest:** The authors declare no conflict of interest, and Ionbond AG, where S.S. and M.T. are employees, owns patents related to similar coatings.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Functional Characteristics and Mechanical Performance of PCU Composites for Knee Meniscus Replacement**

#### **Adijat Omowumi Inyang \* and Christopher Leonard Vaughan**

Division of Biomedical Engineering, Human Biology Department, Faculty of Health Sciences, University of Cape Town, Anzio Road, Observatory, Cape Town 7925, South Africa; kit.vaughan@uct.ac.za

**\*** Correspondence: wumi.inyang@uct.ac.za; Tel.: +27-62-207-6069

Received: 27 February 2020; Accepted: 30 March 2020; Published: 17 April 2020

**Abstract:** The potential use of fiber-reinforced based polycarbonate-urethanes (PCUs) as candidate meniscal substitutes was investigated in this study. Mechanical test pieces were designed and fabricated using a compression molding technique. Ultra-High Molecular Weight Polyethylene (UHMWPE) fibers were impregnated into PCU matrices, and their mechanical and microstructural properties evaluated. In particular, the tensile moduli of the PCUs were found unsuitable, since they were comparatively lower than that of the meniscus, and may not be able to replicate the inherent role of the meniscus effectively. However, the inclusion of fibers produced a substantial increment in the tensile modulus, to a value within a close range measured for meniscus tissues. Increments of up to 227% were calculated with a PCU fiber reinforcement composite. The embedded fibers in the PCU composites enhanced the fracture mechanisms by preventing the brittle failure and plastic deformation exhibited in fractured PCUs. The behavior of the composites in compression varied with respect to the PCU matrix materials. The mechanical characteristics demonstrated by the developed PCU composites suggest that fiber reinforcements have a considerable potential to duplicate the distinct and multifaceted biomechanical roles of the meniscus.

**Keywords:** composite; fibers; polycarbonate-urethane; meniscal replacement; mechanical properties; meniscus

#### **1. Introduction**

The meniscus is a complex and vital biomechanical fibrocartilaginous tissue in the knee joint. The menisci are important structures, as they have a participatory role in shock absorption, joint lubrication and joint congruency [1–4]. As a result of a large amount of force borne by the meniscus, it is frequently subjected to tear, and it wears out with time. Meniscal tears have been extensively reported as one of the most recurrent knee injuries [5–9]. Damage to the meniscus affects its load sharing and distribution roles performed in the knee, which has been linked to the degeneration of the articular cartilage and a high risk of the devastating disease, osteoarthritis [9–11].

Meniscus removal has considerable repercussions for the joint, as it causes abnormal contact pressures, resulting in joint degeneration [12]. As a result of the consequences of meniscectomy, alternative measures have involved the repair or replacement of the meniscus. However, limited success has been achieved with available options owing to various limitations, such as repair only being possible when the tear occurs in the vascularized area, which does not heal well due to shortage of blood supply [13]. Allografts are alternatives for replacing the meniscus of younger patients with a meniscectomized knee joint [14]. Although allografts have acceptable clinical outcomes, the long-term examinations revealed debatable protective effects to the cartilage [15]. Besides, meniscal allografts have difficulty with remodeling and lack adequate strength [12]. Consequently, allografts cannot

be an absolute cure for post-meniscectomy pain [14]. On the other hand, meniscal replacements are biomechanically suitable, and have a distinct advantage of acting as substitutes in the cases of multifaceted tears, as well as alleviating the excruciating pain related to meniscus damage [16].

Different replacements have been sought for the meniscus using either synthetic scaffolds [17,18], natural meniscal tissues or composite materials [19–24]. Amongst the first meniscus replacements developed were permanent meniscal substitutes made from teflon and dacron [11,25,26]. These alternatives were found incompatible after in-vivo testing in rabbits as particles resulting from wear were deposited on the implant. Besides, mechanical integrity was compromised. Similarly, researchers have worked on scaffolds of poly(lactic-co-glycolic acid) embedded with polyglycolic acid fiber [27]. Although the in-vivo studies showed that the scaffolds produced meniscus-resembling tissue after ten weeks' implantation, mechanically the modulus of the tissue was inadequate, as it was not comparable with that of the native meniscus. Another development was a composite of hyaluronic acid and the polycaprolactone matrix with poly-lactic acid reinforcement fibers. The in-vivo studies established that the composite materials support meniscal tissue growth devoid of any adverse effect on the cartilage after implantation, but there was failure resulting from implant extrusion [22,28].

Balint et al. [29] used a different approach in developing a total meniscal substitute where a porous scaffold of collagen-hyaluronan matrix with degradable poly (desaminotyrosyl-tyrosine dodecyl ester dodecanoate) reinforcement fibers was studied. These scaffolds proved to be successful, with considerable mechanical properties suitable as meniscal substitutes; however, implant extrusion remains a challenge [30,31]. The use of polyvinyl alcohol (PVA) hydrogel as a choice material for permanent meniscus replacement has been investigated [32,33]. The study showed the inability of the implant to protect the articular cartilage, and the authors concluded that the failure could be due to size incongruity and inefficient fixation [34].

Studies have explored the possibility of incorporating Ultra High Molecular Weight Polyethylene (UHMWPE) fibers into different matrices, such as poly (vinyl alcohol) hydrogels and polycarbonateurethane [35,36]. The former group fabricated meniscal substitutes that showed promising mechanical properties and manufacturability. They reported limitations, including delamination and implant extrusion [37,38]. In the latter study, the mechanical behavior of the meniscal prosthesis was determined mathematically and experimentally using a knee model subjected to compressive loading. However, it is indispensable to fully characterize the mechanical properties of the developed composite material in-vitro. Moreover, the finite element model is an approximation of the numerical model [39].

Although several researchers have worked on the development of meniscal substitutes, most of their attempts have been focused on investigating the biological characteristics of the implants from a tissue engineering perspective. While these factors are critical for body replacement parts, the mechanical properties of the developed implants are not given adequate consideration. This gap is evident from the failure reported in literature [40–42]. As a result, there is a dearth of information describing the mechanical requirements and functional performance of replacement for the meniscal tissue. Due to the load-bearing capabilities of the meniscus and its exposure to millions of cycles on an annual basis [43,44], the importance of evaluating the mechanical characteristics of the implant material intended to replace the meniscus cannot be overemphasized, since this will provide critical information, such as safety before implantation. Furthermore, current failures associated with existing permanent synthetic meniscal implants, such as insufficient strength, durability, dislocation, wear and fracture [40] further buttressed the need for assessing the mechanical behavior of a material proposed to replace the meniscus. Thus, the choice of appropriate materials for design, the geometry and the mechanical attributes, could produce a suitable candidate for replacing the meniscus.

The meniscus, being a bundle of cartilaginous tissue, has a complex make-up of material properties which vary with location as well as direction [45]. Its peculiar and particular roles emanate from its unique chemical, physical and biomechanical composition, as well as its distinctive structural architecture [46].

Therefore, in order to restore the biomechanical tasks of a worn-out meniscus, it is important to circumspectly select a replacement material having biomechanical properties as close as possible to those of a native meniscus. Though the meniscus demonstrates site-dependent properties, both in tension and compression, it is anticipated that a composite material can be precisely fitted to duplicate these properties, and hence replace the meniscus. Thus, an isotropic matrix material may not be able to produce the inhomogeneous and anisotropic characteristics exhibited by the natural meniscus [45], but fiber-reinforced polycarbonate-urethanes (PCU) could be customized and adapted based on a suitable choice of reinforcement fiber and its orientation within the PCU matrix. Additionally, the use of fiber-reinforced composite will enable the reinforcing fibers to act as a channel for attaching the implant to the joint capsule to cater for dislocation during motion. Incorporating fibers with high strength into soft polymeric matrices to engineer synthetic meniscal replacement has not been extensively explored. Consequently, this work aimed at developing a structured, tailor-made, PCU-reinforced UHMWPE (referred to as PE) composite as a meniscal substitute. Several surface engineering approaches have been introduced in order to modify the surface structure of UHMWPE for improved biomedical applications. These methods, which include plasma techniques [47], laser surface modification [48] and incorporating particulate or fibrous reinforcements [49,50], have been used to enhance the mechanical, tribological and biological properties of UHMWPE substrates.

Medical grade polyurethanes have been widely promoted for biomedical applications [51]. In particular, the use of PCU has drawn considerable attention in the orthopedic device industry as a result of their excellent mechanical properties, biostability and biocompatibility [52]. PCUs have been extensively utilized in vascular grafts, stents, catheters, pacemaker leads and artificial heart valves [53]. Specifically, bionate thermoplastic PCU, commercially produced by DSM PTG (Berkeley, CA, USA), has been of great interest in the field of orthopedics, because of its outstanding load-bearing properties and excellent wear resistance, which enables it to overcome the setback of osteolysis. Its superior characteristics have made it an excellent material selected for hip and knee joint prostheses, prosthetic spinal discs and the shoulder joint system [54]. Besides, they offer long term durability and resistance to hydrolytic degradation, making them outstanding for in-vivo orthopedic applications [55]. The use of PCU for meniscus replacement stemmed from its unique weight-bearing capabilities, the ability to withstand intense forces within the knee joint [56], and an ease of lubrication due to its hydrophilic nature. Moreover, it proffers low friction properties [57–59] to promote movement within the meniscal compartment, while withstanding repeated stresses from the femoral condyle during flexion and extension motions.

Therefore, the overall goal of this study was to develop a meniscal substitute with mechanical properties closely matching those of the native meniscus. For that reason, test pieces of PCU-PE composites were designed, fabricated and evaluated to determine their suitability as a replacement for the meniscus capable of replicating the closest possible mechanical behavior of the native meniscal tissue. The mechanical test pieces were made up of longitudinally-arranged fibers, such as to duplicate the orientation of the circumferential collagen fiber existing in the human meniscus. The effects of the reinforcing fibers on the mechanical properties of the PCU were investigated, and the composites examined and appraised as meniscal implants. The fibers were able to provide a substantial increase in the stiffness of the PCU matrix and enhance the fatigue and abrasion resistance for long term implantation in the knee joint capsule.

#### **2. Materials and Methods**

#### *2.1. Materials and Processing*

Biomedical Bionate 80A and 90A polycarbonate urethane (PCU) from DSM (PTG, Berkeley, CA, USA) were used as the matrices, and the reinforcing fibers were Ultra High Molecular Weight Polyethylene (UHMWPE) continuous strand fibers, Dyneema Purity® UG from DSM (designated as PE). Initial drying of the PCU pellets was done at 100 ◦C for 14 h in a vacuum oven, as stipulated by the supplier and established by Geary et al. [55] as the optimal drying conditions for PCUs to reduce the moisture content to about 0.01%. The intrinsic properties of the PCUs and the reinforcing fibers are detailed in Table 1, as stated in the supplier's data sheets.

**Table 1.** Characteristic properties of the polycarbonate-urethanes (PCUs) and PCU-reinforced Ultra-High Molecular Weight Polyethylene (PE) fiber.


#### *2.2. Composite Preparation*

A stainless-steel, tailor-made mold was fabricated to produce the composite mechanical test samples (Figure 1). The mold, which encloses the fiber, consists of different parts that are assembled to provide the facile removal of cured samples. The composite samples were formulated using a 5% fiber volume fraction. The percentage of the various constituents was based on a computation of fiber diameter, fiber length and the number of fibers, as discussed in our previous work [60]. The composite material was prepared using the combinations of the different PCUs and the reinforcement fibers. The specimen types and their constituents are described in Table 2.

**Figure 1.** The fabricated mold for the test pieces showing (**a**) The entire mold assemblage (**b**) The fiber holes within the mold.


Bionate 90A — MX2

PE MP1

PE MP2

**Table 2.** Material constituents for the specimen types.


#### *2.3. Mechanical Evaluation*

The mechanical test samples were tested for tensile, Ft, and compressive, Fc, properties (Figure 2) using a Zwick/Roell 1484 Material Testing Machine (Zwick GmbH & Co. KG, Ulm, Germany). Each rectangular cuboid tensile specimen of 70 × 19 × 6 mm, was tested at a crosshead speed of 12 mm/min (Figure 2a). Cubic specimens of 6 × 6 × 6 mm were tested for compression, Fc, at a crosshead speed of 5 mm/min (Figure 2b). The moduli for the tensile and compressive tests were calculated from the slope of the linear region of the stress–strain plots. The tensile modulus was taken as the slope of a linear curve fit between 0% to 5% strain, while the compressive modulus was between 2% and 8% strain. These tests were also performed on 100% virgin PCU samples. Three specimens were analyzed in each of the tests. All the results were computed as mean ± standard deviation, and with Excel software (Microsoft, Washington, USA), the unpaired Student's t-test was used to assess the data statistically for significant difference at *p* < 0.05.

**Figure 2.** Graphical representation of the mechanical test pieces describing the fiber arrangement and direction for (**a**) tensile testing, Ft and (**b**) compression testing, Fc. The diagrams are not drawn to scale.

#### *2.4. Microstructural Analysis*

The arrangements and alignments of the fibers in the matrices were studied using a Wild M400 photomacroscope (Wild Heerbrugg, Gais, Switzerland). In addition, the Nova NanoSem 230 scanning electron microscope (SEM) (FEI, Holland, Netherlands) was used to examine the morphology of the PCUs and their composites, and to investigate the fractured surfaces of the tested samples after failure during the mechanical testing.

#### **3. Results and Discussion**

#### *3.1. Microstructural Characterization*

In order to study the distribution of fibers in polymeric matrices, a useful approach is to examine their fractured surfaces after mechanical testing through scanning electron microscopy. An accurate and detailed investigation of the fractured surfaces provides information on the nature of the interfacial bonding existing within the fiber–matrix interface. It explains the phenomenon taking place during the deformation process. Also, the knowledge of the failure mechanisms of the PCU and their composites is crucial in evaluating their long-term mechanical reliability for the proposed applications.

Consequently, the surface morphology of the fabricated samples was microscopically viewed to investigate the effects of the reinforcement fibers. The observed micrographs of the PCU matrices and their reinforced composites from the light microscope are presented in Figure 3a–d. The photomicrographs showed that the surface of the reinforced matrices appeared relatively smoother and cleaner with minimal surface imperfections, compared to the as-molded polymeric matrices.

**Figure 3.** Photomicrographs showing the molded samples of (**a**) MX1 (**b**) MX2 (**c**) MP1 (**d**) MP2.

The fibers were almost evenly distributed within the PCU matrix and at equal distances to one another (Figure 4a). A micrograph of the cross-sectional view of the sample observed from the same microscope corroborated the previous observation (Figure 4b).

Furthermore, the SEM examination, as indicated in Figure 4c, ascertained the uniform distribution of the fiber orientation. The fiber holes, as seen in Figure 4c, were respectively arranged parallel to one another. At higher magnification, the PE fiber was observed to be thoroughly embedded in the polymeric PCU matrix with the surrounding formed ring of PCU-PE fiber interface (Figure 4d).

The fracture behavior of the as-molded PCU samples is shown in Figure 5a,b. The illustrative SEM images showed a sharp, rough and angular fractured surface of the PCU polymer. This surface is characterized by striations suggesting the direction of the crack propagation. The large surface area characterized by smooth regions reveals fast brittle fracture, which is typical of elastomeric PCUs [61,62]. The minimal protrusions disintegrating at the fractured side on the relatively smooth surface (Figure 5b), suggest somewhat the plastic deformation of the PCU polymeric matrix.

On the other hand, PCU-PE composites showed a smooth surface with the fibers firmly held intact in the matrix (Figure 5c,d). A magnified view of the fiber showed an irregular, well-bonded PCU-PE surface. A comparison of the fiber arrangement in the PCU matrix before and after tensile loading gives the information on what ensued when the tensile load was applied (Figures 4b and 5c). SEM images showed the fractured surfaces of the composite material under tensile loading along the fiber direction, with none of the fibers displaced from their average initial positions (Figure 5c).

The SEM images of the PCU-PE composites after the tensile loading showed no evidence of PE fiber pull-out from the PCU matrix, as no void was observed on the microscopically examined surfaces (Figure 5c,d). This observation suggests that there is a strong interfacial bonding that exists between the PCU and the PE fibers. The strong adhesion transmits the applied load from the PCU matrix to the PE fibers, producing a significant rise in the overall mechanical characteristics of the composite material. The existence of the PCU matrix on the surface of the fiber shows there is a strong interaction between the fibers and the PCU, as seen in Figure 5d. The interfacial bonding strength has a corresponding positive effect on the mechanical properties of composite materials [63], hence the increased stiffness observed in the reinforced PCU compared to the virgin PCUs.

**Figure 4.** Representative photomicrographs showing (**a**) a transverse sectional view of the arrangement and alignment of the fibers within the composite (**b**) a cross-sectional view of the arrangement and alignment of the fibers within the composite. Scanning electron microscopy showing a cross-section of (**c**) the distribution of the PE fibers within the PCU (**d**) a magnified view of the PE fiber embedded in the PCU matrix.

**Figure 5.** Scanning electron microscopy (SEM) showing fractured surfaces of the samples under tensile loading for (**a**) B8 matrix (**b**) cross-sectional view of the fractured surface of B8 matrix. (**c**) PE-PCU composite along the fiber direction B8 composite (**d**) a closer view of the fractured PE fiber.

#### *3.2. Mechanical Properties*

The stress–strain graphs for the average values of the tested samples for both tension and compression tests were plotted in Figures 6 and 7, respectively. These curves exhibited a linear pattern at low strains. Subsequently, a considerable change followed in the slope presenting a nonlinear behavior that continued until the tested samples began to fail. Both MX1 and MX2 demonstrated typical elastomeric stress–strain behavior [55,64]. The PCUs were not as stiff as their fiber-reinforced composites, which indicates that the increase is a function of the stiffness of both the matrix and the interspersed fibers.

**Figure 6.** Average tensile stress–strain plots for the PCUs and the PCU-PE composites.

**Figure 7.** Average compressive stress–strain plots for the PCUs and the PCU-PE composites.

A meniscal substitute must be able to perform similar functions as the natural weight-bearing meniscal structure. To this end, the tensile and compressive properties of the PCUs and their fiber-reinforced composites have been evaluated relative to the meniscus tissue to optimize them as potential meniscal replacements. The empirically calculated tensile moduli for the PCU matrices were rather disparate to those of the supplier (Table 1). These discrepancies could be as a result of the manufacturer's test conducted at 37 ◦C conditioned in water, since PCU properties are temperature-dependent (DSM-PTG). Besides, their samples were annealed for 24 h at 70 ◦C before testing. Generally, the moduli of all reinforced specimens were higher than their unreinforced counterparts. The tensile modulus of the meniscus is site-dependent, and hence varies relative to the area and direction. The circumferential tensile modulus of the human meniscus varies between approximately 58 MPa and 295 MPa, while the radial tensile modulus varies between approximately 3 MPa and 60 MPa [65–68]. Therefore, it is expedient that a meniscal device should possess a circumferential tensile modulus of at least 58 MPa in order to prevent deformation, as well as implant extrusion maximally. Both of the PCUs studied offer much lower stiffness (Figure 8). Thus, they will not appropriately perform the rigorous tasks that the meniscal tissue is subjected to on a routine basis. Reinforcing the soft polymeric matrices with durable, high-performance fibers such as PE could therefore possibly construct a composite material which is biomechanically acceptable to replace the worn-out meniscus.

**Figure 8.** Tensile and compressive properties of the PCU-PE composites compared with their unreinforced matrices.

The tensile moduli of composites of MX1 and MX2 increased appreciably with the incorporation of the PE fibers. MP1 and MP2 exhibited higher stiffness than the PCU matrices, as the fibers significantly (*p* < 0.05) enhanced the tensile properties of the PCUs. Percentage increases of 227% and 148% were obtained for MX1 and MX2, respectively. Interestingly, a similar trend was observed in the tensile characteristics of the curves of the following pairs: MX1 and MX2, and also MP1 and MP2, which revealed similar patterns (Figures 6 and 7). This suggests that the behavior of the fiber is similar, irrespective of the matrix material, further establishing the role of the fibers in influencing the overall performance of a composite material. The function of the fibers could be further understood by the details seen in the micrographs (Figure 5c,d), where the fibers were oriented, and remained in their original positions as they "fought" to withstand the applied tensile load.

The compression behavior of the implant device is pivotal to its overall performance, since the meniscus provides support for knee joint stability [69]. In addition, the compressive modulus is of great significance, as it resists the high stresses and transmits the compression loads exerted by the femur on the tibia.

Unlike the tensile moduli, there was a variation in the changes observed in the compressive moduli with the inclusion of fibers in the PCUs compared to their unreinforced counterparts. All the specimen types behave similarly in compression irrespective of their constituent's composition. The menisci are reported to transmit approximately 50%, and about 85%, of the compressive forces exerted in the knee in extension and 90◦ in flexion, respectively [70,71]. This role is made possible by the distinctive arrangement of the collagen fibers, which withstand the high stresses produced during the load-bearing. Consequently, a meniscal replacement must be able to reproduce the aforementioned characteristics and peculiarities associated with the native meniscus. An extensive range of compressive moduli values has been published, in which the variation was controlled by strain and loading rates, as well as the type of test conducted. The aggregate compressive modulus varying between 0.10 and 1.13 MPa has been reported for the native meniscus [65,72–74]. In this study, the minimum compressive modulus was recorded for MP1 composite with a 4% rise with the addition of fibers, while MP2 produced a considerable 55% reduction with the inclusion of fibers. The addition of the PE fibers to MX2 were found to be statistically significant in compression. The difference observed in the compressive moduli of MX1 and MX2, and their composites, showed the influence of fiber reinforcements in the compressive properties of the PCU matrices. Although the values obtained are not comparable to those reported for the human meniscus, a higher compressive modulus will be tolerable and acceptable for a meniscal substitute, since some polymers and metals whose compressive moduli are much higher have been utilized as spacers in knee replacement devices [75]. The mechanical characteristics of composites are ultimately determined by the interfacial bond strength between the fiber and the matrix, which is dependent upon the type, shape, orientation and texture of the fiber surface. Consequently, the PCU-reinforced composites can be customized to mimic the desired properties of the native meniscus.

While the PCU composites exhibited excellent mechanical properties in tension and compression, MP1 produced a relatively high tensile modulus close to the natural meniscal range of values and a lower desirable compressive modulus. This comparatively low stiffness could be attributed to a drawback of the fibers encountered during the processing of the composite samples. The PE molecules tend to exhibit some form of relaxation and reorientation, even below the melting point. Besides, under rigorous loading conditions, the fiber molecules can slide, forming new arrangements, which phenomenon in the long run elongates the fiber, thus causing a reduction in tension leading to failure. These molecular changes may trigger a loss in the tensile properties depending on factors like temperature, time and loading conditions (Dyneema Purity® UG, DSM). Therefore, it is anticipated that PE fibers with a higher melting point than the MX1 will produce exceptional results, both in tension and compression. During repetitive tensile stress, failure of the PCU composite can result from fiber fracture or fiber-matrix interfacial debonding. In such a case, the fibers can be detached from the matrix. When the applied tensile load extends the matrix beyond the fibers, the PCU composite will withstand shear at the fiber–matrix interface, which could cause it to fracture [76].

When the meniscus is subjected to an axially compressive force, the load is distributed over its surface area. Due to the meniscal structure, the transmitted force tends to cause the tissue to extrude radially. This structural malalignment is opposed by the hoop stresses generated in the circumferential collagen fibers [2,3]. These tensile stresses, developed within the meniscus during loading, control their function, and are responsible for failure [67]. The ultimate tensile stress of the native meniscus varies with respect to region. The average maximum stress has been reported to be 18.8 MPa, 17.6 MPa and approximately 4 MPa for lateral, medial and radial meniscal tissue, respectively [66]. Consequently, an average ultimate tensile stress of at least 18.8 MPa would be ideal for a meniscal substitute material. Of the composites, the MP1 performs extremely well within this limit.

The ultimate tensile strengths of the PCUs were lower than those from the manufacturer's data sheets (Table 1 and Figure 9). This could be as a result of the disparities in the shape and size of the tested samples, and also the testing technique employed. The average ultimate tensile strength of both composites decreased with fiber reinforcement, and reductions of 36% and 70% were calculated for both MP1 and MP2, respectively. These decreases could be because of the integral sample property, following that tensile strength is a characteristic of both the component materials and the composite samples under examination. Moreover, the tensile properties of polymeric matrix composites are considerably dependent on several factors like the matrix–fiber interface, geometry, distribution and micromechanical deformation [77].

**Figure 9.** Elongation at break and ultimate tensile strength of the PCU-PE composites compared with their unreinforced matrices.

The elongation at break decreased when the PE fibers were incorporated into the PCU matrices (Figure 9). Elongation at break is a measure of the ductility of a polymeric material, and indicates its ability to resist changes in shape without failing. With PE fibers-incorporated PCU composites, the elongation at break for MX1 and MX2 were reduced by about 28.5% and 61.8%, respectively. Fiber reinforcement of polymeric matrices increases the stiffness and toughness of composites. As the stiffness increases (Figure 6), there is a decrease in the ductility of the composite material, and hence the calculated elongation at break. Previous work has reported similar occurrences [78–80]. The value reported in the manufacturer's datasheet for the PCU suggests that it can attain percentage elongation as high as 531%. At the same time, results from this study showed a higher maximum strain to failure of 619%. The changes in the percentage fracture strains could be due to factors like better surface finish and the firmness of the tensile grips. The wide-ranging tensile and compressive characteristics of the native meniscus complicates the development of a comparable meniscal implant.

Since the functioning of the meniscus is critically reliant on its multiplex shape and structure, it is anticipated that a "true" reflection of the mechanical properties of the fabricated composites could be appropriately exhibited when the prosthesis is designed and manufactured to reproduce the structural configuration of the normal meniscus.

#### **4. Conclusions**

The mechanical performance of the PCU matrices determined from this study for the PCU matrices showed they are inadequate, and cannot replace nor sufficiently perform the load-bearing functions of the meniscus. In general, the effect of the fiber reinforcement was favorable, as the tensile modulus was significantly raised to a value within an acceptable tensile modulus of the human meniscus. The results from this study demonstrate that the PCUs can be customized to fit that of the meniscal tissue, by methodically implanting circumferential fibers into the PCU matrix to obtain a meniscal device with desirable mechanical properties. These results visibly revealed the positive effects of the reinforcing fibers.

Furthermore, the microstructural analysis revealed the failure mechanisms during mechanical testing. The embedded fibers in the PCU-PE composites prevented the brittle failure and plastic deformation exhibited in the fractured PCUs. The excellent interfacial bonding strength within the PCU-PE composites produced a corresponding positive effect on the mechanical properties of composite materials. Hence the increased stiffness observed in the reinforced PCUs compared to the virgin PCUs.

This work provides an insight into the mechanical and microstructural performance exhibited by the PCUs and their composites, hence their suitability for artificial bearing surfaces. Further characterization of the composite materials is required to determine their tribological behavior as a meniscal replacement.

**Author Contributions:** Conceptualization, A.O.I. and C.L.V.; Methodology, A.O.I.; Investigation, A.O.I.; Writing—Original Draft Preparation, A.O.I.; Writing—Review and Editing, A.O.I. and C.L.V.; Supervision, C.L.V. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no funding from any funding agency in the public, commercial, or not-for-profit sectors.

**Acknowledgments:** The authors will like to thank Mrs. Miranda Waldron for her assistance with the SEM at the Aaron Klug Centre for Imaging and Analysis, Electron Microscopy Unit, University of Cape Town.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article* **Quantitative Measurements of Backside Wear in Acetabular Hip Joint Replacement: Conventional Polyethylene Versus Cross-Linked Polyethylene**

#### **Ste**ff**en Braun, Sebastian Jaeger, Robert Sonntag , Stefan Schroeder and J. Philippe Kretzer \***

Laboratory of Biomechanics and Implant Research, Heidelberg University Hospital, Schlierbacher Landstraße 200a, 69118 Heidelberg, Germany; Steffen.Braun@med.uni-heidelberg.de (S.B.); Sebastian.Jaeger@med.uni-heidelberg.de (S.J.); Robert.Sonntag@med.uni-heidelberg.de (R.S.); Stefan.Schroeder@med.uni-heidelberg.de (S.S.)

**\*** Correspondence: Philippe.Kretzer@med.uni-heidelberg.de; Tel.: +49-6221-56-29209

Received: 25 February 2020; Accepted: 7 April 2020; Published: 15 April 2020

**Abstract:** As shown in previous studies, the modification of conventional polyethylene (CPE) to cross-linked polyethylene (XLPE) and the contribution of antioxidants result in a reduction in total wear. The aim of this study was to evaluate XLPE inserts with vitamin E and CPE regarding their resistance to the backside wear mechanism. A cementless hip cup system (Plasmafit® Plus 7, Aesculap) was dynamically loaded using CPE and XLPE inserts. The backside wear was isolated, generated and collected using the two-chamber principle. The chambers were filled with ultrapure water. After 2 <sup>×</sup> <sup>10</sup><sup>6</sup> cycles, the fluids were examined for wear particles according to a particle analysis. Using XLPE inserts, the backside wear was significantly reduced by 35%. While XLPE backside wear particles are significantly larger than CPE particles, they do not differ in their morphology. This study confirms the greater resistance to backside wear of XLPE compared to CPE. It can be assumed that the improved fatigue resistance of the vitamin E-stabilized XLPE inserts demonstrates XLPE's effectiveness against micro-motion and the resulting changing tensions in interface areas like surface breakdown, pitting and the release of very small particles.

**Keywords:** backside wear; cross-linked; total hip replacement; hip cup system

#### **1. Introduction**

The modification of conventional polyethylene (CPE) to cross-linked polyethylene (XLPE) and the introduction of antioxidants results in a reduction in joint articular wear [1–3].

While CPE has been the gold standard in hip joint replacement for many years, it is being increasingly replaced by XLPE at a rate of over 95% in hip arthroplasty [4,5]. There is a continuing trend towards bearings with ceramic heads and XLPE inserts [4,5]. Polymer chains that are cross-linked as a result of defined gamma or electron radiation in oxygen-free settings and subsequent thermal treatment show a higher articular wear resistance compared to CPE [6]. In addition, antioxidants such as vitamin E are used to bind free radicals in order to improve the mechanical stability, fatigue strength and oxidation resistance of XLPE [7], which results in a further reduction in articular wear [1–3]. Furthermore, XLPE with vitamin E shows an effective prevention of oxidation even with long aging and leads to more consistent wear behavior compared to CPE and XLPE [8].

Established and internationally standardized examination methods are frequently used to evaluate the articular wear resistance of joint replacements. In addition, a new investigation method enables the isolated quantitative measurement of polyethylene (PE) backside wear in modular acetabular cups [9]. While the resistance to articular wear processes of XLPE has already been demonstrated in

many experimental [1,10–13] and clinical studies [3,14,15], no quantitative evidence exists about the supposed wear advantage of XLPE in terms of backside wear.

Therefore, the aim of this study was to compare the backside wear behavior of CPE and XLPE inserts. The questions were defined as follows: Is less PE backside wear generated by using XLPE than using CPE? Is there a difference in the size and morphology of the backside wear particles of XLPE compared to those of CPE?

#### **2. Materials and Methods**

#### *2.1. Experimental Groups*

Two test groups were compared to each other (CPE versus XLPE). Each group included three cups and three inserts, which were tested independently of each other under the same test conditions.

#### *2.2. Analyzed Components*

The cementless cup system Plasmafit® Plus 7 (Aesculap, Tuttlingen, Germany, Ref: NV352T) with a cup size of 52 mm was used. The locking mechanism for fixing the PE inserts was based on a conical width and rough striking surface (Ra = 3.7 μm, Rz = 24.7 μm). PE inserts with an inner diameter of 32 mm were used for all investigations. For the first group (CPE), conventional PE (Ref: NV203) in accordance with ISO 5834-2 was used. Cross-linked PE Vitelene® (Aesculap, Tuttlingen, Germany, Ref: NV203E) with vitamin E as the antioxidant was used for the second group (XLPE).

#### *2.3. Test Setup*

The method used for the quantitative measurement of PE backside wear has been validated and described in detail [9].

The cup system (Figure 1(5)) was fixed in polyurethane and dynamically loaded with a frequency of 3 Hz and a force of up to 2.5 kN. Simultaneously, a moment of about 5 Nm (Figure 1) was induced in the cup system. The inclination of the cup was 30◦ to the load axis (Figure 1(1)). The test duration was 2 <sup>×</sup> <sup>10</sup><sup>6</sup> cycles.

**Figure 1.** Test setup according to the two-chamber principle: (**1**) load axis, (**2**) artificial femoral head, (**3**) backside wear particle, (**4**) articulation area, (**5**) cup system, (**6**) PE insert, (**7**) backside area [9].

According to the two-chamber principle, the backside area (Figure 1(7)) was separated from the articulation area (Figure 1(4)) and each area represented a chamber system. Both chambers were filled with ultrapure water as the test fluid. However, the interface between the PE insert and the cup could not be considered a reliable seal between the articulation area and the backside cup area. To ensure an isolated generation of backside wear particles, the articulation of the sliding partners (head and PE insert) and thus the articular wear must be prevented. Due to the rigid and resistant cohesive connection (Loctite® 406/SF770, Henkel AG & Co. KGaA, Duesseldorf, Germany) (Figure 1) between the articulation surfaces (femoral head and insert (Figure 1(2,6)), an isolated generation of insert backside wear (Figure 1(3)) was provided. Thereby, backside-generated PE particles were collected in the test fluid. The integrity of the rigid connection was checked and confirmed before and after each test.

#### *2.4. Wear Analysis*

After the test was completed, the test fluid was removed from the chambers and analyzed for wear particles. For this analysis, the test fluid was vacuum filtered (pore size: 0.02 μm) and the filters were examined by a scanning electron microscope (SEM Leo 1530, Carl Zeiss AG, Oberkochen, Germany) at a 20,000 times magnification (3 filters/3 SEM images per chamber). The SEM images created were then analyzed using image processing software (ImageJ, version 1.48, public domain) and a particle analysis was carried out in accordance with ASTM F1877-16 [16]. The wear particles were characterized in terms of their size (ECD) and morphology (aspect ratio (AR) and roundness (R)). The particle shapes were characterized as round, oval and fibril-like according to their AR [17]. In addition, the number of analyzed particles was extrapolated to obtain a total number of particles (ETN) as a measure of PE wear [9].

#### *2.5. Statistics*

Descriptive statistics with the mean and standard deviations of nine individual values were given for all results. An independent t-test was used to compare all the mean parameters (ETN, ECD, AR and R) of the particle analysis of the two groups. All the requirements for the implementation of statistical procedures were confirmed. The software SPSS (Version 22, IBM, Amonk, NY, USA) was used for the statistical evaluation. The level of significance was set at 5% (*p* < 0.05).

#### **3. Results**

Figure 2 shows an example of isolated backside PE wear particles. Figure 3 shows the mean and standard deviations of the PE backside wear in a direct comparison between the use of CPE and XLPE in the investigated cup system. The results of the CPE backside wear with the parameters ETN, ECD, AR and R were partially published in a previous study [9]. In addition, the amounts of different particle shapes and the median (max and min) are shown in Figure 3.

**Figure 2.** Filter images of the analyzed backside PE wear particles.

**Figure 3.** Results of the particle analysis and comparison between the conventional polyethylene (CPE) and cross-linked polyethylene (XLPE) backside wear.

Using the XLPE, about 35% less PE backside wear was generated compared to when the CPE was used. This difference was statistically significant (ETN: *p* < 0.001).

The morphology parameters AR and R did not differ significantly between the CPE and XLPE (AR: *p* = 0.465 and R: *p* = 0.126). The particles were predominantly characterized by a round morphology. The XLPE tended to have a larger amount of round particles than the CPE. In addition, the XLPE particles were slightly but significantly larger (*p* = 0.028). The PE backside wear particles were generally nanoparticles in a size range of between 40 and 100 nm (Figure 4).

**Figure 4.** The size distribution of the analyzed particles from the CPE and XLPE inserts.

Figure 5 shows an example of the backside of the CPE and XLPE inserts. It seems that micro-scratches are much more pronounced on the CPE inserts than on the XLPE inserts, whereas the XLPE inserts show a higher proportion of pitting. However, the damage to the XLPE inserts is generally less pronounced.

**Figure 5.** Photographs (Digital Microscope VHX-5000, Keyence, Japan) of the backside of the tested PE inserts.

#### **4. Discussion**

The use of XLPE generated significantly less PE backside wear than the use of CPE for the investigated cup system. With the CPE and XLPE inserts, predominantly round particles (56–62%) were generated on the backside and the XLPE particles were significant larger.

The particle analysis of the reported articular wear showed comparable results regarding the morphology [10,12,18]. Illgen et al. described predominantly round particles from CPE (amount: 78%) and from XLPE (amount: 87%) [10]. The amount of elongated fibrils was significantly higher for CPE (14%) than for XLPE (6%). In addition, the CPE particles in published articular wear studies (approx. 196–710 nm) are significantly larger than the XLPE particles (approx. 110–260 nm) [10,12,13]. There was a significant difference in the size of backside-generated particles in this study. The difference in size between the CPE and XLPE particles was only around 3 nm. Considering the size distribution (about 40–100 nm) of the analyzed PE particles, the difference seems negligibly small. However, this 3 nm was a difference of 4%. Fewer wear particles that are of a larger size could result in a higher volume of wear. However, the significantly higher amount of wear particles generated by the CPE inserts (by 35%) seems to have been more relevant than the 4% larger size of the XLPE particles. In addition, we only have two-dimensional images of the particles, which makes it impossible to determine the volume of wear.

The reason for the comparable particle shape was probably the rough locking surface of the cup. The rough peaks of the titanium locking surface engaged the soft PE insert. Due to the micro-motion at this interface, the rough peaks plowed grooves into the soft surface layer of the PE. Therefore, the resulting wear mechanism was dominated by scratching and micro-machining and thus produced comparable wear particles in terms of size and shape. This could be confirmed by the similar type of damage on the backside of the CPE and XLPE inserts. In addition, the permanent alternating stress could cause material fatigue in the PE interface areas and could lead to the release of very small particles.

The higher articular wear resistance of XLPE has not only already been demonstrated in some experimental studies [1,10–13], but also in clinical trials [3,14,15,19,20], which examined the primary wear process. In vitro wear simulating showed a great decrease in wear rate, significantly higher fatigue resistance and improved mechanical properties for XLPE compared to CPE inserts [1,8,11]. In this study, the XLPE also showed greater resistance to backside wear. An explanation for this might be that the improved fatigue resistance of the vitamin E-stabilized inserts was effective against the occurring micro-motions and resulting variations in stress in the PE interface areas (surface disruption, pitting and release of very small particles).

In addition to lower total wear, previous studies have documented a significantly lower occurrence of osteolysis and lower revision rate in patients with XLPE components [3,15,21]. Hanna et al. described a survival rate of 86% for CPE and 100% for XLPE after 13 years with revisions due to excessive wear or osteolysis [15]. While osteolysis was identified in patients with CPE in up to 25% of cases, osteolysis could only be detected in about 2% of cases in patients with XLPE [3,15,21]. According to Fukui et al., patients with XLPE had no wear rate above the osteolysis threshold of 0.1 mm/year, in contrast to 35% of cases in patients with CPE [3]. Cellular responses to the wear particles of XLPE are controversial. While no cell responses [22] or significantly reduced functional biological activity [23] were detected in patients with XLPE in some studies, other studies have shown a significantly larger inflammatory response to XLPE particles [10,13,24,25] than to CPE particles. Illgen et al. showed a concentration-dependent inflammatory response of macrophages against CPE and XLPE particles [10]. At low concentrations, no significant differences in the biological inflammatory response could be observed between CPE and XLPE. However, a significantly higher response was shown to XLPE particles at higher particle concentrations than to CPE wear particles [10]. This higher inflammatory response with a high particle concentration could certainly play an important role with regards to pelvic lysis behind the cups. A particle migration within the cup system [26] could lead to a local accumulation of wear products in the area of screw holes on the bony acetabulum and thereby steadily increase the concentration of PE particles. Therefore, the migration of articulating and backside-generated PE particles should be avoided as far as possible [26].

#### **5. Limitations**

This study was an experimental investigation. Deviations from the clinical situation were unavoidable. Therefore, the following limitations must be taken into account.

Instead of bovine serum or a similar lubricant used for in vitro wear studies in accordance with ISO 14242, ultrapure water was used as the test fluid. However, the backside wear mechanism was completely different to the articulation wear mechanism. On the backside, the rough peaks of the harder inner surface of the cup plowed grooves into the surface of the softer PE. Therefore, the influence of the lubricant on the backside wear mechanism was assumed to be negligible. In addition, ultrapure water proved to be advantageous due to its simpler handling and lower risk of contamination by proteins or other residues of biological substances. On the basis of the structure of the artificial cup system, a reliable hermetic sealing of the articulation area from the backside of the PE insert could hardly be achieved. Therefore, articulation between the head and PE insert was prevented (due to the rigid connection). Subsequently, the usual articulation between the sliding bearings was not possible and the applied force represented a simplified load condition compared to physiological hip loads. However, together with the introduction of moments, the load situation approximated the clinical situation in the cranial–caudal axis [27].

#### **6. Conclusions**

XLPE showed a significantly higher backside wear resistance compared to CPE. It can be assumed that the improved fatigue resistance of the vitamin E-stabilized XLPE inserts demonstrated XLPE's effectiveness against the occurring micro-motion and resulting changing tensions in and below the interface areas like surface breakdown, pitting and the release of very small particles. The detected backside wear particles were smaller than the reported particles of articular wear, while the morphology of the backside-generated CPE particles did not differ from those of the XLPE.

**Author Contributions:** Conceptualization, S.B. and J.P.K.; methodology, S.B.; formal analysis, S.B. and S.J.; investigation, S.B., R.S. and S.S.; resources, J.P.K.; data curation, S.B. and S.J.; writing—Original draft preparation, S.B.; review and editing, S.J., R.S., S.S. and J.P.K.; visualization, S.B.; supervision, S.B. and J.P.K.; project administration, J.P.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Aesculap AG, Tuttlingen, Germany who provided the implant components.

**Acknowledgments:** The authors would like to thank Aesculap, Tuttlingen, Germany for the study support.

**Conflicts of Interest:** The authors declare no conflict of interest in relation to this study. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Comparison of the Primary Stability of Porous Tantalum and Titanium Acetabular Revision Constructs**

**Nicholas A. Beckmann 1,2,\*, Rudi G. Bitsch 3, Mareike Schonho**ff **4, Klaus-Arno Siebenrock 2, Martin Schwarze <sup>1</sup> and Sebastian Jaeger <sup>4</sup>**


Received: 29 February 2020; Accepted: 7 April 2020; Published: 10 April 2020

**Abstract:** Adequate primary stability of the acetabular revision construct is necessary for long-term implant survival. The difference in primary stability between tantalum and titanium components is unclear. Six composite hemipelvises with an acetabular defect were implanted with a tantalum augment and cup, using cement fixation between cup and augment. Relative motion was measured at cup/bone, cup/augment and bone/augment interfaces at three load levels; the results were compared to the relative motion measured at the same interfaces of a titanium cup/augment construct of identical dimensions, also implanted into composite bone. The implants showed little relative motion at all load levels between the augment and cup. At the bone/augment and bone/cup interfaces the titanium implants showed less relative motion than tantalum at 30% load (*p* < 0.001), but more relative motion at 50% (*p* = n.s.) and 100% (*p* < 0001) load. The load did not have a significant effect at the augment/cup interface (*p* = 0.086); it did have a significant effect on relative motion of both implant materials at bone/cup and bone/augment interfaces (*p* < 0.001). All interfaces of both constructs displayed relative motion that should permit osseointegration. Tantalum, however, may provide a greater degree of primary stability at higher loads than titanium. The clinical implication is yet to be seen

**Keywords:** porous implants; tantalum; titanium; acetabulum; hip arthroplasty; hip replacement; revision hip arthroplasty; acetabular revision; primary stability

#### **1. Introduction**

Total hip arthroplasty (THA) is a highly successful surgical intervention that is being performed with increasing frequency in cases of advanced osteoarthritis, and in patients of decreasing age [1]. The increased incidence of primary THA is accompanied by a corresponding increase in revision THA with the associated concerns of diminished bone quality, bone loss and compromised soft tissue [2]. Earlier interventions that addressed these concerns included the use of large structural allografts that had mixed results with loosening and migration rates of up to 70% [3]. Utilization of metal cages for large defects reduced the loosening rate to 14% at 6 year follow up [3,4]. In addition, the recognition of cement disease as a major cause of loosening and later failure in cemented constructs [5] led to the increasing use of cementless porous metal components that allowed for bone ingrowth that facilitated

stable fixation. The introduction of porous metal implants with a range of accessory porous metal augments, buttresses and shims has led to a further improvement in revision THA outcome.

Currently, the most frequently used porous metal implants have either a tantalum or titanium porous metal surface and with press-fit implantation they provide a stable mechanical surface between implant and host bone in the short term (primary stability), and facilitate osseointegration in the mid and long term [6,7]. Tantalum in the form of Trabecular Metal™ (TM) (Zimmer Biomet, Warsaw, Indiana) is currently one of the more frequently used porous implants [8], and has been used to treat very extensive acetabular defects [9] as well as for neoplastic periacetabular lesions [10].

Optimal primary stability and ultimately adequate osseointegration and successful outcome is dependent on minimal relative motion at the component/bone interface. Prior experimental studies have shown that successful osseointegration occurs with relative motion between surfaces of up to 40μm, and that fibrous attachment occurs at 150 μm [11]. In addition, increased relative motion between components and bone can lead to particle generation and shedding that promotes later loosening and failure [12]. The use of additional components such as augments and buttresses increases the number of opposed surfaces and also the potential for increased relative motion in the construct as a whole, with possible consequences for the stability of the construct [13].

The aim of this study was to utilize an experimental biomechanical set up to evaluate the primary stability of a tantalum acetabular cup and augment construct as used in the treatment of larger acetabular defects [14], and compare the results with those of a similar porous titanium construct and augment that were published previously in an identical experimental set up [13].

#### **2. Materials and Methods**

Six Trabecular Metal™ acetabular cup components of 56 mm diameter and corresponding Trabecular Metal™ augments of 54/56 × 1 cm size were utilized in our biomechanical set up (see Figure 1). We also utilized six large fourth generation composite left hemipelvises (#3405 Sawbones; Sawbones Europe AB, Malmo, Sweden) each with a created Paprosky 2b defect of 1 cm thickness that was segmental and constituted less than one third of the acetabular circumference. Each defect was created in an identical standardized manner at the postero-cranial aspect of each acetabulum, with the edge of the defect adjacent to the antero-inferior iliac spine. To accomplish this, the periphery of the defect was first marked on each hemipelvis and the central synthetic bone was reamed and burred to the peripheral mark and to 1cm depth. This created defect was then completely covered with a TM augment according to manufacturer's instructions, and fixed to host bone with two 5.5 × 30 mm screws. Prior to acetabular cup implantation, premixed cement was then applied to the aspect of the augment that apposed the acetabular cup. A medium viscosity bone cement (Palacos R + G pro; Heraeus Medical Gmbh, Wehrheim, Germany) was used at this interface. The cement was vacuum mixed (Optivac Cement Mixing System; Zimmer Biomet, Warsaw, Indiana) and applied 120 s after the start of mixing. The cement was dispensed with a cement gun and 1.5 cm<sup>3</sup> was hand-modeled on the augment surface. Excess cement was carefully removed from the multi-hole acetabular component. All cementing was done under standardized conditions with the same mean room temperature and humidity as in the prior experiment [13]. The acetabular cup component was then implanted according to manufacturer's directions, and attached to host composite bone with one each of 6.5 × 40 mm and 6.5 × 30 mm screws, with the screws directed towards the sacro-iliac joint. The acetabular cup was press-fit as well as the rim defect would allow. All augments and cups were implanted by a single experienced surgeon (R.G.B.).

**Figure 1.** Photograph of titanium Gription cup and augment (**left**) and tantalum Trabecular Metal augment and cup (**right**) after implant explantation, demontrating the differences in their hole positions and augment geometry.

Following implantation, the hemipelvises were secured along the sacral side of the ilium using polyurethane foam (RenCast FC 53 A/B; Goessl + Pfaff Gmbh, Karlskron, Germany) in a containment device [13]. The symphysis was also secured to a two-component casting resin block that had an attached stainless-steel ball on the under-side that was placed on a metal plate. This constituted a two-point pelvic fixation, with the pelvis fixed in only one degree of freedom to allow for multi-planar movement and rotation of the symphysis that mimics a physiologic fixation, as described in prior studies [13].

Optical markers of 0.8 mm diameter (uncoded passive white markers, GOM Item Number 21874; GOM Gmbh, Braunschweig, Germany) were placed in adjacent rows along the rims of the acetabular cup component, adjacent augment and the host bone [13]. These adjacent rows of markers were detected in grey-scale by a stereo camera system that provided 3D discrimination and recording of relative motion between components and bone during loading. This was achieved by using 3D point triangulation to calculate the 3D marker position in the x, y and z axes of the defined coordinate system [15]. The 3D relative motion in the x, y and z axes were measured simultaneously between the acetabular component and bone, acetabular component and augment and augment and bone using an optical measuring system (PONTOS, GOM Gmbh, Braunschweig, Germany).

We pre-tested the hemipelvis set-up using a materials testing machine (MTS Mini Bionix 359; MTS Systems Corporation, Eden Prairie, Minnesota), with the load applied in the direction of the greatest load that occurs during the normal gait cycle, as defined by Bergmann et al. [16]. The maximum load during normal walking was found to be 233% of the individual's body weight at 31 degrees of rotation around the x axis and 5 degrees around the z axis relative to the acetabular component system described by Bergmann et al. [17]. We arbitrarily chose a body weight of 80 kg for each specimen, as in our prior study [13], that was equivalent to 1.8 kN at 100% load at the hip during normal gait.

Three load levels were chosen; 3–30% load (equal to 0.5 kN), 5–50% load (equal to 0.9 kN) and 10–100% load (equal to 1.8 kN) (see Figure 2). A total of 1000 cycles were applied sequentially in a sinusoidal wave-form at 1 Hz at each of the three load levels. To ensure good force closure between force plate and testing sample, 0.2 kN was applied prior to testing. The dependent variable (measured in μm, with average and variance) was the relative motion between components and bone, measured at the following groups of cycles; 1 to 50, 51 to 200, 201–500, 501–800, 801–995.

The results of measurements obtained as described above were compared to results obtained in an identical manner during a prior experimental set-up using Gription® titanium components instead of tantalum components of the same diameter/size [13].

**Figure 2.** Schematic graph displaying the load applied for each sample over the 3000 test cycles.

#### **3. Statistical Analysis**

Statistical evaluation was carried out descriptively (arithmetic mean, standard deviation, minimum and maximum). After confirmation of normal distribution using a Shapiro–Wilk test, a t-test of independent variables was performed. To evaluate differences in both groups during the cyclic loading, we performed an analysis of variance with repeated measures (ANOVA). The effects with regard to implant type and time points were evaluated. A *p*-value of ≤0.05 was considered significant. Results were presented using statistical graphics when necessary. Statistical evaluation was performed using Microsoft Excel (Microsoft Corporation, Santa Rosa, CA, USA), and the analytical software SPSS 25 (IBM Inc., Armonk, New York, USA).

#### **4. Results**

One of the six samples was excluded, since the fixation of the hemipelvis in the containment device failed. Table 1 shows the average relative motion between the tantalum augment/cup, tantalum augment/bone and tantalum cup/bone interfaces at 30%, 50% and 100% load for the remaining samples and compares it to the average relative motion between titanium cups/augment, titanium augment/bone and titanium cup/bone interfaces.

The t-test revealed a statistically significant difference in the relative motion between titanium augment/cup and tantalum augment/cup at all load levels (30% load: t(8) = −20.34, *p* < 0.001; 50% load: t(8) = −30.06, *p* < 0.001; 100% load: t(8) = −14.32, *p* < 0.001) (see Figure 3).

The titanium augment/sawbone interface displayed less relative motion at 30% load than the tantalum augment/sawbone interface (30% load: t(8) = −8.81, *p* < 0.001). At 50% (t(8) = 1.59, *p* = 0.151) and 100% (t(8) = 15.47, *p* < 0.001) load there was an increased average relative motion of the titanium augment/Sawbone interface when compared to the relative motion at the tantalum/Sawbone interface (see Figure 4).

At 30% load, the titanium displayed significantly lower relative motion (t(8) = −13.00, *p* < 0.001) at the bone/cup interface, while at 50% load (t(8) = −0.20, *p* = 0.843) and at 100% load (t(8) = 11.76, *p* < 0.001) the tantalum displayed lower relative motion (see Figure 5).

No significant difference was noted at the augment/cup interface with regard to the load level (F(2, 16) = 2.87, *p* = 0.086). The load level did, however, have a significant effect on the relative motion at the bone/augment (F(2, 16) = 352.66, *p* < 0.001) and bone/cup (F(2, 16) = 331.96, *p* < 0.001) interfaces.


**Table 1.** Table showing the mean and standard deviation (SD) of the relative motion (μm) of tantalum (Trabecular Metal) and titanium (Gription) implants at the respective implant/bone interfaces and load levels.

**Figure 3.** Graph displaying the average relative motion (μm) at the tantalum and titanium augment and cup interfaces at the three tested load levels (30%, 50% and 100% load).

**Figure 4.** Graph shows the average relative motion (μm) between the tantalum and titanium augment and adjacent composite bone at the three tested load levels (30%, 50% and 100% load).

**Figure 5.** Graph showing the average relative motion (μm) between the tantalum and titanium cup and composite bone at the three tested load levels (30%, 50% and 100% load).

#### **5. Discussion**

The incidence of revision total hip arthroplasty (RTHA) is continuing to increase, particularly in younger patients [1,18] and is predicted to increase to 14.5% of all THAs and to increase by 174% from 2005 to 2030 [1].

Aseptic loosening has been reported to be the major reason for THA revision [2,18], and the frequently associated osteolytic defects that result from particulate debris and component wear can present a significant surgical challenge [2,19,20]. Revision THA consequently has a greater incidence of failure than primary THA because of the compromised soft tissue, bone loss and increased complexity of the procedure. This has prompted the ongoing search for improved components. Cemented acetabular cups allow only bone ongrowth rather than bone ingrowth [21,22] and have been associated with poor integration into the sclerotic host bone, increased rates of bone resorption and increased difficulty with later revision procedures [23]. Porous coated uncemented acetabular implants depend upon press-fit implantation to provide adequate primary stability during the surgical and early postoperative phase and secondary stability from later adequate osseointegration [22].

Currently tantalum and titanium are the most frequently used metals in uncemented porous components due to their biologically inert nature and their physical properties that are close to those of cancellous bone [24–27]. In addition, a recent study by Brüggemann et al. has shown little systemic response to tantalum implants, underscoring their safety in joint replacement procedures [28]. A large body of literature documents the success of porous trabecular tantalum constructs in RTHA [4,29,30]. In contrast, there is a relatively small body of literature documenting outcome with titanium constructs that vary in type and physical properties as a result of differing manufacturing processes [6,30].

The goal of our study was the evaluation of relative motion occurring at all interfaces of an implanted TM acetabular/augment construct and comparison of the results with the previously recorded relative motion occurring at the same interfaces of a porous titanium (Gription®) acetabular cup/augment construct, implanted under identical technical and environmental conditions. The inclusion of an augment in the construct added an additional interface (cup/augment) with the potential for additional relative motion. In all our tantalum and titanium constructs we found minimal relative motion at the cup/augment interface at all load levels, and we therefore interpreted this interface as having no significant negative impact on the stability of the construct as a whole. The tantalum and titanium constructs also displayed minimal relative motion of 30–50 μm at the bone/cup and bone/augment interfaces at the 30% and 50% load level. At the 100% load only, the Gription constructs displayed increased relative motion of the bone/cup and bone/augment interfaces of 107 and 84 μm, respectively. This may be due to the properties of the materials and implants, and their respective elastic modulus. Differences in the coefficient of friction alone have been shown in

a prior study to have little impact on the primary stability of the acetabular component [31]. In all instances these values are below the previously recorded levels of relative motion that are thought to result in fibrous attachment. Prior in-vivo animal studies and studies on human autopsy bones have shown that successful osseointegration occurs with up to 40 μm relative motion between implant and bone, and fibrous attachment occurs with 150 μm relative motion [11,32]. It has also been shown that successful osseointegration can occur with bony attachment that involves substantially less than 100% of the bone/implant interface [33,34] and most of the osseointegration occurs around the acetabular rim, and decreases towards the pole [34,35]. One study by Bondarenko et al. showed that osteoporotic bone has worse osseointegration than healthy bone, and also that the implant can have a significant effect on the osseointegration, or bone-implant-contact [36]. In their study the tantalum implant Trabecular Metal® and the Trabecular Titanium® showed better osseointegration than the titanium implants Stiktite®, titanium with Gription® coating or Tritanium® [36].

The minimal levels of relative motion between porous implant and bone promote successful osseointegration, secondary stability and good surgical outcome, as documented in clinical reports. In RTHA in particular, tantalum components have been reported to have excellent results in complex cases, even with large bone deficiencies [21,37,38]. The ancillary use of porous tantalum augments as buttresses in cases of insufficient acetabular rim support has also been reported to have superior results [2,39,40]. Konan et al. reported a 96% survivorship of the TM acetabular component and good functional outcome at a mean 11 year follow up in patients with Paprosky 11 and 111 defects [41]. Morselized allograft was used in most cases, and no augments. Survivorships of 10 years for tantalum cup/augment constructs have been reported by a number of authors as 91–97% [20,27,37]. There are very few studies of porous titanium components in RTHA, and also little research of the titanium Gription cup/augment constructs in RTHA. One study evaluating 146 Pinnacle Gription cups, 1 of which was used in combination with an augment, showed good short-term results after RTHA [42]. In addition, studies with Gription augments used with various other cup types have shown good functional results [2]. However, there have been recent reports of studies that used other titanium components of different composition and manufacture, such as Trabecular Titanium and Tritanium® [43]. Hosny in 2018 reported a 98.4% aseptic acetabular cup survivorship at mean follow up of 87.6 months using Tritanium® revision cups in 62 patients with Paprosky 1–3 defects [6]. No augments were used. Delanois reported a 97% aseptic acetabular cup survivorship in 35 patients with a mean 6 year follow up, also using Tritanium® cups [23].

Our study has several limitations. Although the tantalum and titanium set-ups were done under identical technical and environmental conditions with implantations performed by a single surgeon in all cases (RGB), small differences in implantation technique cannot be ruled out.

Results for the Gription® samples cannot be extrapolated to other types of titanium implants in biomechanical experiments or in the clinical scenario. There are currently several different titanium product lines that differ in composition, architecture and manufacture, with differing biophysical properties.

We chose to use synthetic composite bone (Sawbone®) rather than cadaveric bone because of the uniformity in composition that is particularly important when working with a small sample, although the biomechanical properties are not identical to bone. Our results therefore may not reflect the clinical scenario.

Increasing loads were applied up to 100% of average normal body weight (80 kg), that was our best estimate of the limited weight bearing experienced during the postoperative period. The maximum load experienced during normal walking conditions is 233% of body weight [16,17]. Joint loading was applied only in the direction of maximal load as defined by Bergmann, and did not reflect the cyclical pattern of loading during normal walking conditions.

#### **6. Conclusions**

The samples in our study showed minimal relative motion that should promote successful osseointegration. The Gription construct showed more relative motion than TM at the cup/bone and augment/bone interfaces at 100% load only, that was below the value thought to promote fibrous attachment. Relative motion at the cup/augment interface of both TM and Gription constructs was of a degree that should not negatively impact the stability of the construct as a whole. Our biomechanical results are consistent with the positive clinical experience with TM components. There are too few reports on Gription constructs to make any clinical correlation, but our test results suggest that they should function satisfactorily.

**Author Contributions:** Conceptualization, N.A.B., M.S. (Martin Schwarze), M.S. (Mareike Schonhoff) and S.J.; methodology, N.A.B. and S.J.; validation, M.S. (Mareike Schonhoff), N.A.B. and S.J.; formal analysis, data curation, and investigation S.J., N.A.B., R.G.B. and M.S. (Mareike Schonhoff); writing—original draft preparation, N.A.B.; writing—review and editing, N.A.B., R.G.B., M.S. (Martin Schwarze), M.S. (Mareike Schonhoff), K.-A.S., S.J.; visualization, N.A.B. and S.J.; supervision, R.G.B., S.J., K.-A.S.; project administration, S.J.; funding acquisition, S.J.. All authors have read and agreed to the published version of the manuscript.

**Funding:** We acknowledge financial support by the Baden-Württemberg Ministry of Science, Research and the Arts and by Ruprecht-Karls-Universität Heidelberg.

**Acknowledgments:** We have no further acknowledgements.

**Conflicts of Interest:** The authors declare no pertinent conflicts of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
