- Email: [email protected]
Total knee replacement with natural rollback
Annals of Anatomy 194 (2012) 195–199
Contents lists available at ScienceDirect
Annals of Anatomy journal homepage: www.elsevier.de/aanat
Total knee replacement with natural rollback Martin Michael Wachowski a,∗, Tim Alexander Walde a, Peter Balcarek a, Jan Philipp Schüttrumpf a, Stephan Frosch a , Caspar Stauffenberg b , Karl-Heinz Frosch a , Christoph Fiedler c , Jochen Fanghänel d , Dietmar Kubein-Meesenburg b , Hans Nägerl b a
Department of Trauma Surgery, Plastic and Reconstructive Surgery, Georg August University of Göttingen, D-37099 Göttingen, Germany Department of Orthodontics, Georg August University, D-37099 Göttingen, Germany Institute of Materials Science and Technology, Friedrich Schiller University of Jena, D-07743 Jena, Germany d Department of Orthodontics, Regensburg University Medical Center, D-93053 Regensburg, Germany b c
a r t i c l e
i n f o
Article history: Received 20 December 2010 Accepted 24 January 2011
Keywords: Tibiofemoral joint Total knee replacement TKR Rollback Aequos Patella tendon angle
s u m m a r y A novel class of total knee replacement (AEQUOS G1) is introduced which features a unique design of the articular surfaces. Based on the anatomy of the human knee and differing from all other prostheses, the lateral tibial “plateau” is convexly curved and the lateral femoral condyle is posteriorly shifted in relation to the medial femoral condyle. Under compressive forces the configuration of the articular surfaces of human knees constrains the relative motion of femur and tibia in flexion/extension. This constrained motion is equivalent to that of a four-bar linkage, the virtual 4 pivots of which are given by the centres of curvature of the articulating surfaces. The dimensions of the four-bar linkage were optimized to the effect that constrained motion of the total knee replacement (TKR) follows the flexional motion of the human knee in close approximation, particularly during gait. In pilot studies lateral X-ray pictures have demonstrated that AEQUOS G1 can feature the natural rollback in vivo. Rollback relieves the load of the patello–femoral joint and minimizes retropatellar pressure. This mechanism should reduce the prevalence of anterior knee pain. The articulating surfaces roll predominantly in the stance phase. Consequently sliding friction is replaced by the lesser rolling friction under load. Producing rollback should minimize material wear due to friction and maximize the lifetime of the prosthesis. To definitely confirm these theses one has to wait for the long term results. © 2011 Elsevier GmbH. All rights reserved.
1. Introduction In diverse publications we have reported on a novel Total Knee Replacement (AEQUOS G1 TKR) which shows an unique design of articulating surfaces (Frosch et al., 2009a, 2009b; Nägerl et al., 2008): The lateral tibial articulating surface is convexly curved in anterior/posterior direction and the medial femoral condyle is slightly displaced to anterior (Fig. 1). By this asymmetric design it was intended to implement an artificial implant that sustains essential biomechanical functions of the natural knee. Nonetheless AEQUOS G1 TKR possesses the conventional material pairing for the articulating surfaces (Fig. 2): cobalt chrome (femur) versus polyethylene (tibia).
∗ Corresponding author at: Department of Trauma Surgery, Plastic and Reconstructive Surgery, Georg August University, Robert-Koch-Strasse 40, D-37075 Göttingen, Germany. Tel.: +49 551396114; fax: +49 551398981. E-mail address: [email protected] (M.M. Wachowski). 0940-9602/$ – see front matter © 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.aanat.2011.01.013
When the articular surfaces are brought into contact by a compressing force, the flexional movement follows the constrained motion of an equivalent four-bar linkage. The virtual 4 pivots are defined by the centres of curvature of the articulating surfaces: MTM , MTL , MFM , MFL (Fig. 1). The femorally defined axes (MFM , MFL ) run through the central area of the femur. Axis MTM , given by the flat medial concavity, lies cranial to the femoral centres, but the axis MTL of the flat lateral convexity caudal thereof. Line T connecting the tibial centres of curvature represents the position of the tibia and line F connecting the femoral centres of curvature defines the position of the femur. Note: The natural lateral tibial compartment also shows convex contours and the femoral condyle a small posterior displacement in the sagittal plane (Nägerl et al., 1993). Fig. 2 shows photos of the AEQUOS G1 components. The dimensions of the four-bar-linkage were optimized to the effect that the natural scales of the knee were achieved by the prosthesis in close approximation (Abicht, 2006).
196
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
Fig. 1. AEQUOS G1 TKR in extension and 60◦ -flexion with measures of the equivalent four-bar linkage. Approximately true to scale. MTM , MTL , MFM , MFL = centres of curvature of the articulating surfaces (T = Tibia, F = Femur, M = medial, L = lateral). IRA = position of the instantaneous rotational axis in 60◦ -flexion.
Since 2007, more than 1000 implantations have been performed. In the following we report on in vivo measurements illustrating unique mechanical properties in relation to the patello–femoral joint, as well as on the medical advantages of the prosthesis. 2. The problem of friction In the natural knee the problem of friction is solved not only by smooth articulating surfaces together with a special lubricant, the synovia, but also by implementing ingenious knee kinematics: Weber and Weber (1836) found out that the tibial articulating surfaces slide and roll along the femoral surfaces when the knee is bent. In his textbook, Fischer (1907) summarized the in vitro measurements of Langer (1858) and of Braune and Fischer (1891), reanalysing the in vivo measurements of knee movements by X-ray series of Zuppinger (1904). He concluded that, in flexion, the articulating surfaces initially rollback, then predominantly slide after an angle of flexion of approx. 20–30◦ , and that in further flexion both articulating contacts become stationary on the “tibial plateau”. Fischer (1907) also explained that the instantaneous rotational axis (IRA), initially in a distal position, migrates with increasing flexional angle into the centre of the femoral condyle. Nägerl et al. (2009) showed by reanalyses of published data of the “Freeman group” (Iwaki et al., 2000; Pinskerova et al., 2001, 2004) that under load the knee rolls to a high degree (≈90%) between 0 and 25◦ flexion. Hence, in the natural knee, the problem of friction is especially solved for the human gait: in the stance phase, when the knee is extremely loaded and sliding friction would achieve extreme values, friction is now reduced by dominant rolling of the articulating surfaces. Note: the coefficient of sliding friction is proportional to the compressive load whereas the much lesser coefficient of rolling friction is comparatively hardly influenced by load. The dimensions of the AEQUOS four-bar linkage are laid-out so that under load the constraint flexional movement imitates the natural rollback in close approximation. In 60◦ -flexion, e.g. both contacts are – conditioned by the design of the four-bar linkage – posteriorly shifted by ≈12 mm on the tibia “plateau” and the IRA lies between the femoral centres and the contacts (Fig. 1). In further flexion the IRA migrates towards the femoral centres as described by Fischer (1907) for the natural knee. Sliding prevails more and more and the contacts become stationary on the tibia. Hence, during the stance phase of gait the articular surfaces effectively roll as revealed by wide contact migrations (Fig. 3). But, in the swing
Fig. 2. Components of the AEQUOS G1 TKR: F = femoral part; T = tibial inlay; TS = tibial support; TP = stabilizing peg. The metallic margin of TS is undercut and represents the enclosure of the tibial inlay. During surgery the cooled polyethylene inlay is attached to TS. By means of the following inevitable thermal expansion the inlay is irremovably fixed.
Fig. 3. Gait characteristics and construction-conditioned position of the medial tibial contact in AEQUOS G1 TKR: in the stance phase the knee effectively rotates between 0 and 25◦ -flexion under high compressive loads and rolls to an extent of 90%: The joint contacts move by ∼7 mm within the first 25◦ of flexion. Between 35◦ and 60◦ flexion the contacts only shift by ∼2.5 mm: sliding is dominant.
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
197
Fig. 4. Lateral X-ray pictures: in vivo measurements of patellar tendon angle ˇ and flexion angle ˛ of an AEQUOS G1. Left: Flexion: 0◦ , PT angle: 15◦ . Right: Flexion 43◦ , PT angle: 3◦ . The PT angle pivoted by ˇ = 12◦ by ˛ = 43◦ of flexion.
phase, when load and therefore friction are absent, the tibial contacts remain largely stationary and the femur predominantly slides by wheel-spin. Mean degrees of rolling are 90% in 0–25◦ flexion and 43% in 35–60◦ flexion. Paradoxically, the idea of minimizing friction by implementing the natural rollback as nearly as possible did not play any role in the history of conventional TKR development. The concepts of these designs were mainly focussed on minimizing sliding friction by adequate material pairings (e.g. metal versus polyethylene) and on the problem of how to avoid possible local overstressing of the polyethylene beyond the critical value of 24 MPa (Postak et al., 1998). The local stress was measured by static experiments in testing machines with the help of prescale-films. In doing so, the dynamic character of the high joint loading during gait – i.e. that, therefore, the high loads only act during short time intervals – was not taken into account. Thus, however, the view of a most critical aspect in assessing the material stability of polyethylene was lost: Above all, increase in the temperature of the polyethylene must be avoided (Galetz, 2009). In the AEQUOS G1, the rolling back and forth of the articular surfaces under load makes the synovia flow back and forth and thus to continually lubricate the tibial polyethylene surfaces. Thus the heat, which is produced by the unavoidable solid viscosity of PE, is dissipated: The temperature of PE hardly increases though it is periodically deformed. Since the rigidity of this material would be sensitively reduced by temperature increase, as shown by Galetz (2009) in experiments on models, the back and forth rolling of the AEQUOS G1 would implement the necessary “cooling mechanism” during gait.
According to the established method by Pandit et al. (2005) we have measured the relationship of the patellar tendon angle ˇ as a function of the flexion angle ˛ for AEQUOS G1 TKR (Nägerl et al., 2008). For the method see Fig. 4: in extension and flexion the anterior margin of the patellar tendon (PT) and the position of the femur are traced and then compared with each other. Here, the patellar tendon was pivoted by 12◦ for 43◦ -flexion. The resulting relative patellar pivoting of 0.28 were consistent with Pandit’s data on the natural knee. Similar data have been published by Nägerl et al. (2008). Both series of lateral X-ray pictures of the “AEQUOS” and the contralateral natural knee of an exemplary patient demonstrated that the AEQUOS-TKR can approximately achieve natural knee rollback (Fig. 5). Rollback enlarges the angle between the patellar tendon and the tendon of the quadriceps muscle. A simple model illustrates this geometric fact (Fig. 6). Thus, rollback reduces the compressive force in the patello–femoral joint, the bending stress of the patella and simultaneously the tension of the patellar tendon. Without rollback, as in conventional TKR, knee flexions involve more acute angles between the patellar tendon and the ten-
3. Rollback and patellofemoral joint Pandit et al. (2005) have shown for natural knees, that, in flexional motion out of extension the patellar tendon pivots to the back by 0.3◦ per 1◦ flexion. Pandit et al. (2005) traced back this relative patellar pivoting of 0.3 to the rollback in the natural tibio-femoral joint. They also showed that, in knees with conventional prostheses, the relative patellar pivoting was small: the values ranged between 0 and <0.1 depending on the type of TKR. From these measurements there must be concluded that conventional prostheses do not restore natural knee kinematics.
Fig. 5. In vivo measurements of ˇ(˛)-relations: AEQUOS G1 and contralateral natural knee. The dashed line illustrates the relative patellar pivoting of 0.3.
198
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
Fig. 6. Illustration of load relieving in the patello–femoral joint by rollback: in extension, the contact PF in the joint is located at the lower end of the patella (left). Though knee flexion around centre C without rollback pivots patella P. The position of the patellar tendon is hardly altered (middle). In the case of rollback, however, the contacts FT between femur and tibia move anteriorly. Thus pivoting of patellar tendon and simultaneous straightening of the angle between patellar (PS) and quadriceps tendon (QS) occur. Furthermore the lever arm LPS of the force in the patellar tendon as to contact FT is prolonged (right).
Fig. 7. Posterior cruciate ligament PCL and rollback. PCLT = tibial insertion, PCLF = femoral insertion.
don of the quadriceps muscle and thus higher loading in the patello–femoral joint. We suppose that the anterior knee pain, which often appears after conventional TKR, is most probably associated with the increased loading of the patello–femoral joint. During rollback contacts of the FT are shifted posteriorly. Therefore the effective lever arm LPS of the force of patellar tendon increases in comparison with the state without rollback (Fig. 6). When this increase of the lever arm (LPS ) is absent, the M. quadriceps suffers a partial loss of the extensional torque (moment) in flexion. Since this loss of extensional moment would disequilibrate the given flexion position it must be compensated by enhancing of M. quadriceps force. Hence, the increase in loading in the patello–femoral joint. In contrast the rollback of the natural knee and the AEQUOS TKR leads to an increase in the lever arm LPS . Therefore the necessary extensional moment can be achieved without enhancing the force of the M. quadriceps. Summary: Under a given amount of force of the M. quadriceps, the rollback of the natural knee and that of the AEQUOS G1 TKR produce decreasing loads in the patello–femoral joint, decreasing tension in the patellar tendon, and at the same time an increase in the extensional moment.
4. Rollback and the posterior cruciate ligament The anterior cruciate ligament (ACL) must always and the posterior cruciate ligament (PCL) must frequently be sacrificed in conventional TKR. Why is the sacrifice of the PCL necessary? Fig. 7 illustrates the problem schematically. A: In extension, the femoral insertion PCLF of PCL is caudally and anteriorly positioned from the centre C of the femur. B: In flexion without rollback PCLF moves from position 1 along a circle around the femoral centre C to position 2 by the rotational angle ˛. By that, the PCL is extended by PCL. Generally PCL becomes too long so that PCL would be overstretched. C: In flexion by ˛ with rollback, insertion PCLF is additionally shifted by s to the back. By that the prolonging of the PCL by the rotation around C is practically cancelled so that in the natural knee only the distribution of slack and tense fibres is altered. The mean length remains practically unaltered. In AEQUOS G1 implantations, the PCL should be retained for 2 reasons: (1) PCL represents an important part of the sensory
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
199
rollback should minimize material wear due to friction, maximize the lifetime of the prosthesis and reduce the loading of the patello–femoral joint. To confirm this thesis one clearly has to wait for the long term results. References
Fig. 8. Follow up study by visual analogue pain scale (VAS).
apparatus for perception of knee position (Nägerl et al., 2002); (2) since AEQUOS G1 closely imitates natural mechanical properties of the knee, it consequently might show similar problems to that of the natural knee when the PCL is absent: the respective posterior pull could be similarly diagnosed as soon as the joint is unloaded by compression. 5. Clinical experience with AEQUOS G1 The maximum flexion hardly exceeds 115◦ . This does not have negative effects on every day living of the patients. First fluoroscopic data suggest that the more important kinematic property in reproduction of natural rollback seems to work and with that the load relieving of the patello–femoral joint. Investigations by means of the VAS pain scale suggest that after implantation of the AEQUOS G1 the general knee pain returns to values of healthy knees (Fig. 8). This information implies that the development of AEQUOS G1 represents a novel type of TKR which, when implemented, could restore or improve knee function for the patients. 6. Conclusions Initial clinical data are promissing. The AEQUOS G1 TKR equals the good to excellent short term results of other models. First kinematical in vivo data suggest that the AEQUOS G1 TKR works as it is constructed, namely producing a physiological rollback. This is the unique feature of this prosthesis. Producing
Abicht, C., 2006. Kniegelenke nach dem Viergelenkprinzip – Konzeption, Design und tribologische Eigenschaften einer neuen Knieendoprothese mit naturnaher Gelenkgeometrie. Diss. Ernst-Moritz-Arndt-Universität, Greifswald. Braune, W., Fischer, O., 1891. Bewegungen des Kniegelenkes, nach einer neuen Methode am lebenden Menschen gemessen. S. Hirtzel. Leipzig. Fischer, O., 1907. Kinematik organischer Gelenke. Vieweg-Verlag, Braunschweig. Frosch, K.-H., Floerkemeier, T., Abicht, C., Adam, P., Date, H., Fanghänel, J., Stürmer, K.M., Kubein-Meesenburg, D., Nägerl, H., 2009a. Eine neuartige Knieendoprothese mit physiologischer Gelenkform – Teil 1: Biomechanische Grundlagen und tribologische Untersuchungen. Unfallchirurg 112 (2), 168–175. Frosch, K.-H., Nägerl, H., Kubein-Meesenburg, D., Dörner, J., Dathe, H., Hellerer, O., Dumont, C., Stürmer, K.M., 2009b. Eine neuartige Kniegelenksendoprothese mit physiologischer Gelenkform – Teil 2: Erste klinische Ergebnisse. Unfallchirurg 112 (2), 176–184. Galetz, M., 2009. Influence of Temperature on Friction and Deformation Mechanisms in Artificial Knee Joints. Diss. Universität Bayreuth. Shaker Verlag, Aachen. Iwaki, H., Pinskerova, V., Freeman, M.A., 2000. Tibiofemoral movement. 1. The shapes and relative movements of the femur and tibia in the unloaded cadaver knee. J. Bone Joint Surg. Br. 82B, 1189–1195. Langer, K., 1858. Das Kniegelenk des Menschen. Sitzungsberichte der Akademie der Wissenschaften in Wien. Mathematisch-naturwissenschaftliche Klasse. Vol. 2 u. 3. Nägerl, H., Kubein-Meesenburg, D., Fanghänel, J., Cotta, H., 1993. Biomechanische Prinzipien bei Diarthrosen und Synarthrosen – Teil III: Mechanik des Tibiafemoralgelenks und Funktion der Kreuzbander. Z. Orthop. 131, 385–396. Nägerl, H., Kubein-Meesenburg, D., Miehe, B., Fanghänel, J., 2002. The sensory apparatus for perception in the tibiofemoral joint and outlines of a functional knee endoprosthesis. Acta Bioeng. Biomech. 4 (Suppl. 1), 319–320. Nägerl, H., Frosch, K.-H., Wachowski, M.M., Dumont, C., Abicht, C., Adam, P., KubeinMeesenburg, D., 2008. A novel total knee replacement by rolling articulating surfaces. In vivo functional measurements and tests. Acta Bioeng. Biomech. 10 (1), 55–60. Nägerl, H., Walters, J., Frosch, K.-H., Dumont, C., Kubein-Meesenburg, D., Fanghänel, J., Wachowski, M.M., 2009. J. Physiol. Pharmacol. 60 (Suppl. 8), 69–72. Pandit, H., Ward, T., Hollinghurt, D., Beard, D.J., Gill, H.S., Thomas, N.P., Murray, D.W., 2005. Influence of surface geometry and the cam-post mechanism on the kinematics of total knee replacement. J. Bone Joint Surg. Br. 87B, 940–945. Pinskerova, V., Iwaki, A., Freeman, M.A., 2001. The shapes and relative movements of the femur and tibia in the unloaded cadaveric knee: a study using MRI as an anatomical tool. In: Insall, J.N., Scott, R. (Eds.), Surgery of the Knee. , 3rd ed. W.B. Saunders, Philadelphia. Pinskerova, V., Johal, P., Nakagawa, S., Sosna, A., Williams, A., Gedroyc, W., Freeman, M.A., 2004. Does the femur roll back with flexion? J. Bone Joint Surg. Br. 86B, 925–931. Postak, P.D., Heim, C.S., Greenwald, A.S., 1998. Tibial Plateau Surface Stress in TKA: A Factor Influencing Polymer Damage – Series VI. Orthopaedic Research Labratories, The Mt. Sinai Medical Centre, Cleveland, Ohio. Weber, W., Weber, F., 1836. Mechanik der menschlichen Gehwerkzeuge. Göttingen. Translated by Maquet, P., Furlong, R. (1992): Mechanics of the human walking apparatus. Section 4: On the knee. Springer, Berlin, p. 75. Zuppinger, H., 1904. Die active Flexion im unbelasteten Kniegelenk. Züricher Habil Schr., Bergmann, Wiesbaden.
Contents lists available at ScienceDirect
Annals of Anatomy journal homepage: www.elsevier.de/aanat
Total knee replacement with natural rollback Martin Michael Wachowski a,∗, Tim Alexander Walde a, Peter Balcarek a, Jan Philipp Schüttrumpf a, Stephan Frosch a , Caspar Stauffenberg b , Karl-Heinz Frosch a , Christoph Fiedler c , Jochen Fanghänel d , Dietmar Kubein-Meesenburg b , Hans Nägerl b a
Department of Trauma Surgery, Plastic and Reconstructive Surgery, Georg August University of Göttingen, D-37099 Göttingen, Germany Department of Orthodontics, Georg August University, D-37099 Göttingen, Germany Institute of Materials Science and Technology, Friedrich Schiller University of Jena, D-07743 Jena, Germany d Department of Orthodontics, Regensburg University Medical Center, D-93053 Regensburg, Germany b c
a r t i c l e
i n f o
Article history: Received 20 December 2010 Accepted 24 January 2011
Keywords: Tibiofemoral joint Total knee replacement TKR Rollback Aequos Patella tendon angle
s u m m a r y A novel class of total knee replacement (AEQUOS G1) is introduced which features a unique design of the articular surfaces. Based on the anatomy of the human knee and differing from all other prostheses, the lateral tibial “plateau” is convexly curved and the lateral femoral condyle is posteriorly shifted in relation to the medial femoral condyle. Under compressive forces the configuration of the articular surfaces of human knees constrains the relative motion of femur and tibia in flexion/extension. This constrained motion is equivalent to that of a four-bar linkage, the virtual 4 pivots of which are given by the centres of curvature of the articulating surfaces. The dimensions of the four-bar linkage were optimized to the effect that constrained motion of the total knee replacement (TKR) follows the flexional motion of the human knee in close approximation, particularly during gait. In pilot studies lateral X-ray pictures have demonstrated that AEQUOS G1 can feature the natural rollback in vivo. Rollback relieves the load of the patello–femoral joint and minimizes retropatellar pressure. This mechanism should reduce the prevalence of anterior knee pain. The articulating surfaces roll predominantly in the stance phase. Consequently sliding friction is replaced by the lesser rolling friction under load. Producing rollback should minimize material wear due to friction and maximize the lifetime of the prosthesis. To definitely confirm these theses one has to wait for the long term results. © 2011 Elsevier GmbH. All rights reserved.
1. Introduction In diverse publications we have reported on a novel Total Knee Replacement (AEQUOS G1 TKR) which shows an unique design of articulating surfaces (Frosch et al., 2009a, 2009b; Nägerl et al., 2008): The lateral tibial articulating surface is convexly curved in anterior/posterior direction and the medial femoral condyle is slightly displaced to anterior (Fig. 1). By this asymmetric design it was intended to implement an artificial implant that sustains essential biomechanical functions of the natural knee. Nonetheless AEQUOS G1 TKR possesses the conventional material pairing for the articulating surfaces (Fig. 2): cobalt chrome (femur) versus polyethylene (tibia).
∗ Corresponding author at: Department of Trauma Surgery, Plastic and Reconstructive Surgery, Georg August University, Robert-Koch-Strasse 40, D-37075 Göttingen, Germany. Tel.: +49 551396114; fax: +49 551398981. E-mail address: [email protected] (M.M. Wachowski). 0940-9602/$ – see front matter © 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.aanat.2011.01.013
When the articular surfaces are brought into contact by a compressing force, the flexional movement follows the constrained motion of an equivalent four-bar linkage. The virtual 4 pivots are defined by the centres of curvature of the articulating surfaces: MTM , MTL , MFM , MFL (Fig. 1). The femorally defined axes (MFM , MFL ) run through the central area of the femur. Axis MTM , given by the flat medial concavity, lies cranial to the femoral centres, but the axis MTL of the flat lateral convexity caudal thereof. Line T connecting the tibial centres of curvature represents the position of the tibia and line F connecting the femoral centres of curvature defines the position of the femur. Note: The natural lateral tibial compartment also shows convex contours and the femoral condyle a small posterior displacement in the sagittal plane (Nägerl et al., 1993). Fig. 2 shows photos of the AEQUOS G1 components. The dimensions of the four-bar-linkage were optimized to the effect that the natural scales of the knee were achieved by the prosthesis in close approximation (Abicht, 2006).
196
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
Fig. 1. AEQUOS G1 TKR in extension and 60◦ -flexion with measures of the equivalent four-bar linkage. Approximately true to scale. MTM , MTL , MFM , MFL = centres of curvature of the articulating surfaces (T = Tibia, F = Femur, M = medial, L = lateral). IRA = position of the instantaneous rotational axis in 60◦ -flexion.
Since 2007, more than 1000 implantations have been performed. In the following we report on in vivo measurements illustrating unique mechanical properties in relation to the patello–femoral joint, as well as on the medical advantages of the prosthesis. 2. The problem of friction In the natural knee the problem of friction is solved not only by smooth articulating surfaces together with a special lubricant, the synovia, but also by implementing ingenious knee kinematics: Weber and Weber (1836) found out that the tibial articulating surfaces slide and roll along the femoral surfaces when the knee is bent. In his textbook, Fischer (1907) summarized the in vitro measurements of Langer (1858) and of Braune and Fischer (1891), reanalysing the in vivo measurements of knee movements by X-ray series of Zuppinger (1904). He concluded that, in flexion, the articulating surfaces initially rollback, then predominantly slide after an angle of flexion of approx. 20–30◦ , and that in further flexion both articulating contacts become stationary on the “tibial plateau”. Fischer (1907) also explained that the instantaneous rotational axis (IRA), initially in a distal position, migrates with increasing flexional angle into the centre of the femoral condyle. Nägerl et al. (2009) showed by reanalyses of published data of the “Freeman group” (Iwaki et al., 2000; Pinskerova et al., 2001, 2004) that under load the knee rolls to a high degree (≈90%) between 0 and 25◦ flexion. Hence, in the natural knee, the problem of friction is especially solved for the human gait: in the stance phase, when the knee is extremely loaded and sliding friction would achieve extreme values, friction is now reduced by dominant rolling of the articulating surfaces. Note: the coefficient of sliding friction is proportional to the compressive load whereas the much lesser coefficient of rolling friction is comparatively hardly influenced by load. The dimensions of the AEQUOS four-bar linkage are laid-out so that under load the constraint flexional movement imitates the natural rollback in close approximation. In 60◦ -flexion, e.g. both contacts are – conditioned by the design of the four-bar linkage – posteriorly shifted by ≈12 mm on the tibia “plateau” and the IRA lies between the femoral centres and the contacts (Fig. 1). In further flexion the IRA migrates towards the femoral centres as described by Fischer (1907) for the natural knee. Sliding prevails more and more and the contacts become stationary on the tibia. Hence, during the stance phase of gait the articular surfaces effectively roll as revealed by wide contact migrations (Fig. 3). But, in the swing
Fig. 2. Components of the AEQUOS G1 TKR: F = femoral part; T = tibial inlay; TS = tibial support; TP = stabilizing peg. The metallic margin of TS is undercut and represents the enclosure of the tibial inlay. During surgery the cooled polyethylene inlay is attached to TS. By means of the following inevitable thermal expansion the inlay is irremovably fixed.
Fig. 3. Gait characteristics and construction-conditioned position of the medial tibial contact in AEQUOS G1 TKR: in the stance phase the knee effectively rotates between 0 and 25◦ -flexion under high compressive loads and rolls to an extent of 90%: The joint contacts move by ∼7 mm within the first 25◦ of flexion. Between 35◦ and 60◦ flexion the contacts only shift by ∼2.5 mm: sliding is dominant.
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
197
Fig. 4. Lateral X-ray pictures: in vivo measurements of patellar tendon angle ˇ and flexion angle ˛ of an AEQUOS G1. Left: Flexion: 0◦ , PT angle: 15◦ . Right: Flexion 43◦ , PT angle: 3◦ . The PT angle pivoted by ˇ = 12◦ by ˛ = 43◦ of flexion.
phase, when load and therefore friction are absent, the tibial contacts remain largely stationary and the femur predominantly slides by wheel-spin. Mean degrees of rolling are 90% in 0–25◦ flexion and 43% in 35–60◦ flexion. Paradoxically, the idea of minimizing friction by implementing the natural rollback as nearly as possible did not play any role in the history of conventional TKR development. The concepts of these designs were mainly focussed on minimizing sliding friction by adequate material pairings (e.g. metal versus polyethylene) and on the problem of how to avoid possible local overstressing of the polyethylene beyond the critical value of 24 MPa (Postak et al., 1998). The local stress was measured by static experiments in testing machines with the help of prescale-films. In doing so, the dynamic character of the high joint loading during gait – i.e. that, therefore, the high loads only act during short time intervals – was not taken into account. Thus, however, the view of a most critical aspect in assessing the material stability of polyethylene was lost: Above all, increase in the temperature of the polyethylene must be avoided (Galetz, 2009). In the AEQUOS G1, the rolling back and forth of the articular surfaces under load makes the synovia flow back and forth and thus to continually lubricate the tibial polyethylene surfaces. Thus the heat, which is produced by the unavoidable solid viscosity of PE, is dissipated: The temperature of PE hardly increases though it is periodically deformed. Since the rigidity of this material would be sensitively reduced by temperature increase, as shown by Galetz (2009) in experiments on models, the back and forth rolling of the AEQUOS G1 would implement the necessary “cooling mechanism” during gait.
According to the established method by Pandit et al. (2005) we have measured the relationship of the patellar tendon angle ˇ as a function of the flexion angle ˛ for AEQUOS G1 TKR (Nägerl et al., 2008). For the method see Fig. 4: in extension and flexion the anterior margin of the patellar tendon (PT) and the position of the femur are traced and then compared with each other. Here, the patellar tendon was pivoted by 12◦ for 43◦ -flexion. The resulting relative patellar pivoting of 0.28 were consistent with Pandit’s data on the natural knee. Similar data have been published by Nägerl et al. (2008). Both series of lateral X-ray pictures of the “AEQUOS” and the contralateral natural knee of an exemplary patient demonstrated that the AEQUOS-TKR can approximately achieve natural knee rollback (Fig. 5). Rollback enlarges the angle between the patellar tendon and the tendon of the quadriceps muscle. A simple model illustrates this geometric fact (Fig. 6). Thus, rollback reduces the compressive force in the patello–femoral joint, the bending stress of the patella and simultaneously the tension of the patellar tendon. Without rollback, as in conventional TKR, knee flexions involve more acute angles between the patellar tendon and the ten-
3. Rollback and patellofemoral joint Pandit et al. (2005) have shown for natural knees, that, in flexional motion out of extension the patellar tendon pivots to the back by 0.3◦ per 1◦ flexion. Pandit et al. (2005) traced back this relative patellar pivoting of 0.3 to the rollback in the natural tibio-femoral joint. They also showed that, in knees with conventional prostheses, the relative patellar pivoting was small: the values ranged between 0 and <0.1 depending on the type of TKR. From these measurements there must be concluded that conventional prostheses do not restore natural knee kinematics.
Fig. 5. In vivo measurements of ˇ(˛)-relations: AEQUOS G1 and contralateral natural knee. The dashed line illustrates the relative patellar pivoting of 0.3.
198
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
Fig. 6. Illustration of load relieving in the patello–femoral joint by rollback: in extension, the contact PF in the joint is located at the lower end of the patella (left). Though knee flexion around centre C without rollback pivots patella P. The position of the patellar tendon is hardly altered (middle). In the case of rollback, however, the contacts FT between femur and tibia move anteriorly. Thus pivoting of patellar tendon and simultaneous straightening of the angle between patellar (PS) and quadriceps tendon (QS) occur. Furthermore the lever arm LPS of the force in the patellar tendon as to contact FT is prolonged (right).
Fig. 7. Posterior cruciate ligament PCL and rollback. PCLT = tibial insertion, PCLF = femoral insertion.
don of the quadriceps muscle and thus higher loading in the patello–femoral joint. We suppose that the anterior knee pain, which often appears after conventional TKR, is most probably associated with the increased loading of the patello–femoral joint. During rollback contacts of the FT are shifted posteriorly. Therefore the effective lever arm LPS of the force of patellar tendon increases in comparison with the state without rollback (Fig. 6). When this increase of the lever arm (LPS ) is absent, the M. quadriceps suffers a partial loss of the extensional torque (moment) in flexion. Since this loss of extensional moment would disequilibrate the given flexion position it must be compensated by enhancing of M. quadriceps force. Hence, the increase in loading in the patello–femoral joint. In contrast the rollback of the natural knee and the AEQUOS TKR leads to an increase in the lever arm LPS . Therefore the necessary extensional moment can be achieved without enhancing the force of the M. quadriceps. Summary: Under a given amount of force of the M. quadriceps, the rollback of the natural knee and that of the AEQUOS G1 TKR produce decreasing loads in the patello–femoral joint, decreasing tension in the patellar tendon, and at the same time an increase in the extensional moment.
4. Rollback and the posterior cruciate ligament The anterior cruciate ligament (ACL) must always and the posterior cruciate ligament (PCL) must frequently be sacrificed in conventional TKR. Why is the sacrifice of the PCL necessary? Fig. 7 illustrates the problem schematically. A: In extension, the femoral insertion PCLF of PCL is caudally and anteriorly positioned from the centre C of the femur. B: In flexion without rollback PCLF moves from position 1 along a circle around the femoral centre C to position 2 by the rotational angle ˛. By that, the PCL is extended by PCL. Generally PCL becomes too long so that PCL would be overstretched. C: In flexion by ˛ with rollback, insertion PCLF is additionally shifted by s to the back. By that the prolonging of the PCL by the rotation around C is practically cancelled so that in the natural knee only the distribution of slack and tense fibres is altered. The mean length remains practically unaltered. In AEQUOS G1 implantations, the PCL should be retained for 2 reasons: (1) PCL represents an important part of the sensory
M.M. Wachowski et al. / Annals of Anatomy 194 (2012) 195–199
199
rollback should minimize material wear due to friction, maximize the lifetime of the prosthesis and reduce the loading of the patello–femoral joint. To confirm this thesis one clearly has to wait for the long term results. References
Fig. 8. Follow up study by visual analogue pain scale (VAS).
apparatus for perception of knee position (Nägerl et al., 2002); (2) since AEQUOS G1 closely imitates natural mechanical properties of the knee, it consequently might show similar problems to that of the natural knee when the PCL is absent: the respective posterior pull could be similarly diagnosed as soon as the joint is unloaded by compression. 5. Clinical experience with AEQUOS G1 The maximum flexion hardly exceeds 115◦ . This does not have negative effects on every day living of the patients. First fluoroscopic data suggest that the more important kinematic property in reproduction of natural rollback seems to work and with that the load relieving of the patello–femoral joint. Investigations by means of the VAS pain scale suggest that after implantation of the AEQUOS G1 the general knee pain returns to values of healthy knees (Fig. 8). This information implies that the development of AEQUOS G1 represents a novel type of TKR which, when implemented, could restore or improve knee function for the patients. 6. Conclusions Initial clinical data are promissing. The AEQUOS G1 TKR equals the good to excellent short term results of other models. First kinematical in vivo data suggest that the AEQUOS G1 TKR works as it is constructed, namely producing a physiological rollback. This is the unique feature of this prosthesis. Producing
Abicht, C., 2006. Kniegelenke nach dem Viergelenkprinzip – Konzeption, Design und tribologische Eigenschaften einer neuen Knieendoprothese mit naturnaher Gelenkgeometrie. Diss. Ernst-Moritz-Arndt-Universität, Greifswald. Braune, W., Fischer, O., 1891. Bewegungen des Kniegelenkes, nach einer neuen Methode am lebenden Menschen gemessen. S. Hirtzel. Leipzig. Fischer, O., 1907. Kinematik organischer Gelenke. Vieweg-Verlag, Braunschweig. Frosch, K.-H., Floerkemeier, T., Abicht, C., Adam, P., Date, H., Fanghänel, J., Stürmer, K.M., Kubein-Meesenburg, D., Nägerl, H., 2009a. Eine neuartige Knieendoprothese mit physiologischer Gelenkform – Teil 1: Biomechanische Grundlagen und tribologische Untersuchungen. Unfallchirurg 112 (2), 168–175. Frosch, K.-H., Nägerl, H., Kubein-Meesenburg, D., Dörner, J., Dathe, H., Hellerer, O., Dumont, C., Stürmer, K.M., 2009b. Eine neuartige Kniegelenksendoprothese mit physiologischer Gelenkform – Teil 2: Erste klinische Ergebnisse. Unfallchirurg 112 (2), 176–184. Galetz, M., 2009. Influence of Temperature on Friction and Deformation Mechanisms in Artificial Knee Joints. Diss. Universität Bayreuth. Shaker Verlag, Aachen. Iwaki, H., Pinskerova, V., Freeman, M.A., 2000. Tibiofemoral movement. 1. The shapes and relative movements of the femur and tibia in the unloaded cadaver knee. J. Bone Joint Surg. Br. 82B, 1189–1195. Langer, K., 1858. Das Kniegelenk des Menschen. Sitzungsberichte der Akademie der Wissenschaften in Wien. Mathematisch-naturwissenschaftliche Klasse. Vol. 2 u. 3. Nägerl, H., Kubein-Meesenburg, D., Fanghänel, J., Cotta, H., 1993. Biomechanische Prinzipien bei Diarthrosen und Synarthrosen – Teil III: Mechanik des Tibiafemoralgelenks und Funktion der Kreuzbander. Z. Orthop. 131, 385–396. Nägerl, H., Kubein-Meesenburg, D., Miehe, B., Fanghänel, J., 2002. The sensory apparatus for perception in the tibiofemoral joint and outlines of a functional knee endoprosthesis. Acta Bioeng. Biomech. 4 (Suppl. 1), 319–320. Nägerl, H., Frosch, K.-H., Wachowski, M.M., Dumont, C., Abicht, C., Adam, P., KubeinMeesenburg, D., 2008. A novel total knee replacement by rolling articulating surfaces. In vivo functional measurements and tests. Acta Bioeng. Biomech. 10 (1), 55–60. Nägerl, H., Walters, J., Frosch, K.-H., Dumont, C., Kubein-Meesenburg, D., Fanghänel, J., Wachowski, M.M., 2009. J. Physiol. Pharmacol. 60 (Suppl. 8), 69–72. Pandit, H., Ward, T., Hollinghurt, D., Beard, D.J., Gill, H.S., Thomas, N.P., Murray, D.W., 2005. Influence of surface geometry and the cam-post mechanism on the kinematics of total knee replacement. J. Bone Joint Surg. Br. 87B, 940–945. Pinskerova, V., Iwaki, A., Freeman, M.A., 2001. The shapes and relative movements of the femur and tibia in the unloaded cadaveric knee: a study using MRI as an anatomical tool. In: Insall, J.N., Scott, R. (Eds.), Surgery of the Knee. , 3rd ed. W.B. Saunders, Philadelphia. Pinskerova, V., Johal, P., Nakagawa, S., Sosna, A., Williams, A., Gedroyc, W., Freeman, M.A., 2004. Does the femur roll back with flexion? J. Bone Joint Surg. Br. 86B, 925–931. Postak, P.D., Heim, C.S., Greenwald, A.S., 1998. Tibial Plateau Surface Stress in TKA: A Factor Influencing Polymer Damage – Series VI. Orthopaedic Research Labratories, The Mt. Sinai Medical Centre, Cleveland, Ohio. Weber, W., Weber, F., 1836. Mechanik der menschlichen Gehwerkzeuge. Göttingen. Translated by Maquet, P., Furlong, R. (1992): Mechanics of the human walking apparatus. Section 4: On the knee. Springer, Berlin, p. 75. Zuppinger, H., 1904. Die active Flexion im unbelasteten Kniegelenk. Züricher Habil Schr., Bergmann, Wiesbaden.