DATA & RESEARCH OVERVIEW

Transcription

1 DATA & RESEARCH OVERVIEW

2 Total knee arthroplasty helps to relieve pain and restore function and mobility for arthritis pain sufferers and is widely recognized as one of the most performed and successful surgical procedures. However, there is still room for improvement. This is why the ATTUNE Knee System was developed. The ATTUNE Knee System is designed to address the unmet needs of patients, surgeons, and hospital providers around the world. Extensive research and science has gone into the design to help improve functional outcomes for patients, performance for surgeons, and efficiency for providers.

3 The Opportunity for Improvement: Patient Satisfaction Historically, 15-2% of patients have not been satisfied with their knee replacement after surgery. 1,2 While the cause of patient dissatisfaction is multi-factorial, extensive studies have highlighted how implant shape affects patient performance and satisfaction. The following is a summary of key implant related factors that can impact performance and satisfaction. Causes of Dissatisfaction: 1. Instability: Data has shown that a greater force or load on the knee after total knee arthroplasty surgery leads to increased symptoms of instability and/or pain. Patient dissatisfaction increases to 28% when patients perform activities like going up or down stairs (Image 1). 1,3 2. Anterior Knee Pain: One of the most common complaints after total knee replacement is anterior knee pain. Data has shown that up to 25% of patients have this issue after TKA. Some of the drivers for anterior knee pain include: improper femoral rotation, patella implant shape, and an overstuffed patellofemoral joint Overhang: a. Tibia Overhang on the posterior medial side of the tibial base can cause irritation of the medial collateral ligament (MCL). Overhang on the posterior lateral side can result in impingement with the popliteal tendon. b. Femoral Overhang of 3 mm is associated with an almost twofold increased risk of knee pain more severe than occasional or mild at 2 years after surgery. Data has shown 4% of men and 68% of women have at least one area of the implant with 3 mm or more overhang. 5 The remainder of this document contains a sampling of the research done on the ATTUNE Knee to address these clinically relevant challenges. A listing of all ATTUNE Knee publications is also provided at the end of the document. 6% Very Dissatisfied Dissatisfied Neutral Satisfied Very Satisfied 5% 4% 3% 2% 1% % Walking on a flat surface Going up or down stairs Sitting or Lying Ascending stairs Getting in or out of car/bus Rising from bed OVERALL PAIN FUNCTION Lying in bed Performing light domestic duties Very Dissatisfied 7.6% 4.% 4.8% 3.5% 4.4% 4.% 3.% 3.1% 2.7% Dissatisfied 4.% 3.3% 8.5% 3.% 7.8% 8.9% 3.8% 2.7% 3.8% Neutral 7.7% 7.2% 15.% 9.1% 15.% 16.9% 11.4% 1.2% 1.7% Satisfied 29.3% 33.2% 36.9% 35.5% 39.2% 39.5% 4.% 37.5% 38.6% Very Satisfied 51.3% 52.3% 34.7% 48.9% 33.6% 3.7% 41.8% 46.6% 44.2% Image 1: This chart looks at dissatisfaction rates among patients for different activities. 1 Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 3

4 Table of Contents SCIENTIFIC EVIDENCE ON THE ATTUNE KNEE SYSTEM: Addressing Stability with the ATTUNE Knee: Page 6: Page 8: Clary, C. W., Fitzpatrick, C. K., Maletsky, L. P., Rullkoetter, P. J. (212, February). Improving dynamic mid-stance stability: an experimental and finite element study. Orthopaedic Research Society, 58th Annual Meeting, Poster No. 144, San Francisco, CA. Fitzpatrick, C. K., Clary, C. W., Rullkoetter, P. J. (212, February). The influence of design on TKR mechanics during activities of daily living. Orthopaedic Research Society, 58th Annual Meeting, Poster No. 234, San Francisco, CA. Page 1: Clary, C. W., Wyss, J. G., Wright, A. P., Bennet, T. D., Auger, D. D., Heldreth, M. A. (212, February). Management of MCL tension in deep flexion: influence of implant design. Orthopaedic Research Society, 58th Annual Meeting; Poster No. 199, San Francisco, CA. Addressing Anterior Knee Pain with the ATTUNE Knee: Page 13: Fitzpatrick, C. K., Clary, C. W., Cyr, A. J., Maletsky, L. P., Rullkoetter, P. J. (213, April). Mechanics of post-cam engagement during simulated dynamic activity. Journal of Orthopaedic Research, 31(9), Page 16: Clary, C.W., Wright, A.P., Komosa, M.C., & Maletsky, L.P. (212). The effect of patella medicalization on patellofemoral kinematics after total knee replacement. 18th European Society of Biomechanics, Presentation number 1262, Lisbon, Portugal. Page 18: Abo-Alhol, T. R., Fitzpatrick, C. K., Clary, C. W., Cyr, A. J., Maletsky, L. P., Laz, P. J., & Rullkoetter, P. J. (213). Patellar mechanics during simulated kneeling in the natural and implanted knee. Journal of Biomechanics, 4, 2-7. Addressing Overhang with the ATTUNE Knee: Page 2: Clary, C. W., Deffenbaugh, D., Leszko, F., Courtis, P. (213, October). Tibial tray design factors affecting tibial coverage after total knee arthroplasty. International Society of Technology in Arthroplasty 26th Annual Congress, Palm Beach, FL. Page 22: Clary, C. W., Fitzpatrick, C.K., Maletsky, L. P., Rullkoetter, P. J. (212, February). Probabilistic modeling of femoral component overhang. Orthopaedic Research Society, 58th Annual Meeting, Poster No. 1988, San Francisco, CA. Additional Publications: Page 24: Full list of all ATTUNE Knee scientific publications 4 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

5 Scientific Evidence on the ATTUNE Knee System: ADDRESSING STABILITY WITH THE ATTUNE KNEE Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 5

6 Addressing Stability with the ATTUNE Knee IMPROVING DYNAMIC MID-STANCE STABILITY: AN EXPERIMENTAL AND FINITE ELEMENT STUDY CONFERENCE MEETING: Orthopaedic Research Society, 58th Annual Meeting, Poster No. 144, February 212, San Francisco, CA AUTHORS: Clary CW, Fitzpatrick CK, Maletsky LP, Rullkoetter PJ Introduction Fluoroscopic kinematic evaluation of total knee arthroplasty (TKA) has shown a sudden anterior shift of the tibiofemoral contact point, frequently of the medial femoral condyle. 1 It has been suggested this motion is tied to abrupt changes in the femoral sagittal radius of curvature (J-Curve) typical of traditional TKA. To evaluate the link between detailed implant geometry and joint mechanics, an experimental or computational model that effectively demonstrates the in vivo behavior is a necessity. The purpose of the current study was to utilize a previously validated computational model of the Kansas knee simulator (KKS) 2 to understand the influence of TKA geometry on the resulting joint mechanics and then as an iterative design-phase tool to develop implant geometry which improves dynamic mid-stance stability. To verify the predictions of the computational model, the new geometry was compared to an existing TKA in a cadaveric study utilizing the experimental simulator. This comparison enabled assessment of the accuracy of the computational model and illustrated whether the simulations were sensitive enough to appropriately differentiate subtle changes in implant design and the resulting kinematic patterns. Methods A previously validated specimen-specific finite element model of a cadaveric knee in the KKS, 2 including specimen-specific bony geometry and soft-tissue representations, was implanted with multiple prototype implant geometries. Design iterations implicated an abrupt change from the sagittal femoral distal radius to the posterior radius as responsible for the anterior slide seen in vivo. Based on this understanding, a gradually reducing sagittal femoral radius was developed and incorporated into the updated femoral design (Fig. 1). Six cadaveric knees were implanted with a traditional multi-radius TKA design and mounted into the KKS. 3 A simulated deep knee bend (DKB) was performed on the knees between 1 and 1 flexion driven by a force applied to the quadriceps tendon to balance a body-weight force applied at the hip. The medial-lateral (M-L) translation and all rotations at the ankle were unconstrained. Subsequently, the traditional TKA was replaced with the updated TKA geometry, incorporating the gradually reducing sagittal femoral radius of curvature, and the cycle repeated (Fig. 2). Knee motion was measured using an Optotrak 32 (Northern Digital Inc., Waterloo, Ontario, Canada) and six-degree-of-freedom tibiofemoral kinematics described using a three-cylindrical open-chain model. 4 Additionally, the contact points between the insert and femoral component were approximated by identifying the lowest point on the femoral geometry along the superior-inferior (S-I) axis of the tibia. Kansas Knee Simulator Computational Simulator Fig. 1: The Kansas knee simulator (left) and the computational representation of the simulator (right). Traditional TKA (Multi-radius) Abrubt radius change in mid-flexion Updated TKA (Continuously Reducing Radius) Gradual radius reduction through 9 flexion Fig. 2: Comparison of femoral sagittal curvatures for the traditional multi-radius TKA (left) and the updated design with a gradually reducing radius (right). Results Both the computational and experimental simulators were able to identify key relationships between the implant shape and the contact mechanics, including the abrupt anterior slide of the femoral condyles of the traditional TKA at the transition from the distal to posterior sagittal radius of curvature (Fig. 3, left). In comparison, the gradually reducing femoral sagittal radius of curvature attenuated the anterior slide of the medial femoral condyle and led to a gradual posterior translation of the lateral condyle with knee flexion (Fig. 3, right). Although not statistically significant, the cadaveric knees on average experienced increased femoral rollback with the updated design. 6 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

7 A-P Translation (mm) Traditional TKA (Multi-radius) Updated TKA (Continuously Reducing Radius) Medial (Exp) -2 Lateral (Exp) -2 Medial (Model) -4 Lateral (Model) Medial (Exp) -1-1 Lateral (Exp) Medial (Model) Lateral (Model) Knee Flexion (º) Knee Flexion (º) Fig. 3: A-P condylar translations measured in the KKS and predicted by the computational simulator for the PFC SIGMA System (left) and ATTUNE Knee (right) Implants. Discussion In vitro experimental and computational simulations are critical preclinical tools in the evaluation of new implant designs. The combined experimental and computational approach described here was able to relate subtle design changes in the sagittal femoral radius of curvature to A-P stability during a DKB and femoral rollback in flexion. While the models were able to identify and enable a solution to a clinically observed phenomenon, current and future work is focused on improving the fidelity and validation of the computational simulations to represent more sophisticated activities of daily living like gait, navigating stairs, and rising from a chair. Significance This study utilized computational and experimental knee simulations to identify the relationship between TKA implant shape and a clinically observed kinematic phenomenon and then enabled design changes to address the paradoxical motion. Article Summary: This in vitro study demonstrated that the position where a multi-radius femoral design changes from its first radius to its second radius, the medial condyle slid forward and the lateral condyle stopped rolling back. This is not reflective of what is seen in the native nonoperated knee. The modeling data demonstrated that a gradually reducing radius on the femoral component, ATTUNE GRADIUS Curve, helped to attenuate paradoxical anterior slide when compared to a traditional multi-radius design. Computational modeling and cadaveric simulation both showed that the ATTUNE GRADIUS Curve kept the medial condyle stable during early flexion, while enabling the lateral condyle to roll back throughout flexion. This is more similar to what is seen in the native knee as compared to the multi-radius design. References: 1. Dennis D.A., Komistek R.D., Mahfouz M.R. In vivo fluoroscopic analysis of fixedbearing total knee replacements. (23). Clinical Orthopaedics and Related Research, 41, Baldwin M.A., Clary C.W., Fitzpatrick C.K., Deacy J.S., Maletsky L.P., Rullkoetter P.J. (212). Dynamic finite element knee simulation for evaluation of knee replacement mechanics. Journal of Biomechanics, 45(3), Maletsky L.P., Hillberry B.M. (25). Simulating dynamic activities using a five-axis knee simulator. Journal of Biomechanical Engineering, 127(1), Grood E.S., Suntay W.J. (1983). A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. Journal of Biomechanical Engineering, 15(2): Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 7

8 Addressing Stability with the ATTUNE Knee THE INFLUENCE OF DESIGN ON TKR MECHANICS DURING ACTIVITIES OF DAILY LIVING CONFERENCE MEETING: Orthopaedic Research Society, 58th Annual Meeting, Poster No. 234, February 212, San Francisco, CA AUTHORS: Fitzpatrick CK, Clary CW, Rullkoetter PJ Introduction Instability of the knee in total knee replacement patients has been reported during high demand activities both through clinical observations and fluoroscopic evaluation. Ploegmakers et al, 1 in an in vivo kinematic evaluation, cited implant design factors as a determinant of knee instability. The objective of the current study was to compare anterior-posterior (A-P) and internal-external (I-E) motions of the knee for four current TKR designs in order to assess the influence of implant geometry on the inherent stability, motion, and contact mechanics of the joint. Each design was assessed in two finite element (FE) models: a laxity/stability test, and a full lower limb model during two high demand activities stepdown (high A-P force) and stance-phase gait (high I-E torque). Methods Implant design was quantified in terms of the tibiofemoral (TF) conformity ratio, calculated by dividing the femoral articular radius at, 15, 3 and 6 flexion by the radius of the insert in the dwell point (Fig. 1). To assess tibiofemoral constraint, a finite element model of the femoral component was positioned in the dwell of the insert at, 15, 3 and 6 flexion under a compressive load of 667N. A 5 mm anterior translation or 1 internal rotation was applied to the femur while the TF reaction force or torque was measured. Subsequently, dynamic simulations of stepdown and stance-phase gait activities were carried out in a FE model of the lower limb (Fig. 1). TF joint loads were taken from in vivo telemetric data 2 and a control system was implemented to apply external loads at the hip and ankle to create the experimentally-measured loading condition (compressive load, A-P force, I-E torque) at the TF joint for each activity using the telemetric implant geometry. The external loading condition was subsequently applied directly in the model, and the simulation was carried out for the four components, including current cruciate-retaining (CR) and posterior-stabilizing (PS) designs from several manufacturers. 6-DOF TF kinematics, medial and lateral condyle lowest point, and contact mechanics were evaluated for each design. Results Conformity ratios correlated well with laxity/constraint of the components (r =.73 for translation tests; r =.78 for rotation tests). Trends during the dynamic activities were in agreement with those predicted during the laxity simulations; higher conformity increased constraint and hence the loads carried by the insert instead of the surrounding soft tissue. These designs with higher conformity had, in general terms, higher contact area, and lower contact pressure than the less conforming components (Fig. 2). Both laxity tests and dynamic simulations highlighted substantial variation in the constraint provided by current implant designs. The range of A-P and I-E motion for the least constrained design was twice that of the most constrained design during dynamic activity (Fig. 3). Conformity Ratio Triathlon NexGen ATTUNE SIGMA hip load Hip Compressive Load (N) TF Reaction Force (N) Ankle I-E Torque (Nm) TF Reaction Moment (Nm) Flexion Angle ( ) ankle I-E ankle flexion Ankle Flexion Force (N) Gait Stepdown Fig. 1: TF conformity ratios (left, top) and laxity test reaction forces (left, middle) and moments (left, bottom) to anterior and internal motions, respectively zero indicates post-cam -5 impingement; FE model of the lower limb (below); external loading profiles implemented in the model to apply joint compressive load, I-E -1 torque and A-P force during gait and stepdown activities (right) Cycle (%) 8 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

9 Gait TF Contact Area (mm 2 ) Triathlon Nexgen ATTUNE SIGMA Gait Stepdown Stepdown Flexion Angle ( ) Triathlon NexGen ATTUNE SIGMA Peak Contact Pressure (MPa) to instability of the knee joint, notably, ligamentous balance/ tension. Some knees, through natural mechanics or injury, have a tendency towards instability. Component designs with inherent geometric stability may aid in maintaining knee stability during dynamic activity for these patients. Significance Understanding the variation in constraint provided by differing current implant designs may aid clinicians in determining which type of implant is most appropriate, given the soft-tissue quality of their patient, to provide adequate stability during activities of daily living while maintaining range of motion. Stepdown Triathlon NexGen ATTUNE SIGMA Fig. 2: Contact area during stepdown (top left); peak contact pressure for PS components during both activities (top right); contact patch for PS components at peak external torque during gait (center) and peak posterior force during stepdown (bottom) Medial A-P Position (mm) Medial A-P Position (mm) 14 Anterior Posterior Stance-phase Gait Cycle (%) 12 Anterior Posterior Stepdown Flexion Angle ( ) 14 Anterior Posterior Stance-phase Gait Cycle (%) Fig. 3: Medial and lateral A-P kinematics for each PS (solid) and CR (dashed) component shown for gait (top) and stepdown (bottom) Lateral A-P Position (mm) Lateral A-P Position (mm) Triathlon NexGen ATTUNE SIGMA 12 Anterior Posterior Stepdown Flexion Angle ( ) Article Summary: This computational study reviewed the sagittal conformity of various implant designs and its impact on kinematics during activities like stair descent and walking. This study showed that a gradually reducing radius design, ATTUNE GRADIUS Curve, and size for size matching of the femoral component and polyethylene insert helped to attenuate paradoxical anterior sliding and provide better contact area without point loading or edge loading. This study showed there is an abrupt change in conformity that occurs with the Zimmer NexGen knee and the PFC SIGMA Knee as a result of their multiradius femoral design. This abrupt change in conformity was shown to lead to paradoxical anterior sliding and greater A/P movement during testing. The Stryker Triathlon knee in this study had a significantly lower femoral to insert conformity ratio than the other designs tested. This contributed to Triathlon having more A/P movement and paradoxical anterior sliding as compared to the other designs. Discussion In the current analysis, each component was analyzed under the same external loading conditions with the same soft-tissue representation, allowing for direct comparison between components. Varying ligament mechanics would alter the magnitude of motions, but relative performance of each implant would be consistent. Other factors, aside from geometry, contribute References 1. Ploegmakers M.J., Ginsel B., Meijerink H.J., de Rooy J.W., de Waal Malefijt M.C., Verdonschot N., Banks S.A. (21). Physical examination and in vivo kinematics in two posterior cruciate ligament retaining total knee arthroplasty designs. The Knee, 17(3), Kutzner I., Heinlein B., Graichen F., Bender A., Rohlmann A., Halder A., Beier A., Bergmann G. (21). Loading of the knee joint during activities of daily living measured in vivo in five subjects. Journal of Biomechanics, 43(11), Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 9

10 Addressing Stability with the ATTUNE Knee MANAGEMENT OF MCL TENSION IN DEEP FLEXION: INFLUENCE OF IMPLANT SHAPE CONFERENCE MEETING: Orthopaedic Research Society, 58th Annual Meeting, Poster No. 199, February 212, San Francisco, CA AUTHORS: Clary CW, Wyss JG, Wright AP, Bennett TD, Auger DD, Heldreth MA Introduction While recent focus on attaining deep flexion after total knee replacement (TKR) has focused on the effects of femoral rollback 1, the interplay between implant design, ligamentous constraint, and maximum knee flexion remains unclear. In this study, the influence of implant shape, in particular the shape of the posterior condyles and amount of femoral rollback, on Medial Collateral Ligament (MCL) elongation was assessed with a computational model of passive flexion and the results substantiated through implantation into cadaveric specimens. Methods Three prototype posterior stabilized rotating platform knee implants were formulated with different levels of mid-flexion femoral thickness, condylar height, and femoral rollback (Fig. 1). Implant 1 had an increased condyle height, decreased mid-flexion thickness, and increased rollback. Implant 2a had a neutral condyle height, neutral mid-flexion thickness, and neutral rollback. Implant 2b utilized the same femur as 2a, but with 1 mm less rollback. Implant 3 had a decreased condyle height, neutral mid-flexion thickness, and decreased rollback. Height Thickness Rollback Imp Imp 2a N N N Imp 2b N N - Imp 3 - N - Mid-Flexion Thickness Condyle Height Fig. 1: Three prototype implants with different femoral geometry. Table indicates level of condyle height, mid-flexion thickness, and rollback (+=increased, N=neutral, -=decreased). Each design was virtually implanted into a finite element (FE) model of the knee. The model included four non-linear spring elements representing the anterior, central, posterior, and distal portions of the medial collateral ligament (MCL), with the attachment site based on measured data from a cadaveric knee implanted with the prototype designs. Passive flexion boundary conditions were applied to the model, with the tibial component fixed in space while a compressive load of 5-N was applied to the femur and flexed from to 14 knee flexion with the remaining femoral degrees of freedom unconstrained. Elongations of the MCL bundles were measured during the simulation. The prototype knee implants were manufactured and implanted into 8 cadaveric knees by eight different orthopaedic surgeons. Four surgeons employed a measured resection technique while the other four employed a balanced gap technique. 2 After implantation, surgeons were blinded to the implant in the knee and asked to quantify the MCL tension by palpation as either loose, normal, or tight at 3 flexion angles (45, 9, and 12 ) and to measure the maximum passive flexion angle using a goniometer. Results According to the FE predictions, increased condylar height and increased femoral rollback led to increased elongation of the distal bundle of the MCL in deep flexion (Fig. 2). These predictions were corroborated by observations of half the surgeons (1% of the balanced gap surgeons) that implant 1 had an unacceptably tight MCL at 12 flexion (Fig. 3) and that fewer surgeons felt that implant 2a had a tighter MCL than implant 2b. On the contrary, while the model predicted that implant 1 had a more lax MCL at 45, the surgeons were not able to distinguish this difference. Despite the observation that the MCL was unacceptably tight in implant 1 in flexion, there was not a significant difference in the measured terminal passive flexion. Discussion In this study, a FE model was used to quantify changes in MCL elongation in deep flexion associated with implant shape. The predicted changes, on the order of 1% 2% of MCL elongation, were perceptible to orthopaedic surgeons performing a balanced gap technique, but not to surgeons performing a measured resection technique. While the differences in implant shape were subtle, the effect of these changes on the soft tissue function and the surgeon s perception was clear. This is the first time such a correlation between implant shape and surgical technique has been documented. Although it s been shown that femoral rollback enables deep flexion by preventing bony impingement of the insert on the posterior femoral bone 1, this data suggests that too much rollback coupled with increased condylar tip height can lead to excessive strain in the MCL in deep flexion. While the cadaveric simulation did not illustrate a measureable change in terminal flexion across designs, the surgeons almost unanimously preferred Implant 3, leading to a new knee design (ATTUNE Knee System, DePuy Synthes Joint Reconstruction*, Warsaw, IN). 11 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

11 Significance This study clarifies the effect of implant shape on soft tissue function in deep flexion and the interaction with surgical technique. Results of this work will enable implant designs and surgical technique to improve high-flexion outcomes for patients receiving TKA. MCL Length (% Initial Length) MCL Elongation with Knee Flexion Implant 1 Implant 2a (Normal Rollback) Implant 2b (Reduced Rollback) Implant 3 Anterior MCL Tight Distal MCL Tight Flexion Angle (Deg) Fig. 2: Composite MCL elongation of the longest bundle of the MCL predicted by the FE model for the three prototype designs. Article Summary: Four variations of the ATTUNE Posterior Stabilized (PS) Knee design were tested with a computational model followed by a surgeon led cadaveric based test to determine the effect of implant shape on MCL function. The surgeons were blinded and asked to select which of the four knee designs provided better MCL tension throughout the range of motion. The surgeons gave the highest satisfaction rating to design 3. This work on MCL tension helped lead to the final ATTUNE PS Knee design. Additionally, this study demonstrated that the MCL changes in tension throughout flexion. It is important for implants to work in harmony with changing MCL tension throughout the range of motion. The study also confirmed the impact that relatively small changes in implant geometry can have on function. References 1. Banks S., Bellemans J., Nozaki H., Whiteside L.A., Harman M., Hodge W.A. (23). Knee motions during maximum flexion in fixed and mobile-bearing arthroplasties. Clinical Orthopaedics and Related Research, 41: Dennis D.A., Komistek R.D., Kim R.H., Sharma A. (21). Gap balancing versus measured resection technique for total knee arthroplasty. Clinical Orthopaedics and Related Research, 468(1), % of Surgeons (All) % of Surgeons (Balanced Gap) Implant: 45 Knee Flexion 9 Knee Flexion 12 Knee Flexion 1 2a 2b 3 1 2a 2b 3 1 2a 2b 3 Fig. 3: Surgeon assessments of MCL tension for the three prototypes (red=tight, green=normal, yellow=loose) by all surgeons (top, n=8) and only balanced gap surgeons (below, n=4). Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 11

12 Scientific Evidence on the ATTUNE Knee System: ADDRESSING ANTERIOR KNEE PAIN WITH THE ATTUNE KNEE

13 Addressing Anterior Knee Pain with the ATTUNE Knee MECHANICS OF POST-CAM ENGAGEMENT DURING SIMULATED DYNAMIC ACTIVITY JOURNAL: Journal of Orthopaedic Research, 31(9), AUTHORS: Fitzpatrick CK, Clary CW, Cyr AJ, Maletsky LP, Rullkoetter PJ Posterior-stabilized (PS) total knee arthroplasty (TKA) components employ a tibial post and femoral cam mechanism to guide anteroposterior knee motion in lieu of the posterior cruciate ligament. Some PS TKA patients report a clicking sensation when the post and cam engage, while severe wear and fracture of the post; we hypothesize that these complications are associated with excessive impact velocity at engagement. We evaluated the effect of implant design on engagement dynamics of the post-cam mechanism and resulting polyethylene stresses during dynamic activity. In vitro simulation of a knee bend activity was performed for four cadaveric specimens implanted with PS TKA components. Post-cam engagement velocity and flexion angle at initial contact were determined. The experimental data were used to validate computational predictions of PS mechanics using the same loading conditions. A lower limb model was subsequently utilized to compare engagement mechanics of eight TKA designs, relating differences between implants to geometric design features. Flexion angle and post-cam velocity at engagement demonstrated considerable ranges among designs (23 89, and.5.22 mm/, respectively). Post-cam velocity was correlated (r ¼.89) with tibiofemoral condylar design features. Condylar geometry, in addition to post-cam geometry, played a significant role in minimizing engagement velocity and forces and stresses in the post. This analysis guides selection and design of PS implants that facilitate smooth post-cam engagement and reduce edge loading of the post. A Hip load B C Quadriceps translation or load Ankle I-E torque Ankle F-E torque Ankle M-L Load Fig 1: In vitro experiments of implanted TKA knees were performed in the KKS (A); computational simulations of virtually implanted TKA knees were performed in an FE model of the KKS (B) and a lower limb (C). Fig. 2: Measurements of the TKA geometries included the radius of the condyles at postcam engagement (R), the center of radius at engagement (C), and the sagittal plane distance from C to the initial point of post-cam contact. C was found by calculating the tangent (T) to the femoral articular geometry lowest point; in the sagittal plane, C was on a perpendicular to T, and a distance R from the point of tangency (PoT). Post-cam Engagement Velocity (mm/ ) Exp: Specimen 1 Exp: Specimen 1.1 Exp: Specimen 1 Exp: Specimen 1 FE Model Flexion Angle ( ) Initial Post-Cam Distance (mm) A-P Post Position (mm) Initial Post-Cam Distance A-P Post Position r=.97 r= Flexion Angle at Engagement ( ) Center of Radius to Cam Distance (mm) Initial Post-Cam Distance (mm) Center of Radius to Cam Initial Post-Cam Distance r=.77 r= Post-Cam Engagement Velocity (mm/ ) Fig. 3: Comparison between experimental and computational post-cam engagement velocity and engagement flexion angle predictions. Markers indicated the point of initial engagement. Fig. 4: Left: Correlation between flexion angle at engagement with the initial post-cam distance, and AP position of the posterior surface of the post. Right: Correlation between post-cam engagement velocity with distance from the center of the condylar radius of curvature at engagement to the point of first contact on the cam, and the initial distance between the post and the cam. Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 11

14 Post-cam Engagement Velocity (mm/ ) Nexgen Vanguard SIGMA.1 ATTUNE Triathlon.5 Scorpio Genesis II Journey Flexion Angle ( ) Medial A(+ve)-P Translation (mm) Nexgen Vanguard SIGMA ATTUNE Triathlon Scorpio Genesis II Journey Flexion Angle ( ) Lateral A(+ve)-P Translation (mm) Flexion Angle ( ) Fig. 5: Post-cam velocity for 8 TKA designs during a squat activity (markers indicate velocity at the instant of post-cam engagement). Fig. 6: AP translation of the lowest points of the femoral medial and lateral condyles for eight TKA designs (markers indicate the point of post-cam engagement). Fig. 7: Cumulative contact pressure over the squat cycle (from 1 to 11 flexion) for eight TKA designs implanted into a right knee. A(+ve)-P Post Force (N) Nexgen Vanguard SIGMA ATTUNE Triathlon Scorpio Genesis II Journey M(+ve)-L Post Force (N) S(+ve)-I Post Force (N) Flexion Angle ( ) Flexion Angle ( ) Flexion Angle ( ) Fig. 8: AP (left), ML (center) and SI (right) force on the tibial post for eight TKA designs. Force directions are described as acting on the post. 11 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

15 Article Summary: This computational study reviewed the interaction between the cam and spine of various PS knee constructs. The Zimmer NexGen PS and Biomet Vanguard PS knees were shown to have the highest femoral component velocity when the femoral engaged with the insert spine. This resulted in abrupt changes in kinematics after contact and increased force on the insert spine. The Smith & Nephew Journey knee had very early engagement of the cam and spine. This engagement occurred just after 2 degrees. The Stryker Triathlon knee and the Genesis II also had earlier engagement than most other designs at around 5 degrees. This early engagement may lead to increased cam/spine interaction during certain activities and indicates a dependence upon the cam/post mechanism to augment the lack of femoral-tibial conformity. The ATTUNE Knee was shown to have a lower femoral velocity than all but the Stryker Triathlon knee and the Smith & Nephew Journey knee, while allowing for engagement to occur after 7 degrees of flexion. An analysis of contact stresses was performed on each design s poly articulating surface and spine. The ATTUNE PS Tibial Insert showed reduced contact forces in part due its smaller post radius and broader contact area during engagement. Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 11

16 Addressing Anterior Knee Pain with the ATTUNE Knee THE EFFECT OF PATELLA MEDIALIZATION DURING TKR CONFERENCE MEETING: 18th European Society of Biomechanics Presentation No. 1262, 212, Lisbon, Portugal AUTHORS: Clary C, Wright A, Komosa M, Maletsky L Introduction Lateral patellofemoral ligament (LPFL) release to remedy patellar mal-tracking during total knee replacement (TKR) remains a common occurrence [Ballantyne, 23], potentially restricting blood supply to the patella and increasing the risk of patella fracture [Ritter, 1999]. Recent research indicates that patella component medialization potentially reduces strain in the LPFL, improving patellar tracking and reducing the risk of lateral release [Anglin, 28]. The purpose of the current study was to measure the effect of patella medialization during TKR on patellar motion. Methods Seventeen cadaveric knees were dissected and mounted into the Kansas knee simulator [Maletsky, 25]. A deep knee bend was performed with the flexion angle controlled by a force applied to the quad tendon. The medial-lateral translation and all rotations at the ankle were unconstrained. After the natural evaluation, each knee received a posterior stabilized TKR with either a centralized patella (PFC Sigma Knee, DePuy Synthes Joint Reconstruction) (n=7) or a medialized patella (ATTUNE Knee, DePuy Synthes Joint Reconstruction) (n=1). Both patellae had identical outer profiles, but the articular peak of the medialized patella was offset 2 3 mm medially. Both femoral components had equivalent trochlear angles. A point at the center of the patella implant was tracked through flexion in both the natural and implanted conditions using an Optotrak 32 (NDI, Waterloo, Canada). The linear distance from the lateral border of the patella to the lateral epicondyle was calculated to quantify strain in the LPFL. Medialized Central Natural Fig. 1: Patella motion with natural (green), central (red), and medialized (blue) patellae. Results Tracking of the natural patella started centrally and moved slightly lateral in flexion, consistent with previous descriptions [Iranpour, 29]. Following TKR, both the centralized and medialized patella groups moved medially with flexion, but the medialized patella group consistently tracked 2 3 mm lateral of the centralized patella group, resulting in a position more similar to the natural patella at 9 flexion (Fig. 1). Medialization of the patella component reduced the length of the LPFL, particularly from 15 to 35 knee flexion (Fig. 2). LPFL Length (mm) Medialized (StD) Central (StD) Natural (StD) Discussion The LPFL is the primary restraint of medial patella translation near extension [Merican, 29]. Patella medialization led to a reduction in the LPFL length through 45 flexion. These results are consistent with surgical techniques which release the LPFL to address lateral patella tracking [Strachan, 29] and suggest that medialization of the patella may reduce lateral release rates during TKR Knee Flexion ( ) Fig. 2: LPFL length of the natural and implanted knees (shaded regions indicate ± std. deviation). 11 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

17 Article Summary: This cadaveric based study at the University of Kansas reviewed patellar tracking for the native knee and patellar tracking for a total knee arthroplasty with a central domed patella and a medialized domed patella. The data showed that the ATTUNE Medialized Dome Patella tracked more similar to the native patella when compared with a central domed patella. The ATTUNE Medialized Dome Patella also reduced lateral patellofemoral ligament tension as compared with a central dome patella. The authors theorized this may help lead to a reduced rate of lateral releases. References 1. Anglin C., Brimacombe J.M., Hodgson A.J., Masri B.A., Greidanus N.V., Tonetti J., Wilson D.R. (28). Determinants of patellar tracking in total knee arthroplasty. Clinical Biomechanics, 23(7), Ballantyne A., McKinley J., Brenkel I. (23). Comparison of the lateral release rates in the press fit condylar prosthesis and the PFC Sigma prosthesis. The Knee, 1(2), Iranpour F., Merican A.M., Baena F.R., Cobb J.P., Amis A.A. (21). Patellofemoral joint kinematics: the circular path of the patella around the trochlear axis. Journal of Orthopedic Research, 28(5), Maletsky L.P., Hillberry B.M. (25). Simulating dynamic activities using a five-axis knee simulator. Journal of Biomechanical Engineering, 127(1), Merican A.M., Kondo E., Amis A.A. (29). The effect on patellofemoral joint stability of selective cutting of lateral retinacular and capsular structures. Journal of Biomechanics, 42(3), Ritter M.A., Pierce M.J., Zhou H., Meding J.B., Faris P.M., Keating E.M. (1999). Patellar complications (total knee arthroplasty). Effect of lateral release and thickness. Clinical Orthopaedics and Related Research, 367, Strachan R.K., Merican A.M., Devadasan B., Maheshwari R., Amis A.A. (29). A technique of staged lateral release to correct patellar tracking in total knee arthroplasty. Journal of Arthroplasty, 24(5), Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 11

18 Addressing Anterior Knee Pain with the ATTUNE Knee PATELLAR MECHANICS DURING SIMULATED KNEELING IN THE NATURAL AND IMPLANTED KNEE JOURNAL: Journal of Biomechanics. 214 March 21;47(5): AUTHORS: Abo-Alhol TR, Fitzpatrick CK, Clary CW, Cyr AJ, Maletsky LP, Laz PJ, Rullkoetter PJ Kneeling is required during daily living for many patients after total knee replacement (TKR), yet many patients have reported that they cannot kneel due to pain, or avoid kneeling due to discomfort, which critically impacts quality of life and perceived success of the TKR procedure. The objective of this study was to evaluate the effect of component design on patellofemoral (PF) mechanics during a kneeling activity. A computational model to predict natural and implanted PF kinematics and bone strains after kneeling was developed and kinematics were validated with experimental cadaveric studies. PF joint kinematics and patellar bone strains were compared for implants with dome, medialized dome, and anatomic components. Due to the less conforming nature of the designs, change in sagittal plane tilt as a result of kneeling at 9 knee flexion was approximately twice as large for the medialized-dome and dome implants as the natural case or anatomic implant, which may result in additional stretching of the quadriceps. All implanted cases resulted in substantial increases in bone strains compared with the natural knee, but increased strains in different regions. The anatomic patella demonstrated increased strains inferiorly, while the dome and medialized dome showed increases centrally. An understanding of the effect of implant design on patellar mechanics during kneeling may ultimately provide guidance to component designs that reduces the likelihood of knee pain and patellar fracture during kneeling. Article Summary: This in-vitro cadaveric and computational study of individuals kneeling demonstrated that the ATTUNE Medialized Anatomic Patella had kinematics that were more similar to the native knee as compared to a dome style patella. It was theorized this may be due to increased patellofemoral joint conformity with the anatomic patella versus the dome style patella. The increased conformity of the ATTUNE Medialized Anatomic Patella led to less flexion/extension tilt and reduced contact forces as compared to a central dome and the ATTUNE Medialized Dome Patella. The ATTUNE Medialized Dome Patella also exhibited reduced contact force and reduced lift off as compared to a standard central domed patella Before floor contact After floor contact 6 5 Before floor contact After floor contact Contact Pressue (MPa) Contact Area (mm 2 ) Patellar sagittal tilt ( ) Natural Anatomic Medialized Dome Dome Fig. 1: Measurement of sagittal plane patellar tilt, and representation of the typical change (reduction) in tilt as a result of kneeling (top); average change (and standard deviation) in sagittal plane tilt for natural and TKR knees as a result of kneeling (bottom). Before Kneeling After Kneeling Natural Natural Anatomic Mod-dome Dome Natural Anatomic Mod-dome Dome Anatomic Medialized Dome Dome Fig. 2: Mean and standard deviations in peak contact pressure and contact area before and after kneeling for natural and TKR conditions (top); contact pressure for a representative subject before and after kneeling (bottom). 11 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

19 Scientific Evidence on the ATTUNE Knee System: ADDRESSING OVERHANG WITH THE ATTUNE KNEE

20 Addressing Overhang with the ATTUNE Knee TIBIAL TRAY DESIGN FACTORS AFFECTING TIBIAL COVERAGE AFTER TOTAL KNEE ARTHROPLASTY CONFERENCE MEETING: International Society of Technology in Arthroplasty 26th Annual Congress, October 213, Palm Beach, United States AUTHORS: Clary CW, Deffenbaugh D, Leszko F, Courtis P Introduction Adequate coverage of the resected tibial plateau with the tibial tray is necessary to reduce the theoretical risk of tibial subsidence after primary Total Knee Arthroplasty (TKA). Maximizing tibial coverage is balanced against avoiding excessive overhang of the tray causing soft tissue irritation, and establishing proper tray alignment improving implant longevity and patella function. 1 Implant design factors, including the number of tray sizes, tray shape, and tray asymmetry influence the ability to cover the tibial plateau. 2 Furthermore, rotating platform (RP) tray designs decouple restoring proper tibial rotation from maximizing tibial coverage, which may enhance the ability to maximize coverage. The purpose of the current study was to assess the ability of five modern tray designs (Fig. 1), including symmetric, asymmetric, fixed-bearing, and RP designs, to maximize coverage of the tibial plateau across a large patient population. Methods Lower limb computed-tomography scans were collected from 14,791 TKA patients and the tibia was segmented. Virtual surgery was performed with an 8 mm tibial resection (referencing the high side) made perpendicular to the tibial mechanical axis in the frontal plane, with 3 posterior slope, and aligned transversely to the medial third of the tibial tubercle. An automated algorithm placed the largest possible tray on the plateau, optimizing the ML and AP placement (and I-E rotation for the RP tray), to minimize overhang. The largest sized tray that fit the plateau with less than 2 mm of tray overhang was identified for each of the five implant systems. The surface area of the tibial tray was divided by the area of the resected plateau and the percentage of patients with greater than 85% plateau coverage was calculated. Results The percentage of patients with greater than 85% plateau coverage across the tray designs ranged from 17.% to 61.4% (Fig. 1). The tray with the greatest number of size options (Tray 4, 1 sizes) had the best coverage among the fixed-bearing trays. The RP variant of the same tray had the best overall coverage. Tibial asymmetry did not significantly improve the overall tibial coverage across the patient distribution for both asymmetric designs. Incorporating a broader medial condyle improved fit along the posterior medial corner for Tray 2, but increased the average underhang along the posterior lateral plateau offsetting any improvement in total coverage. Discussion This analysis represents the most comprehensive assessment of tray coverage to date across a large TKA-patient population. Large variations exist in the size and shape of the proximal tibia among TKA patients. 3 Developing a tray design which provides robust coverage despite this variation remains challenging. This analysis suggests that tibial asymmetry may not robustly improve coverage. Conversely, incorporating an increased number of tray sizes and utilizing an RP implant to decouple coverage from alignment may provide the most reliable solution for maximizing coverage across the patient population. Fig. 1: Tibial plateau coverage for five current TKA tray designs. Designs 1 and 2 have an asymmetric medial plateau while designs 3 through 5 are symmetric about the M-L axis. Trays 4 and 5 have identical outer profiles, but tray 5 has a RP insert which allows flexibility in the rotational alignment of the tray on the bone. The colored border around the tray periphery indicates the average under-hang of the tray (in mm) across the patient population. Significance This study utilized computational and experimental knee simulations to identify the relationship between TKA implant shape and a clinically observed kinematic phenomenon and then enabled design changes to address the paradoxical motion. 22 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

21 Article Summary: This study of over 14, CT scans reviewed the theoretical fit of different tibial base designs based on their shape and number of sizes available. Tibial base rotation was set by aligning all trays to the medial 1/3 of the tibial tubercle. The ATTUNE Knee provided better coverage (without overhang) as compared to PFC SIGMA Knee, Zimmer Persona, and Smith & Nephew Journey. This was due to the refined posterior aspect and increased number of sizes available with the ATTUNE Knee. The ATTUNE Rotating Platform (RP) Tibial Base showed further coverage benefits as compared to the fixed bearing designs due to its ability to be free rotated for best coverage without affecting tibio-femoral kinematics. References 1. Hofmann S., Romero J., Roth-Schiffl E., Albrecht T. (23). Rotational malalignment of the components may cause chronic pain or early failure in total knee arthroplasty. Orthopade, 32(6), Wernecke G.C., Harris I.A., Houang M.T., Seeto B.G., Chen D.B., MacDessi S.J. (212). Comparison of tibial bone coverage of 6 knee prostheses: a magnetic resonance imaging study with controlled rotation. Journal of Orthopedic Surgery (Hong Kong), 2(2), Fitzpatrick C.K., FitzPatrick D.P., Auger D.D. (28). Size and shape of the resection surface geometry of the osteoarthritic knee in relation to total knee replacement design. Proceedings of the Institution of Mechanical Engineers, 222(6), Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 22

22 Addressing Overhang with the ATTUNE Knee PROBABILISTIC MODELING OF FEMORAL COMPONENT OVERHANG CONFERENCE MEETING: Orthopaedic Research Society, 58th Annual Meeting, Poster No. 1988, February 212, San Francisco, CA AUTHORS: Clary CW, Fitzpatrick CK, Maletsky LP, Rullkoetter PJ Introduction The adequacy of femoral implant fit in total knee arthroplasty (TKA), often characterized by the extent of overhang, is a matter of continuous debate. In this work we present a computational method for determining the probability of excess overhang or underhang (beyond an acceptable threshold value). Unlike most previous work which relied on simple landmark data or resection contours, we consider the entire distal femur, real implant models, and clinically accurate resection surfaces. The technique is demonstrated using a commercially available implant and a modified design intended to reduce overhang. are reported as scalar values along the implant edge adjacent to the resection surfaces of the bone, indicating the probability of excess overhang greater than a user specified tolerance value. The shape modeling and Monte-Carlo process were implemented with Arthron: a morphometric analysis interface based on the opensource Visualization Toolkit (VTK). (Fig. 2) Methodology The methodology is based upon a model which describes variation in 3D femoral geometry constructed using sample bones segmented from computed tomography images. Using a principal component analysis (PCA) based surface registration technique optimized for shape model construction, the segmented data is consolidated into a point distribution model (PDM) with a size variable (PC 1 ) and several additional parameters (i.e. PC 2, PC 3, etc.) which represent the distal femur s major modes of shape variation. (Fig. 1) Fig. 2: Femoral components are automatically sized and aligned using landmarks anchored to the bone surface; conformity is estimated using the shortest distance from the implant edge to the resected bone and mapped. Under-hanging edges are colored red, over-hanging edges blue. -3 PC PC 3 3 Fig. 1: Varying PDM control parameters PC2 and PC3, affecting anterior-posterior length and inter-condylar bone width. The PDM is used as the basis for a Monte-Carlo style simulation where thousands of sample femurs are generated and virtually implanted to assess the overhang characteristics of implant designs. Surface landmarks, anchored to the PDM, are used to automatically fit and align each implant. Overhang and underhang characteristics Results & Discussion The commercial implant and modified design were evaluated using a femoral PDM constructed from the left knees of thirty-one Caucasian and forty-five Japanese subjects. The pre-existing commercial design (DePuy Synthes Joint Reconstruction PFC SIGMA Knee System) features seven non-uniform size increments (1.5, 2, 2.5, 3, 4, 5, 6) and a fixed inter-condylar notch width. The 4N component was not included. A newly developed commercial design (DePuy Synthes Joint Reconstruction ATTUNE Knee System) consisted of ten uniform size increments, an inter-condylar notch width proportional to size, and a narrower medial-anterior flange. To compare the two designs, one thousand femurs were automatically generated from Latin hypercube sampling of the PDMs control parameters. During the simulation, the probability of excess overhang was tabulated along the resection edge of each implant size. The overhang limit k was defined to be proportional to the medial-lateral bone width ML (the distance between the lateral and medial femoral epicondyles). We specified the limit to be k = (3/7) ML [mm], representing a 3 mm overhang limit for a 7 mm wide bone. 22 DePuy Synthes Joint Reconstruction ATTUNE Knee System Data & Research Overview

23 Conclusions Results indicate a significant reduction in the incidence of excess overhang on the medial-side of the anterior flange and around the intercondylar notch with the modified implant design (Fig. 3 4). Future work will include analysis of patient variation factors and surgical factors (i.e. implant alignment) and their effects on implant conformity. Significance Recent evidence links excess overhang to poor patient prognoses including soft-tissue irritation and reduced joint mobility; 1 we present a computational method for predicting and comparing the excess overhang of femoral component designs beyond an acceptable threshold. P (o.h. > K) 1 Article Summary: 1. Computational modeling was used to assess the femoral component fit of the ATTUNE Knee and PFC SIGMA Knee on 1, different femoral variations. 2. The results showed a significant reduction in the incidence of excess overhang on the medial-side of the anterior flange and around the intercondylar notch with the ATTUNE Knee design. References 1. Mahoney O.M., Kinsey T. (21). Overhang of the femoral component in total knee arthroplasty: risk factors and clinical consequences. Journal of Bone and Joint Surgery American, 92(5), Bellemans J., Banks S., Victor J., Vandenneucker H., Moemans A. (22). Fluoroscopic analysis of the kinematics of deep flexion in total knee arthroplasty. Influence of posterior condylar offset. Journal of Bone and Joint Surgery British, 84(1), Fig. 3: Probability (p) excess implant overhang for SIGMA Knee component size increments 2 5. Yellow indicates p > 5%. P (o.h. > K) Fig. 4: Probability excess implant overhang for modified component size increments 2 9. Yellow indicates p > 5%. Data & Research Overview ATTUNE Knee System DePuy Synthes Joint Reconstruction 22