- Email: [email protected]
A convex lateral tibial plateau for knee replacement
The Knee 13 (2006) 122 – 126 www.elsevier.com/locate/knee
A convex lateral tibial plateau for knee replacement J.V. Bare´, H.S. Gill *, D.J. Beard, D.W. Murray OOEC/Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Oxford, UK Nuffield Orthopaedic Centre, NHS Trust, Oxford, UK Received 4 August 2005; accepted 15 September 2005
Abstract Unicompartmental knee replacements have not performed as well in the lateral compartment as in the medial. This may be because the tibial components have flat or slightly concave surfaces which match the medial plateau but not the convex lateral plateau. The aim of this study was to find the optimal radius for a convex lateral tibial component. Twelve normal lateral tibial plateau were retrieved at knee replacement, and their surface contour in their mid sagittal plane was determined. The optimal circle was fitted and its radius measured. A series of different shaped tibial components were superimposed. From published information about the position of the femoral condyle relative to the tibia in different degrees of flexion, the flexion gap at these angles was determined. The average radius of the lateral tibial plateau was 40 mm. However, as the surface was polyradial it was not clear if this average radius would be optimal. In full flexion, a flat tibial plateau distracted the knee by 8 mm ( p < 0.001). A 75 mm radius spherical tibial plateau did not alter the knee kinematics significantly and gave rise to a change in joint distraction of 1.5 mm. Spherical tibial plateau of 50 mm and 25 mm radii significantly altered knee kinematics ( p < 0.001) and resulted in changes in distraction of 3 mm and 4 mm respectively. The optimal shape for a unicompartmental lateral tibial plateau is likely to be a spherical dome with radius of about 75 mm. The incorporation of this shape in the lateral side of a total knee replacement might improve its flexion. D 2005 Elsevier B.V. All rights reserved. Keywords: Lateral tibial plateau; Knee replacement
1. Introduction Isolated lateral compartment arthritis of the knee is rare and difficult to treat, it occurs much less commonly than isolated medial compartment arthritis [1]. The treatment options include osteotomy, unicompartmental knee replacement (UKR) and total knee replacement, none of which give very satisfactory results [2–6]. Potentially the best solution is unicompartmental replacement which preserves normal anatomy thereby restoring the kinematics of the knee to normal. However, the results of lateral unicompartmental knee replace* Corresponding author. Botnar Research Centre, Nuffield Orthopaedic Centre, Oxford, OX3 7LD, UK. Tel.: +44 1865 227457; fax: +44 1865 227671. E-mail address: [email protected] (H.S. Gill). 0968-0160/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.knee.2005.09.001
ment have been disappointing and have not been as good as medial unicompartmental replacement [7–10]. In the majority of published series the survival at 10 years is about 80% [8,11]. The lateral compartment of the knee is fundamentally different from the medial compartment. This may explain why UKR components that work well in the medial compartment are less appropriate for the lateral, and it may justify the use of different components in the lateral side. The main difference between the medial and lateral compartments is that the medial tibial plateau is slightly concave, whereas the lateral is convex. In addition, the lateral ligaments are much slacker and weaker than the medial, and dynamic stability, provided by the ileotibial tract and popliteus, is important. During flexion there is a small amount of movement of the medial femoral condyle posterior on the tibia. In contrast, on the lateral side the femoral condyle
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126
123
to the long axis of the tibia using either intra or extra medullary instrumentation. The medio-lateral width of the intact tibial plateau was measured. A sagittal saw cut was made in the middle of the lateral tibial plateau. The cut surface was photographed, together with a scale, in order to display the mid-sagittal contour of the lateral tibial plateau (Fig. 2). Care was taken to ensure that the principal axis of the camera was aligned perpendicular to the plane through the midsagittal section of each specimen. 2.2. Image analysis
Fig. 1. MRI view of lateral compartment of the knee at high flexion, from Nakagawa. [Note to Editor: we request permission to show this figure from Nakagawa et al. [12]].
moves so far posteriorly that in full flexion its centre is posterior to the tibial plateau (Fig. 1, from Nakagawa [12]). This posterior displacement of the femur could not be achieved with a unicompartmental replacement that had a flat tibial plateau. Therefore, the only way normal kinematics could be achieved would be with a convex tibial component. The wear rate of polyethylene is related exponentially to the contact pressure [13], which in turn is inversely related to the contact area. With conventional flat designs of fixed bearing unicompartmental replacement polyethylene wear is a problem but usually only in the long term [3,11]. If the bearing surface of the polyethylene was convex the contact area between it and the femoral component would decrease and the contact stress consequently increase, resulting in higher wear. The wear would probably occur early and be catastrophic. Therefore, for a convex tibial surface it will probably be necessary to use a mobile bearing and ideally one which remains in full congruous contact with the tibia. This can be best achieved with a spherical tibial surface, and bearing with a matching spherically concave under-surface. The aim of this study was firstly to determine the radius of the lateral tibial plateau in the sagittal plane, and secondly to investigate how tibial components of varying radii would influence knee kinematics.
Each mid-sagittal plane image was transferred to a workstation and then imported into a drawing software package (CorelDraw version 10, Corel Corporation, Canada). The image was then scaled such that the pixel dimensions matched the actual dimensions of the specimen, so that measurement could be directly made on the image. The approximate sites of contact in full extension and full flexion were identified. A ‘‘best fit’’ circle was superimposed by eye onto the articulating surface of the lateral tibial plateau between these two contact points. The radius of curvature of this circle was measured (Fig. 2). Outlines of four different UKR prosthetic tibial components were superimposed onto each image. The first tibial component was a flat tray, representing the currently used tray for a meniscal bearing design of UKR. The other three were circular in crosssection, representing hypothetical convex components with varying radii of curvature (25, 50 and 75 mm). The bottom surfaces of the circular components were flat, and were placed with a 7posterior slope relative to the original tibial cut. The upper surface of each outline was placed approximately 4 mm below the actual articular cartilage at the point of contact with the femur in full extension (defined below). Four millimetres represented the optimal thickness of a polyethylene meniscal bearing. The lateral femoral condyle was represented by a circle of radius 22 mm. The centre of this femoral condyle circle was placed at four positions relative to the tibia. These were at 7 mm, 8 mm, 21 mm and 27 mm anterior to the posterior edge of the articular cartilage of the lateral tibial plateau, and were representative of the position of the femur in a normal knee at 160-, 130-, 90- and 0- of flexion respectively, as described by Iwaki et al. [14], Hill et al.
2. Materials and methods The shapes of a series of tibial plateau were determined. Then previously published experimentally derived information about the position of the femoral condyle relative to the tibial plateau at various angles of knee flexion was used to determine how the different radii of convex tibial components would influence the flexion gap at various points in the flexion range. 2.1. Specimens Twelve tibial plateau were collected at the time of total knee arthroplasty performed for medial compartment and/or patellofemoral osteoarthritis. Each specimen had an intact lateral tibial plateau with intact cartilage. The tibial cut was made perpendicular
Fig. 2. Excised and sectioned lateral tibial plateau, showing fitted circle to surface.
124
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126
Fig. 3. Method of flexion gap measurement for a) flat tibial component and b) 25 mm radius tibial component.
[15] and Nakagawa et al. [12] in a series of MRI studies on cadaveric and living knees. The femoral condyle circle was positioned so that it just touched the actual articular cartilage of the tibial cross-section. The closest distance from the femoral condyle circle to each tibial prosthetic component surface at each of the four above positions was measured. This distance was representative of the gap that would be present between the femoral condyle and prosthetic tibial component surface (Fig. 3); the flexion gap was termed f. During meniscal bearing lateral UKR surgery, full extension is taken as the reference point at which the thickness of the meniscal bearing is selected. Therefore, any differences between the measurement made in full extension and the other three measurements were indicative of articular surface distraction as the joint flexes, this change was termed Df. An increase in articular surface distraction (positive values of Df) with the bearing inserted would be associated with distraction of the knee in flexion and tightening of the soft tissues. Conversely a decrease in articular surface distraction (negative values of Df) would be associated with slackening of the soft tissues.
average anterior – posterior (AP) length was 43.3 mm (95% CI T 2.7 mm, range 35.5 to 53.7 mm). The mid lateral sagittal radius of curvature of the articular surface averaged 41.4 mm (95% CI T 4.0 mm, range 30.0 to 48.5 mm). The radius was related to the AP length of the plateau, this was a significant correlation with p = 0.011 and r 2 = 0.49. There was no significant correlation between radius and medio-lateral width. It was observed that the surface of the plateau was polyradial not circular, with the radius decreasing posteriorly on the plateau. The median change in articular surface distraction from that in extension, Df, at each flexion angle (0-, 90-, 130- and 160-) is
2.3. Statistical analysis To determine if the various designs of tibial component resulted in a statistically significant change in kinematics throughout the knee flexion range, the values of articular surface distraction, Df, for all flexion angles was analysed for each design using a one-way ANOVA. For zero degrees of knee flexion, Df was zero by definition. The statistical analysis was performed using the Statistics Toolbox version 4.0 available in Matlab 6.5 (The MathWorks Inc., MA, USA).
3. Results The average medio-lateral width of the 12 tibial plateau was 78.2 mm (95% CI T 4.7 mm, range 70.3 to 87.8 mm), and the
Fig. 4. Change in Df over flexion range from 0- to 160- for each design of tibial component, medians and interquartiles are shown. The *** indicates a significant change ( p < 0.001) in mean Df over the flexion range; NS indicates no significant change.
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126 Table 1 Change in Df over flexion range from 0- to 160- and results of ANOVA for Df at 0-, 90-, 130- and 160- knee flexion angles, for each design of tibial component Design
Change in Df over flexion range (mm)
ANOVA results
Flat 75 mm radius 50 mm radius 25 mm radius
9.9 1.5 3.0 3.9
p < 0.0001 p = 0.13 p < 0.001 p < 0.0001
shown for each design of tibial component (flat, 75 mm, 50 mm, 25 mm) in Fig. 4. The overall change in values of Df over the entire flexion range and statistical significance of the changes in articular surface distraction are given in Table 1. The results of the ANOVA showed that the articular surface distraction did not change statistically over the knee flexion range for the 75 mm radius component. However, for all the other designs, the flat, 50 mm and 25 mm radii, the changes in flexion gap (Df) were highly significant ( p < 0.001). For the 75 mm radius component the median change in articular distraction was 1.5 mm over the entire flexion range. The median change in articular surface distraction over the entire flexion range for the 50 mm radius component was 3 mm, the 25 mm radius and flat designs had much larger changes (approximately 4 mm and 10 mm respectively).
4. Discussion The average mid-sagittal radius of the 12 lateral tibial plateau was 41 mm. The radius was related to the AP length; the larger the length, the larger the radius. It was observed that the optimum radius varied from patient to patient and, more importantly, a circle did not fit well. In general, in the portion of the plateau that was in contact with the femur at low knee flexion angles had a larger radius, whereas the portion in contact at higher flexion angles had a smaller radius. Thus, a radius of 41 mm may not be the optimum. A better way to calculate the ideal radius is to determine the influence of different radii on the articular surface distraction at different angles of flexion. It was found that a flat tibial plateau resulted in, on average, 1 mm of articular surface approximation at 90- and 2 mm at 130-, relative to that at extension. This is probably acceptable. However, at 160- a flat plateau resulted in an increase in the articular surface distraction by, on average, 8 mm. Therefore, for the knee to flex fully with a flat component inserted, either the lateral side would have to be distracted by 8 mm or alternatively the kinematics would have to be abnormal, with the lateral side not rolling back normally. It is unlikely that either of these could occur, hence a flat plateau would be expected to block flexion. We have indeed observed per-operatively that a flat plateau does provide a block to flexion. With a tibial plateau of radius 75 mm, the articular surface distraction did not change significantly (ANOVA p = 0.13) over the full flexion range and the average decrease in articular surface distraction was less than 2
125
mm. It would therefore seem that with a 75 mm radius, the knee should function with virtually normal kinematics. With the 50 mm radius although the flexion gap changed significantly over the full range of flexion the change was only 3 mm. It is therefore likely that tibial plateau with a radius in the region of 50 mm would also probably be satisfactory. A 25 mm radius plateau would not be satisfactory as it results in a significant change in flexion gap of approximately 4 mm. There was considerable variation between patients in the articular surface distraction at different angles of flexion. It would therefore be necessary to make small adjustments to the component position to achieve optimal kinematics in all patients. For example, altering the tibial slope would influence the articular surface distraction, as would altering the position of the femoral component. However, alterations to component position would probably not accommodate the large abnormalities in articular surface distraction that occur with a flat or 25 mm radius component. For the flat tibial plateau the median changes in flexion gap were 2 mm up to 130- flexion and only became large above 130-. Therefore for traditional open lateral UKR which has an average range of motion (ROM) of 110- [8] a flat plateau is probably acceptable. However, evidence from the medial compartment suggests that with a minimally invasive (MIS) approach the ROM increases from about 110- to 130[16]. We would therefore expect that the average ROM with a minimally invasive lateral UKR also to be about 130-, and that some could potentially flex considerably more. These would be expected to have kinematic problems with a flat tibial plateau. It is therefore not surprising that the average ROM of MIS lateral UKR is actually 120- [17] and not 130-. We believe that if a 75 mm radius component was used higher flexion would be achieved. For a mobile bearing to achieve fully congruous contact in all positions and thus low wear, the sagittal plane shape of the tibial plateau should be circular. However, the three dimensional shape of the plateau could be cylindrical or spherical. Lateral UKR with unconstrained mobile bearings have had a 10% dislocation rate [18], with the majority of dislocations being medial into the intercondylar notch. A major advantage of a spherical plateau is that it should decrease the dislocation rate. This is because a spherically convex plateau with a corresponding concave bearing would substantially increase entrapment in a medio-lateral direction. It is appreciated this study has a number of limitations. In particular the position of the femoral component in different angles of flexion was determined from published average data. Substantial differences exist in rollback between patients. Although this would not influence the conclusions about the relative merits of different shaped plateau it could influence the angle of flexion at which the different plateau would cause problems. For example, patients with substantial early rollback would be likely to encounter problems with a flat plateau at much lower flexion angles than those predicted by this study. Further cadaveric and
126
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126
clinical studies need to be undertaken. In these the influence of flat and 75 mm radius tibial plateau should be compared. Total knee replacements (TKR) have in general not achieved excellent flexion. Recently, some new TKRs have been designed to achieve high flexion, but even these have often failed to achieve full flexion, or if they do they have had problems. We believe this is in part because the lateral femoral condyle of the normal knee subluxes off the back of the tibial plateau in high flexion and current designs of TKR do not allow this. If a circular lateral tibial plateau and lateral mobile bearing were incorporated in a TKR this might allow the TKR to achieve high flexion.
5. Conclusions The average radius in the sagittal plane of the lateral tibial plateau is about 40 mm. The tibial plateau, however, is polyradial so 40 mm may not be the ideal radius. The optimum radius of a lateral tibial component is probably about 75 mm as this radius causes the smallest change in articular surface distraction.
References [1] Altman RD, Fries JF, Bloch DA, Carstens J, Cooke TD, Genant H, et al. Radiographic assessment of progression in osteoarthritis. Arthritis Rheum 1987;30-11:1214 – 25. [2] Coventry MB. Proximal tibial varus osteotomy for osteoarthritis of the lateral compartment of the knee. J Bone Joint Surg Am 1987;691:32 – 8. [3] Witvoet J, Peyrache MD, Nizard R. Single-compartment ‘‘Lotus’’ type knee prosthesis in the treatment of lateralized gonarthrosis: results in 135 cases with a mean follow-up of 4.6 years. Rev Chir Orthop Reparatrice Appar Mot 1993;79-7:565 – 76. [4] Cameron HU, Botsford DJ, Park YS. Prognostic factors in the outcome of supracondylar femoral osteotomy for lateral compartment osteoarthritis of the knee. Can J Surg 1997;40-2:114 – 8.
[5] Mathews J, Cobb AG, Richardson S, Bentley G. Distal femoral osteotomy for lateral compartment osteoarthritis of the knee. Orthopedics 1998;21-4:437 – 40. [6] Newman J, Ackroyd C, Shah N. Unicompartmental or total knee replacement? Five-year results of a prospective, randomised trial of 102 osteoarthritic knees with unicompartmental arthritis. J Bone Joint Surg Br 1998;50:862 – 5. [7] Goodfellow JW, Kershaw CJ, Benson M.K, O’Connor JJ. The Oxford Knee for unicompartmental osteoarthritis. The first 103 cases. J Bone Joint Surg Br 1988;70-5:692 – 701. [8] Gunther TV, Murray DM, Miller R, Wallace DA, Carr AJ, O’Connor JJ, et al. Lateral compartment arthroplasty with the Oxford Meniscal knee. Knee 1996;3:33 – 9. [9] Murray DW, Goodfellow JW, O’Connor JJ. The Oxford medial unicompartmental arthroplasty: a ten-year survival study. J Bone Joint Surg Br 1998;80-6:983 – 9. [10] Svard UC, Price AJ. Oxford medial unicompartmental knee arthroplasty. A survival analysis of an independent series. J Bone Joint Surg Br 2001;83-2:191 – 4. [11] Newman JH. Unicompartmental knee replacement. Knee 2000;7: 63 – 70. [12] Nakagawa S, Kadoya Y, Todo S, Kobayashi A, Sakamoto H, Freeman MA, et al. Tibiofemoral movement 3: full flexion in the living knee studied by MRI. J Bone Joint Surg Br 2000;82-8:1199 – 200. [13] Rostoker W, Galante JO. Contact pressure dependence of wear rates of ultra high molecular weight polyethylene. J Biomed Mater Res 1979;13:957 – 64. [14] Iwaki H, Pinskerova V, Freeman MA. Tibiofemoral movement 1: the shapes and relative movements of the femur and tibia in the unloaded cadaver knee. J Bone Joint Surg Br 2000;82-8:1189 – 95. [15] Hill PF, Vedi V, Williams A, Iwaki H, Pinskerova V, Freeman MA. Tibiofemoral movement 2: the loaded and unloaded living knee studied by MRI. J Bone Joint Surg Br 2000;82-8:1196 – 8. [16] Rees JL, Price AJ, Beard DJ, Dodd CA, Murray DW. Minimally invasive Oxford unicompartmental knee arthroplasty: functional results at 1 year and the effect of surgical inexperience. Knee 2004;11-5:363 – 7. [17] Pandit H, Beard DJ, Jenkins C, Dodd CA, Murray DW. Oxford unicompartmental knee arthroplasty using a minimally invasive surgical technique—a five year survival study. Hollywood (FL)’ ISAKOS; 2005. [18] Robinson BJ, Rees JL, Price AJ, Beard DJ, Murray DW, Mc Lardy Smith P, et al. Dislocation of the bearing of the Oxford lateral unicompartmental arthroplasty. A radiological assessment. J Bone Joint Surg Br 2002;84-5:653 – 7.
A convex lateral tibial plateau for knee replacement J.V. Bare´, H.S. Gill *, D.J. Beard, D.W. Murray OOEC/Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Oxford, UK Nuffield Orthopaedic Centre, NHS Trust, Oxford, UK Received 4 August 2005; accepted 15 September 2005
Abstract Unicompartmental knee replacements have not performed as well in the lateral compartment as in the medial. This may be because the tibial components have flat or slightly concave surfaces which match the medial plateau but not the convex lateral plateau. The aim of this study was to find the optimal radius for a convex lateral tibial component. Twelve normal lateral tibial plateau were retrieved at knee replacement, and their surface contour in their mid sagittal plane was determined. The optimal circle was fitted and its radius measured. A series of different shaped tibial components were superimposed. From published information about the position of the femoral condyle relative to the tibia in different degrees of flexion, the flexion gap at these angles was determined. The average radius of the lateral tibial plateau was 40 mm. However, as the surface was polyradial it was not clear if this average radius would be optimal. In full flexion, a flat tibial plateau distracted the knee by 8 mm ( p < 0.001). A 75 mm radius spherical tibial plateau did not alter the knee kinematics significantly and gave rise to a change in joint distraction of 1.5 mm. Spherical tibial plateau of 50 mm and 25 mm radii significantly altered knee kinematics ( p < 0.001) and resulted in changes in distraction of 3 mm and 4 mm respectively. The optimal shape for a unicompartmental lateral tibial plateau is likely to be a spherical dome with radius of about 75 mm. The incorporation of this shape in the lateral side of a total knee replacement might improve its flexion. D 2005 Elsevier B.V. All rights reserved. Keywords: Lateral tibial plateau; Knee replacement
1. Introduction Isolated lateral compartment arthritis of the knee is rare and difficult to treat, it occurs much less commonly than isolated medial compartment arthritis [1]. The treatment options include osteotomy, unicompartmental knee replacement (UKR) and total knee replacement, none of which give very satisfactory results [2–6]. Potentially the best solution is unicompartmental replacement which preserves normal anatomy thereby restoring the kinematics of the knee to normal. However, the results of lateral unicompartmental knee replace* Corresponding author. Botnar Research Centre, Nuffield Orthopaedic Centre, Oxford, OX3 7LD, UK. Tel.: +44 1865 227457; fax: +44 1865 227671. E-mail address: [email protected] (H.S. Gill). 0968-0160/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.knee.2005.09.001
ment have been disappointing and have not been as good as medial unicompartmental replacement [7–10]. In the majority of published series the survival at 10 years is about 80% [8,11]. The lateral compartment of the knee is fundamentally different from the medial compartment. This may explain why UKR components that work well in the medial compartment are less appropriate for the lateral, and it may justify the use of different components in the lateral side. The main difference between the medial and lateral compartments is that the medial tibial plateau is slightly concave, whereas the lateral is convex. In addition, the lateral ligaments are much slacker and weaker than the medial, and dynamic stability, provided by the ileotibial tract and popliteus, is important. During flexion there is a small amount of movement of the medial femoral condyle posterior on the tibia. In contrast, on the lateral side the femoral condyle
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126
123
to the long axis of the tibia using either intra or extra medullary instrumentation. The medio-lateral width of the intact tibial plateau was measured. A sagittal saw cut was made in the middle of the lateral tibial plateau. The cut surface was photographed, together with a scale, in order to display the mid-sagittal contour of the lateral tibial plateau (Fig. 2). Care was taken to ensure that the principal axis of the camera was aligned perpendicular to the plane through the midsagittal section of each specimen. 2.2. Image analysis
Fig. 1. MRI view of lateral compartment of the knee at high flexion, from Nakagawa. [Note to Editor: we request permission to show this figure from Nakagawa et al. [12]].
moves so far posteriorly that in full flexion its centre is posterior to the tibial plateau (Fig. 1, from Nakagawa [12]). This posterior displacement of the femur could not be achieved with a unicompartmental replacement that had a flat tibial plateau. Therefore, the only way normal kinematics could be achieved would be with a convex tibial component. The wear rate of polyethylene is related exponentially to the contact pressure [13], which in turn is inversely related to the contact area. With conventional flat designs of fixed bearing unicompartmental replacement polyethylene wear is a problem but usually only in the long term [3,11]. If the bearing surface of the polyethylene was convex the contact area between it and the femoral component would decrease and the contact stress consequently increase, resulting in higher wear. The wear would probably occur early and be catastrophic. Therefore, for a convex tibial surface it will probably be necessary to use a mobile bearing and ideally one which remains in full congruous contact with the tibia. This can be best achieved with a spherical tibial surface, and bearing with a matching spherically concave under-surface. The aim of this study was firstly to determine the radius of the lateral tibial plateau in the sagittal plane, and secondly to investigate how tibial components of varying radii would influence knee kinematics.
Each mid-sagittal plane image was transferred to a workstation and then imported into a drawing software package (CorelDraw version 10, Corel Corporation, Canada). The image was then scaled such that the pixel dimensions matched the actual dimensions of the specimen, so that measurement could be directly made on the image. The approximate sites of contact in full extension and full flexion were identified. A ‘‘best fit’’ circle was superimposed by eye onto the articulating surface of the lateral tibial plateau between these two contact points. The radius of curvature of this circle was measured (Fig. 2). Outlines of four different UKR prosthetic tibial components were superimposed onto each image. The first tibial component was a flat tray, representing the currently used tray for a meniscal bearing design of UKR. The other three were circular in crosssection, representing hypothetical convex components with varying radii of curvature (25, 50 and 75 mm). The bottom surfaces of the circular components were flat, and were placed with a 7posterior slope relative to the original tibial cut. The upper surface of each outline was placed approximately 4 mm below the actual articular cartilage at the point of contact with the femur in full extension (defined below). Four millimetres represented the optimal thickness of a polyethylene meniscal bearing. The lateral femoral condyle was represented by a circle of radius 22 mm. The centre of this femoral condyle circle was placed at four positions relative to the tibia. These were at 7 mm, 8 mm, 21 mm and 27 mm anterior to the posterior edge of the articular cartilage of the lateral tibial plateau, and were representative of the position of the femur in a normal knee at 160-, 130-, 90- and 0- of flexion respectively, as described by Iwaki et al. [14], Hill et al.
2. Materials and methods The shapes of a series of tibial plateau were determined. Then previously published experimentally derived information about the position of the femoral condyle relative to the tibial plateau at various angles of knee flexion was used to determine how the different radii of convex tibial components would influence the flexion gap at various points in the flexion range. 2.1. Specimens Twelve tibial plateau were collected at the time of total knee arthroplasty performed for medial compartment and/or patellofemoral osteoarthritis. Each specimen had an intact lateral tibial plateau with intact cartilage. The tibial cut was made perpendicular
Fig. 2. Excised and sectioned lateral tibial plateau, showing fitted circle to surface.
124
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126
Fig. 3. Method of flexion gap measurement for a) flat tibial component and b) 25 mm radius tibial component.
[15] and Nakagawa et al. [12] in a series of MRI studies on cadaveric and living knees. The femoral condyle circle was positioned so that it just touched the actual articular cartilage of the tibial cross-section. The closest distance from the femoral condyle circle to each tibial prosthetic component surface at each of the four above positions was measured. This distance was representative of the gap that would be present between the femoral condyle and prosthetic tibial component surface (Fig. 3); the flexion gap was termed f. During meniscal bearing lateral UKR surgery, full extension is taken as the reference point at which the thickness of the meniscal bearing is selected. Therefore, any differences between the measurement made in full extension and the other three measurements were indicative of articular surface distraction as the joint flexes, this change was termed Df. An increase in articular surface distraction (positive values of Df) with the bearing inserted would be associated with distraction of the knee in flexion and tightening of the soft tissues. Conversely a decrease in articular surface distraction (negative values of Df) would be associated with slackening of the soft tissues.
average anterior – posterior (AP) length was 43.3 mm (95% CI T 2.7 mm, range 35.5 to 53.7 mm). The mid lateral sagittal radius of curvature of the articular surface averaged 41.4 mm (95% CI T 4.0 mm, range 30.0 to 48.5 mm). The radius was related to the AP length of the plateau, this was a significant correlation with p = 0.011 and r 2 = 0.49. There was no significant correlation between radius and medio-lateral width. It was observed that the surface of the plateau was polyradial not circular, with the radius decreasing posteriorly on the plateau. The median change in articular surface distraction from that in extension, Df, at each flexion angle (0-, 90-, 130- and 160-) is
2.3. Statistical analysis To determine if the various designs of tibial component resulted in a statistically significant change in kinematics throughout the knee flexion range, the values of articular surface distraction, Df, for all flexion angles was analysed for each design using a one-way ANOVA. For zero degrees of knee flexion, Df was zero by definition. The statistical analysis was performed using the Statistics Toolbox version 4.0 available in Matlab 6.5 (The MathWorks Inc., MA, USA).
3. Results The average medio-lateral width of the 12 tibial plateau was 78.2 mm (95% CI T 4.7 mm, range 70.3 to 87.8 mm), and the
Fig. 4. Change in Df over flexion range from 0- to 160- for each design of tibial component, medians and interquartiles are shown. The *** indicates a significant change ( p < 0.001) in mean Df over the flexion range; NS indicates no significant change.
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126 Table 1 Change in Df over flexion range from 0- to 160- and results of ANOVA for Df at 0-, 90-, 130- and 160- knee flexion angles, for each design of tibial component Design
Change in Df over flexion range (mm)
ANOVA results
Flat 75 mm radius 50 mm radius 25 mm radius
9.9 1.5 3.0 3.9
p < 0.0001 p = 0.13 p < 0.001 p < 0.0001
shown for each design of tibial component (flat, 75 mm, 50 mm, 25 mm) in Fig. 4. The overall change in values of Df over the entire flexion range and statistical significance of the changes in articular surface distraction are given in Table 1. The results of the ANOVA showed that the articular surface distraction did not change statistically over the knee flexion range for the 75 mm radius component. However, for all the other designs, the flat, 50 mm and 25 mm radii, the changes in flexion gap (Df) were highly significant ( p < 0.001). For the 75 mm radius component the median change in articular distraction was 1.5 mm over the entire flexion range. The median change in articular surface distraction over the entire flexion range for the 50 mm radius component was 3 mm, the 25 mm radius and flat designs had much larger changes (approximately 4 mm and 10 mm respectively).
4. Discussion The average mid-sagittal radius of the 12 lateral tibial plateau was 41 mm. The radius was related to the AP length; the larger the length, the larger the radius. It was observed that the optimum radius varied from patient to patient and, more importantly, a circle did not fit well. In general, in the portion of the plateau that was in contact with the femur at low knee flexion angles had a larger radius, whereas the portion in contact at higher flexion angles had a smaller radius. Thus, a radius of 41 mm may not be the optimum. A better way to calculate the ideal radius is to determine the influence of different radii on the articular surface distraction at different angles of flexion. It was found that a flat tibial plateau resulted in, on average, 1 mm of articular surface approximation at 90- and 2 mm at 130-, relative to that at extension. This is probably acceptable. However, at 160- a flat plateau resulted in an increase in the articular surface distraction by, on average, 8 mm. Therefore, for the knee to flex fully with a flat component inserted, either the lateral side would have to be distracted by 8 mm or alternatively the kinematics would have to be abnormal, with the lateral side not rolling back normally. It is unlikely that either of these could occur, hence a flat plateau would be expected to block flexion. We have indeed observed per-operatively that a flat plateau does provide a block to flexion. With a tibial plateau of radius 75 mm, the articular surface distraction did not change significantly (ANOVA p = 0.13) over the full flexion range and the average decrease in articular surface distraction was less than 2
125
mm. It would therefore seem that with a 75 mm radius, the knee should function with virtually normal kinematics. With the 50 mm radius although the flexion gap changed significantly over the full range of flexion the change was only 3 mm. It is therefore likely that tibial plateau with a radius in the region of 50 mm would also probably be satisfactory. A 25 mm radius plateau would not be satisfactory as it results in a significant change in flexion gap of approximately 4 mm. There was considerable variation between patients in the articular surface distraction at different angles of flexion. It would therefore be necessary to make small adjustments to the component position to achieve optimal kinematics in all patients. For example, altering the tibial slope would influence the articular surface distraction, as would altering the position of the femoral component. However, alterations to component position would probably not accommodate the large abnormalities in articular surface distraction that occur with a flat or 25 mm radius component. For the flat tibial plateau the median changes in flexion gap were 2 mm up to 130- flexion and only became large above 130-. Therefore for traditional open lateral UKR which has an average range of motion (ROM) of 110- [8] a flat plateau is probably acceptable. However, evidence from the medial compartment suggests that with a minimally invasive (MIS) approach the ROM increases from about 110- to 130[16]. We would therefore expect that the average ROM with a minimally invasive lateral UKR also to be about 130-, and that some could potentially flex considerably more. These would be expected to have kinematic problems with a flat tibial plateau. It is therefore not surprising that the average ROM of MIS lateral UKR is actually 120- [17] and not 130-. We believe that if a 75 mm radius component was used higher flexion would be achieved. For a mobile bearing to achieve fully congruous contact in all positions and thus low wear, the sagittal plane shape of the tibial plateau should be circular. However, the three dimensional shape of the plateau could be cylindrical or spherical. Lateral UKR with unconstrained mobile bearings have had a 10% dislocation rate [18], with the majority of dislocations being medial into the intercondylar notch. A major advantage of a spherical plateau is that it should decrease the dislocation rate. This is because a spherically convex plateau with a corresponding concave bearing would substantially increase entrapment in a medio-lateral direction. It is appreciated this study has a number of limitations. In particular the position of the femoral component in different angles of flexion was determined from published average data. Substantial differences exist in rollback between patients. Although this would not influence the conclusions about the relative merits of different shaped plateau it could influence the angle of flexion at which the different plateau would cause problems. For example, patients with substantial early rollback would be likely to encounter problems with a flat plateau at much lower flexion angles than those predicted by this study. Further cadaveric and
126
J.V. Bare´ et al. / The Knee 13 (2006) 122 – 126
clinical studies need to be undertaken. In these the influence of flat and 75 mm radius tibial plateau should be compared. Total knee replacements (TKR) have in general not achieved excellent flexion. Recently, some new TKRs have been designed to achieve high flexion, but even these have often failed to achieve full flexion, or if they do they have had problems. We believe this is in part because the lateral femoral condyle of the normal knee subluxes off the back of the tibial plateau in high flexion and current designs of TKR do not allow this. If a circular lateral tibial plateau and lateral mobile bearing were incorporated in a TKR this might allow the TKR to achieve high flexion.
5. Conclusions The average radius in the sagittal plane of the lateral tibial plateau is about 40 mm. The tibial plateau, however, is polyradial so 40 mm may not be the ideal radius. The optimum radius of a lateral tibial component is probably about 75 mm as this radius causes the smallest change in articular surface distraction.
References [1] Altman RD, Fries JF, Bloch DA, Carstens J, Cooke TD, Genant H, et al. Radiographic assessment of progression in osteoarthritis. Arthritis Rheum 1987;30-11:1214 – 25. [2] Coventry MB. Proximal tibial varus osteotomy for osteoarthritis of the lateral compartment of the knee. J Bone Joint Surg Am 1987;691:32 – 8. [3] Witvoet J, Peyrache MD, Nizard R. Single-compartment ‘‘Lotus’’ type knee prosthesis in the treatment of lateralized gonarthrosis: results in 135 cases with a mean follow-up of 4.6 years. Rev Chir Orthop Reparatrice Appar Mot 1993;79-7:565 – 76. [4] Cameron HU, Botsford DJ, Park YS. Prognostic factors in the outcome of supracondylar femoral osteotomy for lateral compartment osteoarthritis of the knee. Can J Surg 1997;40-2:114 – 8.
[5] Mathews J, Cobb AG, Richardson S, Bentley G. Distal femoral osteotomy for lateral compartment osteoarthritis of the knee. Orthopedics 1998;21-4:437 – 40. [6] Newman J, Ackroyd C, Shah N. Unicompartmental or total knee replacement? Five-year results of a prospective, randomised trial of 102 osteoarthritic knees with unicompartmental arthritis. J Bone Joint Surg Br 1998;50:862 – 5. [7] Goodfellow JW, Kershaw CJ, Benson M.K, O’Connor JJ. The Oxford Knee for unicompartmental osteoarthritis. The first 103 cases. J Bone Joint Surg Br 1988;70-5:692 – 701. [8] Gunther TV, Murray DM, Miller R, Wallace DA, Carr AJ, O’Connor JJ, et al. Lateral compartment arthroplasty with the Oxford Meniscal knee. Knee 1996;3:33 – 9. [9] Murray DW, Goodfellow JW, O’Connor JJ. The Oxford medial unicompartmental arthroplasty: a ten-year survival study. J Bone Joint Surg Br 1998;80-6:983 – 9. [10] Svard UC, Price AJ. Oxford medial unicompartmental knee arthroplasty. A survival analysis of an independent series. J Bone Joint Surg Br 2001;83-2:191 – 4. [11] Newman JH. Unicompartmental knee replacement. Knee 2000;7: 63 – 70. [12] Nakagawa S, Kadoya Y, Todo S, Kobayashi A, Sakamoto H, Freeman MA, et al. Tibiofemoral movement 3: full flexion in the living knee studied by MRI. J Bone Joint Surg Br 2000;82-8:1199 – 200. [13] Rostoker W, Galante JO. Contact pressure dependence of wear rates of ultra high molecular weight polyethylene. J Biomed Mater Res 1979;13:957 – 64. [14] Iwaki H, Pinskerova V, Freeman MA. Tibiofemoral movement 1: the shapes and relative movements of the femur and tibia in the unloaded cadaver knee. J Bone Joint Surg Br 2000;82-8:1189 – 95. [15] Hill PF, Vedi V, Williams A, Iwaki H, Pinskerova V, Freeman MA. Tibiofemoral movement 2: the loaded and unloaded living knee studied by MRI. J Bone Joint Surg Br 2000;82-8:1196 – 8. [16] Rees JL, Price AJ, Beard DJ, Dodd CA, Murray DW. Minimally invasive Oxford unicompartmental knee arthroplasty: functional results at 1 year and the effect of surgical inexperience. Knee 2004;11-5:363 – 7. [17] Pandit H, Beard DJ, Jenkins C, Dodd CA, Murray DW. Oxford unicompartmental knee arthroplasty using a minimally invasive surgical technique—a five year survival study. Hollywood (FL)’ ISAKOS; 2005. [18] Robinson BJ, Rees JL, Price AJ, Beard DJ, Murray DW, Mc Lardy Smith P, et al. Dislocation of the bearing of the Oxford lateral unicompartmental arthroplasty. A radiological assessment. J Bone Joint Surg Br 2002;84-5:653 – 7.