and Screw-Home in a Mobile-Bearing Total Knee Arthroplasty

Transcription

1 The Journal of Arthroplasty Vol. 14 No In Vivo Determination of Condylar Lift-off and Screw-Home in a Mobile-Bearing Total Knee Arthroplasty James B. Stiehl, MD,* Douglas A. Dennis, MD, t Richard D. Komistek, PhD,-~ and Hal S. Crane, MDt Abstract: Twenty subjects implanted with the low-contact stress (LCS) cruciatesacrificing, mobile-bearing total knee arthroplasty underwent dynamic videofluoroscopy during in vivo weight-bearing conditions using a 3-dimensional computer-aided design (CAD) interactive modeling method. Ninety percent of the subjects demonstrated significant lift-off during stance phase of gait. Condylar lift-off was present at both the medial and the lateral condyles. The maximal medial lift-off was 2.12 mm, whereas the greatest lateral lift-off was 3.53 ram. The maximal positive screw-home was 9.6, whereas the maximal negative or reverse screw-home was 6.2. The average screw-home rotation was positive 0.5. In 50% of patients, medial condylar translation was unexpectedly greater than lateral condylar motion. Condylar lift-off and screw-home motion are significant kinematic functions in this rotationally unconstrained total condylar knee arthroplasty. Key words: total knee arthroplasty, kinematics, condylar lift-off, screw-home rotation. In total knee arthroplasty (TKA), knowledge of kinematic function is necessary if the in vivo weightbearing forces and shear stresses applied to the bearing surfaces are to be determined. Posterior cruciate-retaining designs tend to have lower conformity to allow greater rotational freedom, believed to be necessary for normal function and to lower stresses at the implant-bone interface. Posterior cruciate-sacrificing implants have higher conformity, which increases implant stability and improves wear characteristics. From a review of the literature, we could find no clear advantage of one technique or implant, although certain posterior cruciate-retaining designs have shown a dis- From the *Midwest Orthopaedic Biomechanical Laboratory, St. Luke's Hospital, Milwaukee, Wisconsin; and ~-Rose Musculoskeletal Research Laboratory, Rose Medical Center, Denver, Colorado. Submitted March 15, 1998; accepted September 1 l, Reprint requests: James B. StiehI, MD, 2015 E. Newport, #703, Milwaukee, WI Copyright 1999 by Churchill Livingstone /99/ /0 turbing incidence of osteolysis not recognized with older posterior cruciate-sacrificing total condylar designs [1-5]. Blunn et al. [6] demonstrated pattern wear and peripheral wear in designs with polyethylene delamination, known to result from the most severe wear. The potential for femoral-tibial separation or condylar lift-off during weight-bearing has been postulated by several authors. Dennis eta]. [7] were able to demonstrate condylar lift-off in both posterior cruciate-retaining and posterior cruciatesacrificing designs using in vivo dynamic videofluoroscopy. Nilsson et al. [8] have also shown this phenomenon using stereoradiography. The clinical implication of this finding is that edge loading at the peripheral surface of the tibia] plateau may be deleterious, particularly if a Jlat-on-flat condylar design is used [9,10]. In vivo rotational movement in TKA has been investigated using several methods, including roentgen stereophotogrammetry, videofluoroscopy, and electromagnetic orthotic fixtures [8,11,I2]. These 293

2 294 The Journal of Arthroplasty Vol, 14 No. 3 April 1999 studies have typically shown screw-home or external rotation of the tibia in extension with internal rotation as the angle of flexion increases. Alterations from the rotation of the normal knee may be related to anterior cruciate deficiency, prosthetic geometry, and differences in surgical technique in individual patients. Knowledge of rotational movement is an important consideration for understanding polyethylene wear patterns, in which exaggerated sliding motion may produce detrimental delamination wear. Dynamic videofluoroscopy has emerged as a valuable scientific tool for investigating in vivo kinematic performance of TKA. Our initial experience with the technique allowed the determination of 1-point contact in the sagittal plane, such as the lateral femoral condyle with the tibial plateau {13]. More recently, we have used 3-dimensional model fitting to determine kinematic relationships accurately. With this method, determination of both medial and lateral condylar position as well as condylar lift-off and screw-home rotation can be measured in gait [ 14]. The purpose of this paper is to use in vivo fluoroscopy with an interactive model-fitting technique to examine coronal and transverse plane motion in a mobile-bearing TKA during gait. The results are compared with our prior experience with fixed bearing designs and with information on normal knees from literature review. The low-contact stress (LCS) rotating platform prosthesis (Depuy, Inc, Warsaw, IN) is a posterior cruciate-sacrificing implant that was designed to have unconstrained rotational freedom and conforming coronal plane articulation that allowed condylar lift-off. Materials and Methods Before inclusion in this study, each patient reviewed and signed an investigation review boardapproved consent form. We performed dynamic fluoroscopy under weight-bearing conditions in 20 patients with a cemented posterior cruciate-sacrificing LCS mobile-bearing (rotating platform) TKA (Deputy, Warsaw, IN). The patients were chosen on the basis of an excellent clinical result (>90/90) using the Knee Society scoring system. The time from surgery to analysis was a minimum of 12 months in all cases. The resultant femoral-tibial alignment was normal in all (range, 5o-7 ) with excellent stability in extension. Each patient underwent 2-dimensional videofluoroscopy using a VJ Works fluoroscopy unit (VF Works, Palm Harbor, FL), which produces images at a rate of 30 Hz. The technique required the patient to take 1 step while walking on an elevated platform. The radiology technician then followed the knee joint by attempting to keep the lateral side of the knee centered on the x-ray machine's fluoroscopic image at all times. This was done to simulate the walking gait cycle from heel-strike to toe-off. Video Analysis We have evolved our video analysis to an interactive model-fitting technique. This method fits 3-dimensional computer-aided design (CAD) solid models of the femoral and tibial implants onto the 2-dimensional fluoroscopic silhouette images. The fluoroscopic images had been stored on videotapes for subsequent redigitization using a frame grabber. The videos were then analyzed on a computer workstation using the interactive computer algorithm (Fig. 1). The femoral and tibial components of the best-fit overlay were rotated into the precise sagittal plane to measure anteroposterior contact of the medial and lateral condyles and screw-home rotation, which is a function of these points. Four distinct positions of the sagittal plane fluoroscopy images were analyzed during the normal gait cycle including i) at heel-strike, ii) at 33% of stance phase, iii) at 66% of stance phase, and 4) at toe-off. These positions were confirmed using a second video camera to determine the exact frame of heel-strike and toe-off and calculating 33 % and 66 % of the weight-bearing gait cycle. The overlay model was then rotated into the frontal plane to measure the distance from the femoral condyle to the tibial plateau to determine femoral-tibial lift-off. Measurement of condylar liftoff is done by comparing the difference of the medial and lateral condylar distances to the tibial baseplate. It is assumed that the condyles are in a similar plane of tibial contact, and there should be little difference in plastic thickness. Rotation was determined from an arbitrary reference line that was perpendicular to the sagittaltibial plane. Medial and lateral condylar translation at each point in the gait cycle were used to calculate rotation. Determination of the specific medial and lateral condyle has been easy for several reasons. First, the leg involved is known, and therefore medial and lateral overlay prosthetic condyles are known. Second, most implants are asymmetric, and condylar shapes become obvious. Third, for symmetric implants, the lateral condyle is always closer to the x-ray tube and is larger. For the anteroposterior reference, a positive reference is denoted as anterior and a negative posterior to the sagittal plane midline of the tibial prosthesis. Normal or positive screw-home motion was defined as tibial internal rotation in relation to the distal femur with increasing flexion [15]. Femoral-tibial

3 Condylar Liftoff and Screw Home in Mobile-Bearing TKA Stiehl et al. 295 Fig. 1. Interactive 3-dimensional CAD modeling showing (A) CAD model overlay superimposed on 2-dimensional videofluoroscopy frame. (B) Example of sagittal view of solid model. (C) Example of coronal view of solid model. (D) Coronal view demonstrating condylar lift-off. A B C D contact of the medial and lateral condyles could have 3 patterns of translation to cause this motion: i) Lateral condyle moves more posterior than the medial; ii) lateral condyle moves posterior, whereas the medial condyle moves anterior; and iii) lateral condyle moves less anterior than the medial condyle. Reverse or negative screw-home was defined as tibial external rotation in relation to the femur with increasing flexion. Femoral-tibial contact of the medial and lateral condyles could have 3 patterns of translation to cause the motion: i) medial condyle moves more posterior than the lateral; ii) medial condyle moves posterior, whereas the lateral moves anterior; and iii) medial condyle moves less anterior than the lateral condyle. Error Analysis An error analysis was performed to determine the reproducibility and accuracy of measurement of our technique. This analysis was done by fluoroscoping implant components mounted on a 6 of freedom apparatus. Accurate positioning of the components was achieved using rotational and translational stages with an accuracy of 15 arc seconds and 0.01 mm. The components were set in an initial position, then rotated and translated to known values. Fluoroscopic images of the components were created at each setting. The 3-dimensional model-fitting process was performed for each setting of the rotational and translational stages to determine the relative pose of the components. A second dynamic test was performed to determine the effect of motion. The components were pulled through the fluoroscopic scene at a variable speed between 0.5 and 1.0 feet per second. The translational and rotational 3-dimensional model-fitting technique was accurate to 0.5 mm and 0.5 [14]. A threshold of 0.75 mm and 0.75 (50% safety factor) was used to account for unknown variables. Condylar Lift-off Results Significant condylar lift-off was seen in 90% of subjects at heel-strike and 66% of subjects at stance phase and toe-off. In 50% of TKAs, we found both medial and lateral lift-off, whereas only 15% showed nonsignificant lift-off (<0.75 mm). The greatest medial lift-off was 2.12 mm, whereas the greatest lateral was 3.53 mm (Table 1). Screw-Home Rotation Seven of 20 patients demonstrated positive screwhome rotation with tibial internal rotation on flexion. In 5 subjects, the overall rotation was minimal or insignificant (< 0.75). Eight patients showed negative screw-home with tibial external rotation on

4 296 The Journal of ArthroplastyVol. 14 No. 3April 1999 flexion. Six patients demonstrated greater than 5 rotation with the maximal positive screw-home of 9.6, whereas the maximal negative screw-home was negative 6.2. The average screw-home for the group was a positive 0.5. We could not find a relationship between condylar lift-off and a specific pattern of screw-home because condylar lik-off was seen with all screw-home possibilities (Table 2; Fig. 2). Normal screw-home with medial condylar pivot and the medial condyle moving less posterior than the lateral condyle with gait was seen in only 2 patients. Five patients demonstrated abnormal positive screw-home, with the medial condyle moving more anterior than the lateral in 4 and the medial condyle moving anterior, whereas the lateral condyle moved posterior in 1 (Table 3). Reverse screw-home was seen in 13 patients. In 5 subjects, the medial condyle translated more posterior than the lateral. For 3 knees, the medial condyle translated posterior, whereas the lateral moved anterior. In 5 cases, the medial condyle translated less anterior than the lateral. Discussion Dynamic fluoroscopy has proved to be invaluable for investigating kinematic performance of TKA. Using computer vector analysis, we have previously identified abnormal anteroposterior translation, loss Table 1. Lift-Off Results* % of Gait Subject 0% 33 % 66% 100% I Average *Negative = medial condyle; positive = lateral condyle. Table 2. Screw-Home Results During 0%, 33%, 66%, and 100% Stance Phase of Gait Cycle* Subject 0%-33% 33%-66% 66%-100% 0%-100% I Average *Positive = tibia internal rotation; negative = tibia external rotation. of posterior femoral rollback, and lateral condylar lift-off in posterior cruciate-retaining TKAs [13]. Evolution of our technique led to an inverse perspective method that required developing a library of CAD implant models that could be matched to each isolated femoral and tibial implant on the video frame [14]. More recently, we have used an interactive modeling system that allows us to manipulate the 3-dimensional CAD model implant precisely onto the video 2-dimensional image. This method does not require precise sagittal plane positioning of the patient's knee because the video image can accurately be superimposed even with some degree of malrotation. The reason for this is that the complex geometry of the implant image allows for only 1 finite position of the CAD model in space. The fluoroscopic image can then be extracted leaving the CAD image from which kinematic calculations are possible. We have been able to demonstrate that condylar lift-off occurs in all TKAs that we evaluated regardless of surgical technique, including posterior cruciate retention, sacrifice, or substitution. Most of the mobile-bearing TKAs of this study showed both medial and lateral condylar lilt-off using a posterior cruciate-sacrificing technique. The largest amount of lift-off was identified in the lateral compartment. Dennis et al. [7] evaluated posterior cruciatesubstituting TKAs, finding that medial condylar lift-off occurred about equal to lateral lift-off. In

5 Condylar Liftoff and Screw Home in Mobile-Bearing TKA Stiehl et al Subject 1 Subject 2 Subject 3 Subject 4 Subject 5 Fig. 2. Five randomly selected patients demonstrating diverse distribution of rotation / Heel-strike / 33%Stance 66%Stance Toe-off % of Stance Phase posterior cruciate-retaining TKAs, condylar lift-off was predominantly lateral. Using a different method, roentgenographic stereophotogrammetry, Nilsson et al. [8] were also able to demonstrate this phenomenon, which they defined as tibial rotation (abduction/adduction) about the sagittal axis. In their study, the LCS mensical-bearing implant revealed a mean of 3 adduction at 50 flexion, similar to normal knees. The fact that condylar lift-off occurs is not surprising considering the adduction and abduction moments that have been hypothesized with gait {16]. Condylar lift-off becomes a significant issue for TKA design when one considers peripheral edge loading that is likely with fiat-on-fiat total condylar designs [9,10]. Wasielewski et al. [17] and Blunn et al. [6] found peripheral wear and pattern wear with severe polyethylene delamination in these designs. Because abnormal medial compartment anteroposterior translation is compounded with lateral condylar lift-off, the posterior medial wear patterns identi- Table 3. Screw-Home Mechanism Internal Tibial No. Rotation Type Condylar Motion Patients + Normal Med Post < Lat Post 2 - Reverse Med Post > Lat Post 5 - Reverse Med Post:Lat Ant 3 + Normal Med Ant:Lat Post 1 + Normal Med Ant:Lat Ant 4 Reverse Med Ant < Lat Ant 5 Med, medial; Post, posterior; Lat, lateral; Ant, anterior. fled in those studies are expected. The coronal geometry of the LCS rotating platform implant used in this study is Conforming in the coronal plane to allow for congruity with lift-off. Retrievals of this implant have revealed minimal wear of the polyethylene surface after extended clinical use [ 18]. High rotational constraint in TKA has been recognized as a predominant cause of failure in early hinge designs. Rotational movement of the normal knee was studied by LaFortune et al. [19] using high-speed photography and implanted Steinman pins as skeletal markers in normal walking volunteers. Internal rotation at heel-strike and toe-off measured slightly less than 5, whereas external rotation increased to 9 during swing phase. Screwhome movement has been described as relative external rotation of the tibia in relation to the femur near full extension [20]. Early investigators postulated that abnormalities of knee function are related to disturbance of this mechanism. Karrholm et al. [21] demonstrated significant alterations in rota~ tion with anterior cruciate-deficient knees finding a more externally rotated tibia in extension followed by decreased internal rotation in flexion. Nilsson et al. [8] reported similar findings with several different posterior cruciate-retaining TKAs, noting less terminal screw-home or terminal external tibial rotation, again relating to the more externally rotated tibia in extension and decreased internal rotation in flexion. Nilsson et al. [9] investigated the LCS meniscalbearing total knee implant, finding that initial extension started with a more externally rotated tibia than normal and had minimal internal rotation

6 298 The Journal ofarthroplasty Vol. 14 No. 3April 1999 (mean, 0.5 ) during flexion. Using the LCS rotating platform implant, which sacrifices the posterior cruciate ligament, our study found similar overall internal rotation (mean, 0.5 ) during flexion, but our patients demonstrated much greater variability with rotation. For example, only 7 of our patients demonstrated internal tibial rotation with flexion. Of that group, only 2 patients had normal screwhome with lateral femoral tibial contact moving more posterior than medial contact on flexion. The other 5 had significant anterior sliding of the medial condyle to cause internal rotation. Reverse screwhome was seen in 13 of our patients, in whom there was actually tibial external rotation in flexion. This reverse screw-home resulted from exaggerated lateral condylar anterior translation in flexion or medial condylar posterior translation that exceeded lateral translation. Two patients in our study demonstrated greater than 9 of tibial internal rotation with gait. Other authors have been able to demonstrate significant tibial rotation of unconstrained TKAs. E1 Nahass et al. [11] evaluated the kinematic condylar posterior cruciate-retaining TKA using electrogoniometers, finding that tibial external rotation on extension varied from 4.4 to 11.3% Markovich et al. [121 used in vivo videofluoroscopy to evaluate weight-bearing step-up activity finding 8 of tibial external rotation from flexion to extension. This rotation occurred from posterior translation of the medial femoral condyle in extension (average, 6.3 mm), whereas lateral condylar translation was limited. We have shown that significant rotation and condylar lift-off were present in a cohort of successfully performed TKAs. The results of our study suggest that a complex rotational motion may exist in many with regard to femoral-tibial contact of the condyles in TKA. Transverse plane rotations and lift-off for each individual patient were highly variable, reflecting the inability to restore perfectly the normal kinematics of the joint. We believe that these kinematic abnormalities, when exaggerated, can be related to some of the current problems of TKA, such as peripheral pattern and posterior medial condylar wear. Further studies are necessary to investigate the complex relationship of biomechanical and biomaterial performance in TKA. These data suggest that designs such as the LCS mobilebearing platform, which accommodate significant rotation and lift-off while maintaining high articular surface congruity, may importantly diminish contract surface stresses. This may represent an optimal solution to these unresolved kinematic derangements. References 1. Malkani AL, Rand JA, Bryan RS, Wallrichs SL: Total knee arthroplasty with the kinematic condylar prosthesis. J Bone Joint Surg Am 77:423, Ranawat CS, Flynn WE Saddler S, et al: Long-term results of the total condylar knee arthroplasty. Clin Orthop 286:94, Stern SH, Insall JN: Posterior stabilized prosthesis: results after follow-up of nine to twelve years. J Bone Joint Surg Am 74:980, Kim YH, Oh JH, Oh SH: Osteolysis around cementless porous-coated anatomic knee prosthesis. J Bone Joint Surg Br 77:236, Becker M, Insall JN, Faris PM: Bilateral total knee arthroplasty: one cruciate retaining and one cruciate substituting. Clin Orthop 271:122, Blunn GW, Joshi AB, Minns R J, et al: Wear in retrieved condylar knee arthroplasties. J Arthroplasty 12:281, Dennis DA, Komistek RD, Cheal E J, Stiehl JB: A determination of condylar lift-off using fluoroscopy. Proceedings 43rd Annual Orthopaedic Research Society Meeting, San Francisco, 1997, p Nilsson KG, Karrholm J, Gadegaard P: Abnormal kinematics of the artificial knee: roentgen stereophotogrammetric analysis of 10 Miller-Galante and five New Jersey LCS knees. Acta Orthop Scand 62:440, Lewis P, Rorabeck CH, Bourne RB, Devane P: Posteromedial tibial polyethylene failure in total knee replacements. Clin Orthop 299:1 l, Feng EL, Stulberg DS, Wixson RS: Progressive subluxation and polyethylene wear in total knee replacements with flat articular surfaces. Clin Orthop 299: 60, 1993 I 1. E1Nahass E, Madson MM, Walker PS: Motion of knee after condylar resurfacing. Proceedings 36th Orthopaedic Research Society Annual Meeting, New Orleans, 1990, p Markovich GD, Banks SA, Hodge WA: Comparison of passive and active knee replacement kinematics. Proceedings 43rd Orthopaedic Research Annual Meeting, San Francisco, 1997, p Stiehl JB, Komistek RD, Dennis DA, et al: Fluoroscopic analysis of kinematics after posterior-cruciateretaining total knee arthroplasty. J Bone Joint Surg Br 77:884, Dennis DA, Komistek RD, Hoff WA, Gabriel SM: In vivo kinematics derived using an inverse perspective technique. Clin Orthop 331:107, Muller W: The knee: form, function and ligament reconstruction. Springer-Verlag, Berlin, Andriacchi TP, Galante JO: Retention of the posterior cruciate in total knee arthroplasty. J Arthroplasty (Suppl) 3:13, Wasielewski RC, Galante JO, Leighty RM, et al: Wear patterns on retrieved polyethylene tibial inserts and

7 Condylar Liftoff and Screw Home in Mobile-Bearing TKA Stiehl et al. 299 their relationship to technical considerations during total knee arthroplasty. Clin Orthop 299:31, Collier JP, Mayor MB, McNamara JL, et al: Analysis of the failure of 122 polyethylene inserts from uncemented tibial knee components. Clin Orthop 273: 232, LaFortune MA, Cavanagh PR, Sommer H J, Kalenak A: Three-dimensional kinematics of the human knee during walking. J Biomech 25:347, Hallen LG, Kindahl O: The "screw-home" movement in the knee joint. Acta Orthop Scand 37:97, Karrholm J, Selvik G, Elmlqvist LG, Hansson LI: Active knee motion after cruciate ligament rupture: stereoradiography. Acta Orthop Scand 59:158, 1988