- Email: [email protected]
Intraoperative joint gaps and mediolateral balance affect postoperative knee kinematics in posterior-stabilized total knee arthroplasty
The Knee 22 (2015) 527–534
Contents lists available at ScienceDirect
The Knee
Intraoperative joint gaps and mediolateral balance affect postoperative knee kinematics in posterior-stabilized total knee arthroplasty☆ Toshifumi Watanabe a,b,c,⁎, Takeshi Muneta c, Ichiro Sekiya c, Scott A. Banks a a b c
Department of Mechanical & Aerospace Engineering, University of Florida, Gainesville, FL, USA Department of Orthopaedic Surgery, Tsuchiura Kyodo General Hospital, Tsuchiura, Ibaraki, Japan Department of Orthopaedic Surgery, Tokyo Medical and Dental University Hospital, Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 26 September 2014 Received in revised form 22 January 2015 Accepted 17 March 2015 Keywords: Intraoperative joint gap Mediolateral balance Postoperative knee kinematics Posterior-stabilized Total knee arthroplasty
a b s t r a c t Background: Adjusting joint gaps and establishing mediolateral (ML) soft tissue balance are considered essential interventions for better outcomes in total knee arthroplasty (TKA). However, the relationship between intraoperative laxity measurements and weightbearing knee kinematics has not been well explored. This study aimed to quantify the effect of intraoperative joint gaps and ML soft tissue balance on postoperative knee kinematics in posterior-stabilized (PS)-TKA. Methods: We investigated 44 knees in 34 patients who underwent primary PS-TKA by a single surgeon. The central joint gaps and ML tilting angles at 0°, 10°, 30°, 60°, 90°, 120° and 135° flexion were measured during surgery. At a minimum of two year follow-up, we analyzed in vivo kinematics of these knees and examined the influence of intraoperative measurements on postoperative kinematics. Results: Gap difference of knee flexion at 135° minus 0° was correlated with the total posterior translation of lateral femoral condyle (r = 0.336, p = 0.042) and femoral external rotation (r = 0.488, p = 0.002) during squatting, anteroposterior position of lateral femoral condyle (r = ‐ 0.510, p = 0.001) and maximum knee flexion (r = 0.355, p = 0.031) in kneeling. Similar correlations were observed between deep flexion gap differences with respect to the 90° reference and postoperative knee kinematics. Well-balanced knees showed less anterior translation of medial femoral condyle in mid- to deep flexion, consistent femoral external rotation, and the most neutral valgus/varus rotation compared with unbalanced knees. Conclusion: These findings indicate the importance of adequate intraoperative joint gaps in deep flexion and ML soft tissue balance throughout the range of motion. Level of evidence: Level II. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Establishing proper intraoperative joint gaps and mediolateral (ML) soft tissue balance are considered essential parts of surgery to achieve the best possible outcomes after total knee arthroplasty (TKA) [1,2]. Various devices have been developed to measure these gaps and ML balance, helping surgeons quantify the procedure [3–8]. These devices have been used to quantify detailed joint gap patterns and have been used, for example, to show the patellofemoral (PF) joint should be reduced during measurements to more closely mimic normal knee mechanics [5,8]. Despite the growing popularity of intraoperative gap measurements, there have been few reports relating intraoperative ☆ This work was mainly performed at Tsuchiura Kyodo General Hospital, Tsuchiura, Ibaraki, Japan and at the University of Florida, Gainesville, FL, USA. ⁎ Corresponding author at: Tokyo Medical and Dental University Hospital, Department of Orthopaedic Surgery, 1-5-45, Yushima, Bunkyo-ku, 113-8510 Tokyo, Japan. Tel./fax: +81 3 5803 4020. E-mail addresses: [email protected] (T. Watanabe), [email protected] (T. Muneta), [email protected] (I. Sekiya), banks@ufl.edu (S.A. Banks).
http://dx.doi.org/10.1016/j.knee.2015.03.006 0968-0160/© 2015 Elsevier B.V. All rights reserved.
joint gaps and balance to weight-bearing knee kinematics postoperatively [9]. Previous studies of TKA utilizing 3D–2D model-image registration techniques have clarified in vivo kinematics of knees with various arthroplasty designs [10–13]. Many studies have reported that both medial and lateral femoral condyles generally translate posteriorly with flexion, with greater lateral condylar posterior translation coupled to femoral external rotation [10–14]. Several studies have reported greater posterior femoral translation has a positive correlation with greater maximum knee flexion [14,15]. However, how intraoperative joint gaps and ML soft tissue balance affect weight-bearing knee kinematics has not been well investigated. One of a surgeon's best opportunities to predict and affect postoperative kinematics is during surgery, so it is important to establish relationships between intraoperative gaps and balance, and postoperative knee kinematics. The goal of this study was to establish how intraoperative joint gaps and ML soft tissue balance affect postoperative kinematics in posteriorstabilized (PS)-TKA. We hypothesized that knees with larger joint gap differences in deep flexion with respect to zero degrees or 90° reference
528
T. Watanabe et al. / The Knee 22 (2015) 527–534
and good ML balance throughout the range of motion would exhibit greater femoral posterior translation, femoral external rotation, and better maximum knee flexion, because adequate joint laxity and balanced ML soft tissue generally should provide more natural knee motions [1,2,16,17]. 2. Materials and methods 2.1. Subjects We studied 44 knees in 34 patients who underwent primary PS-TKA between January 2006 and August 2008. A single surgeon (TW) performed all surgeries using generally accepted techniques for minimally invasive surgery (MIS) and the same prosthesis (NexGen LPS Flex, Zimmer, Warsaw, Indiana, USA). Joint gaps and ML soft tissue balance were recorded during each case. The patients were studied at a minimum of two years follow-up. Knees with previous high tibial osteotomy and knees with severe deformity requiring metal block augmentation were excluded. All patients gave informed consent to participate in this institutional review board approved study. Subjects averaged 71 ± 7 years (average ± SD) at the time of surgery. The study cohort included 36 female and eight male knees with a preoperative diagnosis of osteoarthritis (OA) in 38 knees and rheumatoid arthritis (RA) in six knees. The body mass index of the patients averaged 26 ± 5 kg/m2. Two knees had valgus alignment with femorotibial angle less than 170° and the other knees had neutral or varus alignment (Table 1). The follow-up period after surgery averaged 2.4 ± 0.7 years. A sample of 44 knees was computed to produce 80% power (1-b) for correlating intraoperative joint gaps and postoperative knee kinematics using an effect size of 0.4, or an r2 value explaining 16% of the data variance. 2.2. Surgical procedure An air tourniquet was inflated to 330 mm Hg during surgery. The mini-midvastus approach with independent bone cutting technique was used for all knees. The distal femur was cut perpendicular to its mechanical axis, removing the amount of bone corresponding to the prosthetic femoral component thickness. The proximal tibia was cut perpendicular to its coronal mechanical axis and with approximately six degrees sagittal posterior tibial slope, removing bone on the intact side corresponding to the prosthetic tibial component thickness. The posterior femoral condyles were cut with approximately three degrees external rotation from the posterior condylar line. Whiteside's line and epicondilar line were used as a femoral rotation reference in valgus knees. Following the bone cutting, the soft tissues were released on a caseby-case basis to obtain ML balance. The posterior cruciate ligament (PCL) was sacrificed in all knees. The deep layer of the medial collateral ligament (MCL) was released in all knees with varus deformity. After osteophyte removal, the superficial layer of the MCL, posterior capsule, semimembranosus, and pes anserinus were released sequentially until adequate ML soft tissue balance was obtained. No soft tissue release of lateral structures was required to obtain acceptable ML soft tissue balance. 2.3. Intra-operative measurements Joint gaps and ML soft tissue balance were measured after performing soft tissue releases. The femoral trial was put in place and a tensor (Offset
Repo-Tensor; Zimmer, Tokyo, Japan) developed by Kobe University [5] was placed between the tibial cut surface and the femoral trial. The tensor consists of three parts: an upper seesaw plate, a lower platform plate, and an extraarticular main body (Fig. 1a, b). The lower platform plate was fixed on the center of the tibial cut surface and held in place with small pins protruding from the bottom side of the plate. The seesaw plate has a post that fits into the intercondylar space and articulates with the cam of the femoral trial. This post-cam mechanism controls the tibiofemoral translation in both the coronal and sagittal planes over the entire arc of knee flexion. The main body connects the other two parts. The tensor is small enough to be applied for MIS procedures and the offset arm allows surgeons to use the device with the patella reduced. The tensor provides numerical measures of joint gap at the center of the knee and the ML tilting angle from full extension to deep flexion. The patella was reduced to the original position and several sutures were used to maintain correct PF position throughout the range of motion (Fig. 1c). A joint distraction force of 40 lb (18 kg) was applied by the tensor, and the central joint gaps and ML tilting angles were measured at zero degrees, 10°, 30°, 60°, 90°, 120° and 135° flexion. Preliminary studies with this tensor indicated the joint gap at full extension with 40 lb of joint distraction force corresponded most closely to the appropriate tibial insert thickness [5]. An assistant supported the thigh during measurement to maintain appropriate sagittal alignment and to reduce the influence of the thigh mass on knee forces. The accuracy of this measurement has been estimated to be ± 0.3 mm in the joint gap [5]. Gap data for 10° and 135° flexion positions are available for 39 and 37 knees, respectively, as these increments were not tested in the first five to seven knees. We defined the “gap difference” as a gap size difference between one gap and another, which represents the gap change between the two knee flexion positions. The gap differences of 10°, 30°, 60°, 90°, 120° and 135° knee flexion against zero degrees averaged −2.7 ± 2.1 mm, 3.9 ± 2.9 mm, 4.8 ± 3.6 mm, 5.4 ± 4.1 mm, 4.1 ± 3.6 mm, and 1.1 ± 2.7 mm (average ± SD), respectively. Gap differences in deep flexion with respect to the 90° were also examined, because we thought these gap differences beyond 90° would affect knee kinematics at maximum flexion position. The gap differences of 120° − 90° and 135° − 90° averaged −1.2 ± 2.6, −3.6 ± 3.7 mm, respectively. ML soft tissue balance was assessed by calculating the mean joint gap-tilting angle over all flexion angles for each patient. Based on the tilting angle, the 44 knees were classified into three groups: The knees with the mean joint gap tilting of less than −1.0° (lateral tight group, 14 knees), between − 1.0 and 1.0° (well-balanced group, 13 knees), and over 1.0° (medial tight group, 17 knees) The mean joint gap tilting angle of all knees averaged 0.0 ± 3.8°: Joint gap tilting averaged −4.6 ± 2.8° in the lateral tight group, 0.1 ± 0.6° in the well-balanced group, and 3.3 ± 2.0° in the medial tight group. 2.4. Postoperative kinematic analysis At follow-up, a series of dynamic radiographs of squatting and three static lateral radiographs were taken for each patient. For dynamic radiographs, we selected a proper height box (15 cm, 25 cm, and 35 cm) for each patient and determined foot position so the patient could stand up with full knee extension and sit down with maximum knee flexion. A series of sagittal radiographic images of sitting down
Table 1 Preoperative and postoperative clinical data with improvement and correlation assessment.
Extension (°) Flexion (°) Femorotibial angle (°) Knee Society knee score Knee Society function score
Preoperative
At follow-up
p value (paired t-test)
p, r value (correlation)
−7 ± 10 123 ± 20 185 ± 8 39 ± 13 44 ± 17
−1 ± 2 127 ± 15 174 ± 2 94 ± 5 75 ± 16
b0.001 0.105 b0.001 b0.001 b0.001
b0.001, r = 0.606 b0.001, r = 0.585 0.845 0.879 0.032, r = 0.324
T. Watanabe et al. / The Knee 22 (2015) 527–534
529
Fig. 1. (a) The tensor consists of an upper seesaw plate, a lower platform plate, and an extraarticular main body. (b) The offset arm allows surgeons to use the device with the patella reduced during minimally invasive surgery. (c) A joint distraction force of 40 lb (18 kg) was applied by the tensor, and the central joint gaps and mediolateral soft tissue balance (tilting angles) were measured.
and standing up on the box were taken using a flat panel detector (Sonialvision Safire II, Shimadzu, Kyoto, Japan: 7.5 frames/s, 367.20 mm × 367.20 mm image area, 3.922 pixels/mm resolution). We used images of the sitting down phase, from extension to flexion, for this study. Static radiographs were taken using the same detector (365.76 mm × 365.76 mm image area, 7.874 pixels/mm resolution) in the following three positions: (1) straight-leg standing, (2) lunge at maximum flexion and (3) kneeling at maximum flexion. For lunge each patient put their foot on a 15 to 35 cm box and bent the knee to maximum comfortable flexion. For kneeling each patient put their shin on a padded 15 to 35 cm box with their foot hanging freely and bent their knee in the same manner. An investigator (TW) was always available to hold the patient's hands or forearms as a safety measure to prevent the patient from losing balance. The radiographs were digitized and analyzed according to published techniques [18]. Briefly, the three-dimensional position and orientation of the implant components were determined using model-based shape matching techniques, using nonlinear least-squares minimization to refine an initial manual solution. A manufacturer-supplied implant surface model was projected onto the digitized image, and its threedimensional pose was iteratively adjusted to match its silhouette with the silhouette of the patient's knee arthroplasty components. The results of this shape-matching process have standard errors of approximately 0.5° to 1.0° for rotations and 0.5 to 1.0 mm for translations in the sagittal plane [18]. Anteroposterior (AP) locations of each femoral condyle, femoral rotation angles (a positive value means femoral external rotation), valgus/varus angles (a positive value means valgus), and
skeletal flexion angles between the femoral and tibial components, were evaluated. Locations of each femoral condyle were estimated as the lowest point on each femoral condyle relative to the transverse plane of the tibial baseplate. The AP location was defined as the distance between each lowest point and the AP center of the tibial baseplate (a negative value means posterior to the centerline of the baseplate). Skeletal flexion angles were obtained by adding an average of 11° to the implant flexion angle, to reflect average values for femoral anterior bow (five degrees) and posterior tibial slope (six degrees) for these study subjects [19]. In dynamic squatting, the total amount of AP translations calculated from maximum and minimum values of each knee averaged 7 ± 2 mm medially and 9 ± 3 mm laterally. The total amount of femoral external rotations calculated from maximum and minimum values of each knee averaged 9 ± 3°. The total amount of valgus/varus rotations calculated from maximum and minimum values of each knee averaged 3 ± 1°. The range of motion calculated from maximum flexion and extension of each knee averaged 113 ± 19°. In static positions, the medial femoral condylar AP position averaged −4 ± 4 mm at maximum lunge and −5 ± 4 mm at maximum kneeling, while the lateral AP position averaged −10 ± 3 mm at maximum lunge and −9 ± 4 mm at maximum kneeling. The femoral external rotation averaged 8 ± 5° at maximum lunge and 6 ± 5° at maximum kneeling. The valgus/varus angle averaged 1 ± 1° at maximum lunge and 0 ± 1° at maximum kneeling. The skeletal flexion angles at standing, maximum lunge, and maximum kneeling averaged 5 ± 7°, 120 ± 14°, and 123 ± 14°, respectively.
530
T. Watanabe et al. / The Knee 22 (2015) 527–534 respectively, Fig. 3a). The well-balanced knees also had slightly anterior shift of lateral femoral condyle against lateral and medial tight group (p b 0.001, p = 0.045, respectively, Fig. 3b), with similar amount of translation. Resulting from these condylar translations, the well-balanced group exhibited constant femoral external rotation with similar amount compared with other two groups (p b 0.001, Fig. 3c). The medial tight group also demonstrated stable rotation with greater angle, whereas the lateral tight group showed decrease of femoral external rotation in deep flexion. No significant pair-wise difference at each knee angle was found between the three groups in medial and lateral femoral condylar translation and in femoral external rotation. In valgus/varus angle, the wellbalanced group exhibited more neutral valgus rotation and there were significant differences among the three groups (p b 0.001, Fig. 3d). Some pair-wise differences were detected through 70° to 95° knee flexion positions between the medial and lateral tight groups (p b 0.05). With regard to clinical range of motion, there were correlations between gap difference of 135° − 0° and preoperative flexion (r = 0.465, p = 0.004) and postoperative flexion (r = 0.479, p = 0.003) (Table 4). No correlation was demonstrated between joint tilting angle and preoperative and postoperative knee extension and flexion (Table 4). Also, there were no significant preoperative or postoperative extension or flexion differences among three tilting groups (Table 3). Clinical assessments were significantly improved after surgery, except the flexion angle (Table 1). There were strong positive correlations between preoperative and postoperative extension (r = 0.606, p b 0.001), and preoperative and postoperative flexion (r = 0.585, p b 0.001). The F subgroup showed the same significant findings as the entire study group relating preoperative and postoperative clinical data (Table 1). In correlations between gap differences and tilt, and postoperative knee kinematics, the F subgroup demonstrated more significant relations than those in the original group: The original group showed eight significant correlations between gap differences of 135° − 0°, 120° − 90°, 135° − 90° and kinematic values, while the F subgroup showed 13 significant correlations between these gap differences and kinematic parameters (Table 5). Similar to the original study group, the F subgroup showed no significant differences comparing kinematic values and clinical range of motion for the three tilting groups. In correlations between gap differences and tilt, and clinical range of motion, the F subgroup showed correlations between gap differences 120° − 0° and 120° − 90°, and preoperative flexion, in addition to the correlations observed in the original group. In kinematic pathway comparison the F subgroup exhibited the same significant differences, except the F subgroup showed no significant differences in medial condylar AP translation among three groups. The OA subgroup showed the same significant differences in preoperative and postoperative clinical data as the entire study group (Table 1). The OA subgroup demonstrated fewer correlations between gap differences and tilt, and postoperative knee kinematics than the entire study group: The whole study group showed eight significant correlations between gap differences of 135° − 0°, 120° − 90°, 135° − 90° and kinematic values, while the OA subgroup showed six significant correlations (Table 5). Similar to the whole study group, the OA subgroup showed no significant differences for comparisons of kinematic values and clinical range of motion among three tilting groups. In correlations between gap differences and tilt, and clinical range of motion, the OA subgroup showed the same significant differences as the whole study group. The OA subgroup exhibited the same significant differences in kinematic pathway comparisons, except the OA subgroup showed no significant differences in medial condylar AP translation among three groups.
2.5. Assessment of correlations and group comparisons The correlations between intraoperative measurements and postoperative kinematics were evaluated. The knee kinematics was also compared among three tilting groups. Furthermore, preoperative and postoperative clinical measurements were compared, and the correlations between intraoperative measurements and preoperative and postoperative range of motion were assessed to find out clinical relevance. We carried out subgroup analyses for the 36 female knees (32 OA knees and four RA knees, F subgroup) and 38 OA knees (32 female knees and six male knees, OA subgroup). Male and RA subgroups would contain only eight knees and six knees, respectively, which are inadequate for statistical analysis in this study. Paired Student's t-test was used to compare pre- and post-operative clinical measurements. Pearson's test was applied for correlation analyses. Analysis of variance (ANOVA) with post hoc Tukey test was used for tilting group comparisons. The probability value (p) below 0.05 was considered significant. Results are presented as mean ± standard deviation. 3. Results Several correlations are demonstrated between gap difference of 135° with respect to the zero degree reference and postoperative knee kinematics (Table 2). Gap difference of 135° − 0° showed positive correlations with total posterior translation of lateral femoral condyle (r = 0.336, p = 0.042) and total femoral external rotation (r = 0.488, p = 0.002, Fig. 2a) during squatting. Gap difference of 135° − 0° also demonstrated negative correlation with AP position of lateral femoral condyle at maximum kneeling (r = −0.510, p = 0.001, Fig. 2b), which has the same meaning as the positive correlation in squatting, because, in kneeling, negative value indicates posterior position of the femoral condyle, while in squatting, we took up the total amount of translations which are always positive values. Maximum knee flexion also had positive correlations with this gap difference (r = 0.355, p = 0.031). Other correlations were observed between deep flexion gap differences with respect to the 90° reference and postoperative knee kinematics (Table 2). Gap difference of 120° − 90° exhibited positive correlations with femoral external rotation at maximum lunge (r = 0.353, p = 0.019). Gap difference of 120° − 90° also showed negative correlation with AP position of lateral femoral condyle at maximum kneeling (r = −0.308, p = 0.042) and positive correlation with femoral external rotation at maximum kneeling (r = 0.492, p = 0.001) Gap difference of 135° − 90° had positive correlations with femoral external rotation at maximum lunge (r = 0.447, p = 0.006). No other correlation was found between joint gap differences and postoperative knee kinematics. No correlation was demonstrated between joint tilting angle and postoperative knee kinematics (Table 2). Also, there were no significant kinematic value differences among three tilting groups in squatting, lunge and kneeling activities (Table 3). However, if we look at the kinematic pathway during squatting, the well-balanced group showed significantly different kinematics compared with other two groups. The well-balanced knees demonstrated less anterior translation of medial femoral condyle in mid- to deep flexion (p = 0.004, p = 0.001 against lateral and medial tight group,
4. Discussion We investigated the relationships between intraoperative gap measurements and weight-bearing kinematics in 44 knees with primary
Table 2 Correlations between gap differences and tilt, and postoperative knee kinematics of 44 knees. Activities
Kinematics
Squatting
Total medal condylar AP translation (mm) Total lateral condylar AP translation (mm) Total femoral external rotation (°) Total valgus rotation (°) Flexion arc (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Maximum extension (°)
Lunge
Kneeling
Standing
p, r value (correlation) GD10-0
GD30-0
GD60-0
GD90-0
GD120-0
GD135-0
GD120-90
GD135-90
Tilt
0.936 0.114 0.063 0.655 0.104 0.907 0.783 0.672 0.666 0.855 0.120 0.316 0.430 0.441 0.189 0.482
0.939 0.819 0.657 0.859 0.350 0.904 0.892 0.948 0.493 0.694 0.311 0.377 0.125 0.891 0.190 0.884
0.756 0.840 0.919 0.410 0.622 0.707 0.676 0.564 0.863 0.830 0.207 0.548 0.086 0.402 0.285 0.614
0.696 0.670 0.530 0.504 0.283 0.571 0.894 0.767 0.968 0.456 0.219 0.743 0.179 0.352 0.181 0.995
0.776 0.171 0.067 0.886 0.133 0.516 0.402 0.177 0.300 0.621 0.582 0.273 0.413 0.192 0.152 0.709
0.788 0.042, r = 0.336 0.002, r = 0.488 0.987 0.169 0.878 0.100 0.112 0.279 0.257 0.243 0.001, r = −0.510 0.081 0.405 0.031, r = 0.355 0.488
0.827 0.232 0.136 0.400 0.715 0.073 0.350 0.019, r = 0.353 0.138 0.631 0.251 0.042, r = −0.308 0.001, r = 0.492 0.747 0.903 0.617
0.335 0.719 0.316 0.880 0.733 0.277 0.328 0.051 0.063 0.881 0.304 0.195 0.006, r = 0.447 0.618 1.000 0.728
0.162 0.451 0.426 0.342 0.154 0.485 0.969 0.667 0.546 0.272 0.242 0.527 0.663 0.106 0.084 0.976
GD: gap difference between two knee flexion angles, AP: anteroposterior.
T. Watanabe et al. / The Knee 22 (2015) 527–534
531
Fig. 2. (a) Gap difference of 135° minus zero degree showed positive correlations with total femoral external rotation (r = 0.488, p = 0.002) during squatting. (b) Gap difference of 135° minus zero degree demonstrated negative correlation with anteroposterior position of lateral femoral condyle in maximum kneeling (r = −0.510, p = 0.001).
PS-TKA in order to establish objective relationships between intraoperative variables and postoperative knee kinematics. Intraoperative joint gaps usually are measured with the patella everted or laterally shifted, which tightens lateral structures and reduces the effect of the extensor mechanism in deep flexion. This may preclude accurate evaluation of joint gaps and ML soft-tissue balance [5,8,20]. Also, intraoperative joint gaps generally are measured in extension and 90° flexion, which provides limited information in mid- and deep flexion. We measured joint gaps with the patella reduced and in smaller flexion intervals and examined the effect on postoperative kinematics. Consistent with our hypotheses, joint gap differences in deep flexion with respect to the zero degree or 90° references correlated with the total amount of posterior lateral condylar translation and femoral external rotation during squatting, femoral external rotation during lunge, posterior lateral translation, femoral external rotation, and maximum knee flexion during kneeling. The gap difference of 135° with respect to the zero degree reference also positively correlated preoperative and postoperative flexion. ML tilting did not have significant correlation with postoperative kinematics. Various kinematic values were not significantly different among three tilting group, either. However, dynamic kinematics during squatting demonstrated the differences of femoral condylar AP translation and femoral external rotation between the well-balanced knees and the other unbalanced knees. The valgus rotations were different
among three groups with the balanced knees in the most neutral rotation. There are a few limitations in this study. First, many factors including surgeries, postoperative treatment, and patients' conditions could affect arthroplasty outcomes, so focusing solely on the relationships between intraoperative laxity measurements and postoperative kinematics is challenging. However, we believe a single-surgeon series is appropriate to assess these relationships. For this study, all surgeries were performed by a single surgeon with the same procedures: surgical approach, bone cutting, soft tissue balancing, and other maneuvers. Having eliminated variation in surgical techniques to the extent possible [21], we focused on the effect of intraoperative joint gaps and ML soft tissue balance on postoperative kinematics. Second, our subject cohort consisted primarily of women with osteoarthritis. Our findings may not generalize to men or patients with rheumatoid arthritis. Also, we studied PS arthroplasties, where PCL resection enlarges the flexion gap and the post-cam mechanism controlled knee kinematics [22]. Arthroplasties that retain the PCL but lack the post-cam mechanism may exhibit different characteristics. Previous studies have shown larger joint gaps at 90° or greater flexion predicted larger postoperative knee flexion [20,23,24]. Different studies have reported greater posterior femoral translation has a positive correlation with greater maximum knee flexion [14,15]. Our results
Table 3 Comparison of kinematic values and clinical range of motion among three tilting groups.
Squatting
Lunge
Kneeling
Standing Clinical range of motion
AP: anteroposterior.
Total medal condylar AP translation (mm) Total lateral condylar AP translation (mm) Total femoral external rotation (°) Total valgus rotation (°) Flexion arc (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Maximum extension (°) Preoperative extension Preoperative flexion Postoperative extension Postoperative flexion
Lateral tight (n = 14)
Well-balanced (n = 13)
Medial tight (n = 17)
p-Value (ANOVA)
7±2 8±3 10 ± 3 3±1 114 ± 12 −3 ± 4 −10 ± 3 9±6 1±1 120 ± 10 −5 ± 4 −9 ± 3 5±6 1±1 127 ± 9 −4 ± 6 −10 ± 10 123 ± 15 −1 ± 2 128 ± 11
8±1 9±4 9±3 3±1 117 ± 25 −5 ± 3 −10 ± 4 7±4 1±1 123 ± 17 −6 ± 3 −9 ± 4 4±4 1±1 125 ± 16 −10 ± 9 −5 ± 11 126 ± 20 −1 ± 3 130 ± 19
7±2 8±3 9±3 3±1 108 ± 19 −3 ± 3 −10 ± 4 9±6 0±1 117 ± 15 −4 ± 4 −9 ± 4 7±5 0±2 119 ± 16 −5 ± 5 −6 ± 8 122 ± 24 −1 ± 2 125 ± 15
0.304 0.838 0.896 0.559 0.266 0.886 0.824 0.449 0.412 0.480 0.274 0.894 0.380 0.213 0.331 0.096 0.419 0.836 0.905 0.630
532
T. Watanabe et al. / The Knee 22 (2015) 527–534
Fig. 3. (a) The well-balanced knees demonstrated less anterior translation of medial femoral condyle in mid- to deep flexion (p = 0.004, p = 0.001 against lateral and medial tight group, respectively). (b) The well-balanced knees also had slightly anterior shift of lateral femoral condyle against lateral and medial tight group (p b 0.001, p = 0.045, respectively), with similar amount of translation. (c) The well-balanced group exhibited constant femoral external rotation with similar amount compared with other two groups (p b 0.001 against lateral and medial tight group). (d) The well-balanced group showed more neutral valgus rotation and there were significant differences among the three groups (p b 0.001). Some pair-wise differences were detected at 70° to 95° knee flexion positions between the medial and lateral tight groups (p b 0.05).
are consistent with these previous studies. Our study also demonstrated correlations between joint gap differences in deep flexion with respect to the zero degree and 90° references and total lateral posterior translations, femoral external rotation, and lateral condylar AP positions. These relationships between joint gaps and postoperative kinematics have not been well discussed. Greater deep flexion gaps, compared to gaps at zero degree and 90°, are indicative of extensor mechanism flexibility. In very deep flexion, e.g. 135°, joints with a still-flexible extensor mechanism could show greater knee flexion and more natural kinematics, including femoral posterior translation and external rotation. It is well known that preoperative range of flexion correlates postoperative flexion angle [25,26]. Preoperative soft tissue condition around the knee including extensor mechanism should affect postoperative knee flexion. Current study demonstrated positive correlations between
preoperative and postoperative range of flexion, and joint gap differences in deep flexion. Large deep flexion gap differences with respect to the zero degree and 90° references would reflect more extensible soft tissue which is closely related to the postoperative better range of flexion. ML soft tissue balance is an important factor for successful TKA [1,2]. Failure to achieve adequate ML balance is a predictor of poor outcomes [27]. Sasanuma et al. reported intraoperative soft tissue balance after releases averaged 2.1° in extension and − 1.6° at 90° flexion [28]. We found an average 0.0° ML tilt in this study, consistent with previous reports [9,28]. Perfect balance is difficult to achieve [3], and it has been questioned whether rectangular gaps are ideal since normal gaps are trapezoidal [29]. In current study, we did not detect any correlation between ML balance in the range of several degrees and knee kinematics
Table 4 Correlations between gap differences and tilt, and clinical range of motion. Clinical range of motion
Preoperative extension Preoperative flexion Postoperative extension Postoperative flexion
p, r value (correlation) GD10-0
GD30-0
GD60-0
GD90-0
GD120-0
GD135-0
GD120-90
GD135-90
Tilt
0.138 0.540 0.597 0.255
0.185 0.711 0.951 0.510
0.100 0.798 0.342 0.491
0.472 0.359 0.624 0.278
0.799 0.055 0.579 0.133
0.698 0.004, r = 0.465 0.563 0.003, r = 0.479
0.443 0.251 0.125 0.725
0.613 0.144 0.902 0.285
0.302 0.983 0.921 0.505
GD: gap difference between two knee flexion angles.
T. Watanabe et al. / The Knee 22 (2015) 527–534
533
Table 5 Correlations between gap differences and tilt, and postoperative knee kinematics of two subgroups. Activities
Kinematics
p, r value (correlation) Female (n = 36)
Squatting
Lunge
Kneeling
Standing
Osteoarthritis (n = 38)
GD120-0
GD135-0
GD120-90
GD135-90
GD120-0
GD135-0
GD120-90
GD135-90
Total medal condylar AP translation (mm) Total lateral condylar AP translation (mm)
0.760 0.306
0.914 0.088
0.365 0.424
0.974 0.182
0.492 0.093
0.511 0.367
0.208 0.991
Total femoral external rotation (°)
0.058
0.314
0.059
0.332
0.819 0.099 0.594
0.898 0.930 0.317
0.853 0.164 0.440
0.003, r = 0.519 0.807 0.303 0.778
0.081
Total valgus rotation (°) Flexion arc (°) Medial condylar AP translation (mm)
0.003, r = 0.517 0.810 0.176 0.963
0.700 0.029, r = 0.364 0.025, r = 0.374 0.469 0.250 0.092
0.932 0.526 0.219
Lateral condylar AP translation (mm)
0.232
0.051
0.135
0.315
0.134
0.515 0.954 0.028, r = 0.356 0.375
Femoral external rotation (°)
0.206
0.165
0.087
0.090
0.623
0.566
0.274
0.394
0.005, r = 0.444 0.127
0.064
Valgus rotation (°) Maximum flexion (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm)
0.423 0.678 0.162
0.663 0.792 0.286
0.317 0.160 0.100
0.597 0.268 0.449
Femoral external rotation (°)
0.225
Valgus rotation (°) Maximum flexion (°)
0.929 0.127
0.019, r = 0.418 0.795 0.659
Maximum extension (°)
0.722
0.226 0.289 b0.001, r = −0.630 0.040, r = 0.371 0.848 0.034, r = 0.381 0.473
0.037, r = 0.376 0.034, r = −0.382 0.851 0.456 0.103
0.642
0.028, r = −0.365 0.002, r = 0.499 0.194 0.991 0.374 0.028, r = −0.365 0.002, r = 0.495 0.604 0.789
0.012, r = 0.445 0.943 0.911
0.306
0.362 0.376 0.006, r = −0.487 0.107
0.203 0.195
0.312 0.086
0.001, r = 0.509 0.625 0.498
0.053
0.940
0.878
0.380
0.885
0.522
0.103
GD: gap difference between two knee flexion angles, AP: anteroposterior.
in various activities. However, some kinematic differences among three tilting groups were demonstrated. Normal and osteoarthritic knee kinematics show continuous and smooth femoral external rotation with knee flexion [30,31]. Femoral external rotation and Condylar posterior translation are the greatest in normal knees, followed by osteoarthritic knees, and the least in knees with an arthroplasty. In this study, the well-balanced group showed less medial condylar anterior translation in mid-flexion and linear femoral external rotation with similar amount compared with the other two groups. The valgus/varus rotation reflected the ML soft tissue balance with the well-balanced knees in the most neutral. Female knees showed results similar to the entire study group, with additional significant relations between gap differences and postoperative knee kinematics, and between gap differences and clinical range of motion. Female joints generally exhibit more flexibility than male joints [32], which would affect joint kinematics. Greater soft tissue elasticity in female knees could enhance the influence of joint gap on the postoperative knee kinematics. The OA subgroup knees exhibited results similar to the whole study group, but fewer correlations between gap differences and postoperative knee kinematics. Both OA and RA knees exhibit greater AP laxity, restricted femoral axial rotation, and an increased valgus/varus rotation compared to normal knees, although AP laxity decreases with disease severity only in OA knees [33]. Unfortunately, our study cohort had too few RA knees to permit direct comparison of OA and RA knees, so conclusions relating disease and knee mechanics cannot be made from our data. Our study demonstrated gap differences of 135° − 0° was correlated with the total posterior translation of lateral femoral condyle and femoral external rotation during squatting, lateral condylar AP position and maximum knee flexion in kneeling. Gap differences of 120° − 90° and 135° − 90° also exhibited similar correlations. Gap difference of 135° − 0° was also correlated with preoperative and postoperative knee flexion. ML tilting affect postoperative kinematic pathway during squatting. Well-balanced knees showed less anterior translation of medial femoral condyle in mid- to deep flexion, consistent femoral external rotation, and the most neutral valgus/varus rotation. These
findings suggest adequate intraoperative joint gap differences beyond 90° flexion with respect to the zero degree or 90° references are important to obtain greater lateral femoral translation, femoral external rotation, and better postoperative flexion. Also, ML imbalance negatively affects knee kinematic pathway during weight-bearing activity, indicating the importance of soft tissue balancing.
Acknowledgments We thank Hirotsugu Muratsu MD, for the advice on using the intraoperative joint gap tensor developed at Kobe University.
References [1] Kurosaka M, Yoshiya S, Mizuno K, Yamamoto T. Maximizing flexion after total knee arthroplasty: the need and the pitfalls. J Arthroplasty 2002;17:59–62. [2] Scuderi G. The stiff total knee arthroplasty causality and solution. J Arthroplasty 2005;20:23–6. [3] Griffin FM, Insall JN, Scuderi GR. Accuracy of soft tissue balancing in total knee arthroplasty. J Arthroplasty 2000;15:970–3. [4] Sambatakakis A, Attfield SF, Newton G. Quantification of soft-tissue imbalance in condylar knee arthroplasty. J Biomed Eng 1993;15:339–43. [5] Matsumoto T, Muratsu H, Tsumura N, Mizuno K, Kuroda R, Yoshiya S, et al. Joint gap kinematics in posterior-stabilized total knee arthroplasty measured by a new tensor with the navigation system. J Biomech Eng 2006;128:867–71. [6] Winemaker MJ. Perfect balance in total knee arthroplasty: the elusive compromise. J Arthroplasty 2002;17:2–10. [7] Wallace AL, Harris ML, Walsh WR, Bruce WJ. Intraoperative assessment of tibiofemoral contact stresses in total knee arthroplasty. J Arthroplasty 1998;13: 923–7. [8] Yoshino N, Watanabe N, Watanabe Y, Fukuda Y, Takai S. Measurement of joint gap load in patella everted and reset position during total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2009;17:484–90. [9] Moro-oka TA, Shiraishi H, Iwamoto Y, Banks SA. Modified gap-balancing technique in total knee arthroplasty: evaluation of the post-operative coronal laxity. Knee Surg Sports Traumatol Arthrosc 2010;18:375–80. [10] Banks SA, Harman MK, Bellemans J, Hodge WA. Making sense of knee arthroplasty kinematics: news you can use. J Bone Joint Surg Am 2003;85-A(Suppl. 4):64–72. [11] Mikashima Y, Tomatsu T, Horikoshi M, Nakatani T, Saito S, Momohara S, et al. In vivo deep-flexion kinematics in patients with posterior-cruciate retaining and anteriorcruciate substituting total knee arthroplasty. Clin Biomech (Bristol, Avon) 2010;25: 83–7.
534
T. Watanabe et al. / The Knee 22 (2015) 527–534
[12] Okamoto N, Breslauer L, Hedley AK, Mizuta H, Banks SA. In vivo knee kinematics in patients with bilateral total knee arthroplasty of 2 designs. J Arthroplasty 2011;26: 914–8. [13] Cates HE, Komistek RD, Mahfouz MR, Schmidt MA, Anderle M. In vivo comparison of knee kinematics for subjects having either a posterior stabilized or cruciate retaining high-flexion total knee arthroplasty. J Arthroplasty 2008;23:1057–67. [14] Banks S, Bellemans J, Nozaki H, Whiteside LA, Harman M, Hodge WA. Knee motions during maximum flexion in fixed and mobile-bearing arthroplasties. Clin Orthop Relat Res 2003:131–8. [15] Victor J, Banks S, Bellemans J. Kinematics of posterior cruciate ligament-retaining and -substituting total knee arthroplasty: a prospective randomised outcome study. J Bone Joint Surg (Br) 2005;87:646–55. [16] Itokazu M, Masuda K, Wada E, Ohno T, Yoshida M, Takatu T. Influence of anteroposterior and mediolateral instability on range of motion after total knee arthroplasty: an ultrasonographic study. Orthopedics 2000;23:49–52. [17] Warren PJ, Olanlokun TK, Cobb AG, Walker PS, Iverson BF. Laxity and function in knee replacements. A comparative study of three prosthetic designs. Clin Orthop Relat Res 1994:200–8. [18] Banks SA, Hodge WA. Accurate measurement of three-dimensional knee replacement kinematics using single-plane fluoroscopy. IEEE Trans Biomed Eng 1996;43:638–49. [19] Banks SA, Harman MK, Hodge WA. Mechanism of anterior impingement damage in total knee arthroplasty. J Bone Joint Surg Am 2002;84-A(Suppl. 2):37–42. [20] Matsumoto T, Mizuno K, Muratsu H, Tsumura N, Fukase N, Kubo S, et al. Influence of intra-operative joint gap on post-operative flexion angle in osteoarthritis patients undergoing posterior-stabilized total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2007;15:1013–8. [21] Nozaki H, Banks SA, Suguro T, Hodge WA. Observations of femoral rollback in cruciate-retaining knee arthroplasty. Clin Orthop Relat Res 2002:308–14. [22] Yagishita K, Muneta T, Ikeda H. Step-by-step measurements of soft tissue balancing during total knee arthroplasty for patients with varus knees. J Arthroplasty 2003;18: 313–20. [23] Higuchi H, Hatayama K, Shimizu M, Kobayashi A, Kobayashi T, Takagishi K. Relationship between joint gap difference and range of motion in total knee arthroplasty: a
[24]
[25] [26]
[27]
[28] [29] [30]
[31]
[32]
[33]
prospective randomised study between different platforms. Int Orthop 2009;33: 997–1000. Takayama K, Matsumoto T, Kubo S, Muratsu H, Ishida K, Matsushita T, et al. Influence of intra-operative joint gaps on post-operative flexion angle in posterior cruciateretaining total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2012;20: 532–7. Kawamura H, Bourne RB. Factors affecting range of flexion after total knee arthroplasty. J Orthop Sci 2001;6:248–52. Ritter MA, Harty LD, Davis KE, Meding JB, Berend ME. Predicting range of motion after total knee arthroplasty. Clustering, log-linear regression, and regression tree analysis. J Bone Joint Surg Am 2003;85-A:1278–85. Wasielewski RC, Galat DD, Komistek RD. Correlation of compartment pressure data from an intraoperative sensing device with postoperative fluoroscopic kinematic results in TKA patients. J Biomech 2005;38:333–9. Sasanuma H, Sekiya H, Takatoku K, Takada H, Sugimoto N. Evaluation of soft-tissue balance during total knee arthroplasty. J Orthop Surg (Hong Kong) 2010;18:26–30. Tokuhara Y, Kadoya Y, Nakagawa S, Kobayashi A, Takaoka K. The flexion gap in normal knees. An MRI study. J Bone Joint Surg (Br) 2004;86:1133–6. Moro-oka TA, Hamai S, Miura H, Shimoto T, Higaki H, Fregly BJ, et al. Dynamic activity dependence of in vivo normal knee kinematics. J Orthop Res 2008;26: 428–34. Hamai S, Moro-oka TA, Miura H, Shimoto T, Higaki H, Fregly BJ, et al. Knee kinematics in medial osteoarthritis during in vivo weight-bearing activities. J Orthop Res 2009; 27:1555–61. Shultz SJ, Shimokochi Y, Nguyen AD, Schmitz RJ, Beynnon BD, Perrin DH. Measurement of varus-valgus and internal–external rotational knee laxities in vivo—part II: relationship with anterior–posterior and general joint laxity in males and females. J Orthop Res 2007;25:989–96. Wada M, Imura S, Baba H, Shimada S. Knee laxity in patients with osteoarthritis and rheumatoid arthritis. Br J Rheumatol 1996;35:560–3.
Contents lists available at ScienceDirect
The Knee
Intraoperative joint gaps and mediolateral balance affect postoperative knee kinematics in posterior-stabilized total knee arthroplasty☆ Toshifumi Watanabe a,b,c,⁎, Takeshi Muneta c, Ichiro Sekiya c, Scott A. Banks a a b c
Department of Mechanical & Aerospace Engineering, University of Florida, Gainesville, FL, USA Department of Orthopaedic Surgery, Tsuchiura Kyodo General Hospital, Tsuchiura, Ibaraki, Japan Department of Orthopaedic Surgery, Tokyo Medical and Dental University Hospital, Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 26 September 2014 Received in revised form 22 January 2015 Accepted 17 March 2015 Keywords: Intraoperative joint gap Mediolateral balance Postoperative knee kinematics Posterior-stabilized Total knee arthroplasty
a b s t r a c t Background: Adjusting joint gaps and establishing mediolateral (ML) soft tissue balance are considered essential interventions for better outcomes in total knee arthroplasty (TKA). However, the relationship between intraoperative laxity measurements and weightbearing knee kinematics has not been well explored. This study aimed to quantify the effect of intraoperative joint gaps and ML soft tissue balance on postoperative knee kinematics in posterior-stabilized (PS)-TKA. Methods: We investigated 44 knees in 34 patients who underwent primary PS-TKA by a single surgeon. The central joint gaps and ML tilting angles at 0°, 10°, 30°, 60°, 90°, 120° and 135° flexion were measured during surgery. At a minimum of two year follow-up, we analyzed in vivo kinematics of these knees and examined the influence of intraoperative measurements on postoperative kinematics. Results: Gap difference of knee flexion at 135° minus 0° was correlated with the total posterior translation of lateral femoral condyle (r = 0.336, p = 0.042) and femoral external rotation (r = 0.488, p = 0.002) during squatting, anteroposterior position of lateral femoral condyle (r = ‐ 0.510, p = 0.001) and maximum knee flexion (r = 0.355, p = 0.031) in kneeling. Similar correlations were observed between deep flexion gap differences with respect to the 90° reference and postoperative knee kinematics. Well-balanced knees showed less anterior translation of medial femoral condyle in mid- to deep flexion, consistent femoral external rotation, and the most neutral valgus/varus rotation compared with unbalanced knees. Conclusion: These findings indicate the importance of adequate intraoperative joint gaps in deep flexion and ML soft tissue balance throughout the range of motion. Level of evidence: Level II. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Establishing proper intraoperative joint gaps and mediolateral (ML) soft tissue balance are considered essential parts of surgery to achieve the best possible outcomes after total knee arthroplasty (TKA) [1,2]. Various devices have been developed to measure these gaps and ML balance, helping surgeons quantify the procedure [3–8]. These devices have been used to quantify detailed joint gap patterns and have been used, for example, to show the patellofemoral (PF) joint should be reduced during measurements to more closely mimic normal knee mechanics [5,8]. Despite the growing popularity of intraoperative gap measurements, there have been few reports relating intraoperative ☆ This work was mainly performed at Tsuchiura Kyodo General Hospital, Tsuchiura, Ibaraki, Japan and at the University of Florida, Gainesville, FL, USA. ⁎ Corresponding author at: Tokyo Medical and Dental University Hospital, Department of Orthopaedic Surgery, 1-5-45, Yushima, Bunkyo-ku, 113-8510 Tokyo, Japan. Tel./fax: +81 3 5803 4020. E-mail addresses: [email protected] (T. Watanabe), [email protected] (T. Muneta), [email protected] (I. Sekiya), banks@ufl.edu (S.A. Banks).
http://dx.doi.org/10.1016/j.knee.2015.03.006 0968-0160/© 2015 Elsevier B.V. All rights reserved.
joint gaps and balance to weight-bearing knee kinematics postoperatively [9]. Previous studies of TKA utilizing 3D–2D model-image registration techniques have clarified in vivo kinematics of knees with various arthroplasty designs [10–13]. Many studies have reported that both medial and lateral femoral condyles generally translate posteriorly with flexion, with greater lateral condylar posterior translation coupled to femoral external rotation [10–14]. Several studies have reported greater posterior femoral translation has a positive correlation with greater maximum knee flexion [14,15]. However, how intraoperative joint gaps and ML soft tissue balance affect weight-bearing knee kinematics has not been well investigated. One of a surgeon's best opportunities to predict and affect postoperative kinematics is during surgery, so it is important to establish relationships between intraoperative gaps and balance, and postoperative knee kinematics. The goal of this study was to establish how intraoperative joint gaps and ML soft tissue balance affect postoperative kinematics in posteriorstabilized (PS)-TKA. We hypothesized that knees with larger joint gap differences in deep flexion with respect to zero degrees or 90° reference
528
T. Watanabe et al. / The Knee 22 (2015) 527–534
and good ML balance throughout the range of motion would exhibit greater femoral posterior translation, femoral external rotation, and better maximum knee flexion, because adequate joint laxity and balanced ML soft tissue generally should provide more natural knee motions [1,2,16,17]. 2. Materials and methods 2.1. Subjects We studied 44 knees in 34 patients who underwent primary PS-TKA between January 2006 and August 2008. A single surgeon (TW) performed all surgeries using generally accepted techniques for minimally invasive surgery (MIS) and the same prosthesis (NexGen LPS Flex, Zimmer, Warsaw, Indiana, USA). Joint gaps and ML soft tissue balance were recorded during each case. The patients were studied at a minimum of two years follow-up. Knees with previous high tibial osteotomy and knees with severe deformity requiring metal block augmentation were excluded. All patients gave informed consent to participate in this institutional review board approved study. Subjects averaged 71 ± 7 years (average ± SD) at the time of surgery. The study cohort included 36 female and eight male knees with a preoperative diagnosis of osteoarthritis (OA) in 38 knees and rheumatoid arthritis (RA) in six knees. The body mass index of the patients averaged 26 ± 5 kg/m2. Two knees had valgus alignment with femorotibial angle less than 170° and the other knees had neutral or varus alignment (Table 1). The follow-up period after surgery averaged 2.4 ± 0.7 years. A sample of 44 knees was computed to produce 80% power (1-b) for correlating intraoperative joint gaps and postoperative knee kinematics using an effect size of 0.4, or an r2 value explaining 16% of the data variance. 2.2. Surgical procedure An air tourniquet was inflated to 330 mm Hg during surgery. The mini-midvastus approach with independent bone cutting technique was used for all knees. The distal femur was cut perpendicular to its mechanical axis, removing the amount of bone corresponding to the prosthetic femoral component thickness. The proximal tibia was cut perpendicular to its coronal mechanical axis and with approximately six degrees sagittal posterior tibial slope, removing bone on the intact side corresponding to the prosthetic tibial component thickness. The posterior femoral condyles were cut with approximately three degrees external rotation from the posterior condylar line. Whiteside's line and epicondilar line were used as a femoral rotation reference in valgus knees. Following the bone cutting, the soft tissues were released on a caseby-case basis to obtain ML balance. The posterior cruciate ligament (PCL) was sacrificed in all knees. The deep layer of the medial collateral ligament (MCL) was released in all knees with varus deformity. After osteophyte removal, the superficial layer of the MCL, posterior capsule, semimembranosus, and pes anserinus were released sequentially until adequate ML soft tissue balance was obtained. No soft tissue release of lateral structures was required to obtain acceptable ML soft tissue balance. 2.3. Intra-operative measurements Joint gaps and ML soft tissue balance were measured after performing soft tissue releases. The femoral trial was put in place and a tensor (Offset
Repo-Tensor; Zimmer, Tokyo, Japan) developed by Kobe University [5] was placed between the tibial cut surface and the femoral trial. The tensor consists of three parts: an upper seesaw plate, a lower platform plate, and an extraarticular main body (Fig. 1a, b). The lower platform plate was fixed on the center of the tibial cut surface and held in place with small pins protruding from the bottom side of the plate. The seesaw plate has a post that fits into the intercondylar space and articulates with the cam of the femoral trial. This post-cam mechanism controls the tibiofemoral translation in both the coronal and sagittal planes over the entire arc of knee flexion. The main body connects the other two parts. The tensor is small enough to be applied for MIS procedures and the offset arm allows surgeons to use the device with the patella reduced. The tensor provides numerical measures of joint gap at the center of the knee and the ML tilting angle from full extension to deep flexion. The patella was reduced to the original position and several sutures were used to maintain correct PF position throughout the range of motion (Fig. 1c). A joint distraction force of 40 lb (18 kg) was applied by the tensor, and the central joint gaps and ML tilting angles were measured at zero degrees, 10°, 30°, 60°, 90°, 120° and 135° flexion. Preliminary studies with this tensor indicated the joint gap at full extension with 40 lb of joint distraction force corresponded most closely to the appropriate tibial insert thickness [5]. An assistant supported the thigh during measurement to maintain appropriate sagittal alignment and to reduce the influence of the thigh mass on knee forces. The accuracy of this measurement has been estimated to be ± 0.3 mm in the joint gap [5]. Gap data for 10° and 135° flexion positions are available for 39 and 37 knees, respectively, as these increments were not tested in the first five to seven knees. We defined the “gap difference” as a gap size difference between one gap and another, which represents the gap change between the two knee flexion positions. The gap differences of 10°, 30°, 60°, 90°, 120° and 135° knee flexion against zero degrees averaged −2.7 ± 2.1 mm, 3.9 ± 2.9 mm, 4.8 ± 3.6 mm, 5.4 ± 4.1 mm, 4.1 ± 3.6 mm, and 1.1 ± 2.7 mm (average ± SD), respectively. Gap differences in deep flexion with respect to the 90° were also examined, because we thought these gap differences beyond 90° would affect knee kinematics at maximum flexion position. The gap differences of 120° − 90° and 135° − 90° averaged −1.2 ± 2.6, −3.6 ± 3.7 mm, respectively. ML soft tissue balance was assessed by calculating the mean joint gap-tilting angle over all flexion angles for each patient. Based on the tilting angle, the 44 knees were classified into three groups: The knees with the mean joint gap tilting of less than −1.0° (lateral tight group, 14 knees), between − 1.0 and 1.0° (well-balanced group, 13 knees), and over 1.0° (medial tight group, 17 knees) The mean joint gap tilting angle of all knees averaged 0.0 ± 3.8°: Joint gap tilting averaged −4.6 ± 2.8° in the lateral tight group, 0.1 ± 0.6° in the well-balanced group, and 3.3 ± 2.0° in the medial tight group. 2.4. Postoperative kinematic analysis At follow-up, a series of dynamic radiographs of squatting and three static lateral radiographs were taken for each patient. For dynamic radiographs, we selected a proper height box (15 cm, 25 cm, and 35 cm) for each patient and determined foot position so the patient could stand up with full knee extension and sit down with maximum knee flexion. A series of sagittal radiographic images of sitting down
Table 1 Preoperative and postoperative clinical data with improvement and correlation assessment.
Extension (°) Flexion (°) Femorotibial angle (°) Knee Society knee score Knee Society function score
Preoperative
At follow-up
p value (paired t-test)
p, r value (correlation)
−7 ± 10 123 ± 20 185 ± 8 39 ± 13 44 ± 17
−1 ± 2 127 ± 15 174 ± 2 94 ± 5 75 ± 16
b0.001 0.105 b0.001 b0.001 b0.001
b0.001, r = 0.606 b0.001, r = 0.585 0.845 0.879 0.032, r = 0.324
T. Watanabe et al. / The Knee 22 (2015) 527–534
529
Fig. 1. (a) The tensor consists of an upper seesaw plate, a lower platform plate, and an extraarticular main body. (b) The offset arm allows surgeons to use the device with the patella reduced during minimally invasive surgery. (c) A joint distraction force of 40 lb (18 kg) was applied by the tensor, and the central joint gaps and mediolateral soft tissue balance (tilting angles) were measured.
and standing up on the box were taken using a flat panel detector (Sonialvision Safire II, Shimadzu, Kyoto, Japan: 7.5 frames/s, 367.20 mm × 367.20 mm image area, 3.922 pixels/mm resolution). We used images of the sitting down phase, from extension to flexion, for this study. Static radiographs were taken using the same detector (365.76 mm × 365.76 mm image area, 7.874 pixels/mm resolution) in the following three positions: (1) straight-leg standing, (2) lunge at maximum flexion and (3) kneeling at maximum flexion. For lunge each patient put their foot on a 15 to 35 cm box and bent the knee to maximum comfortable flexion. For kneeling each patient put their shin on a padded 15 to 35 cm box with their foot hanging freely and bent their knee in the same manner. An investigator (TW) was always available to hold the patient's hands or forearms as a safety measure to prevent the patient from losing balance. The radiographs were digitized and analyzed according to published techniques [18]. Briefly, the three-dimensional position and orientation of the implant components were determined using model-based shape matching techniques, using nonlinear least-squares minimization to refine an initial manual solution. A manufacturer-supplied implant surface model was projected onto the digitized image, and its threedimensional pose was iteratively adjusted to match its silhouette with the silhouette of the patient's knee arthroplasty components. The results of this shape-matching process have standard errors of approximately 0.5° to 1.0° for rotations and 0.5 to 1.0 mm for translations in the sagittal plane [18]. Anteroposterior (AP) locations of each femoral condyle, femoral rotation angles (a positive value means femoral external rotation), valgus/varus angles (a positive value means valgus), and
skeletal flexion angles between the femoral and tibial components, were evaluated. Locations of each femoral condyle were estimated as the lowest point on each femoral condyle relative to the transverse plane of the tibial baseplate. The AP location was defined as the distance between each lowest point and the AP center of the tibial baseplate (a negative value means posterior to the centerline of the baseplate). Skeletal flexion angles were obtained by adding an average of 11° to the implant flexion angle, to reflect average values for femoral anterior bow (five degrees) and posterior tibial slope (six degrees) for these study subjects [19]. In dynamic squatting, the total amount of AP translations calculated from maximum and minimum values of each knee averaged 7 ± 2 mm medially and 9 ± 3 mm laterally. The total amount of femoral external rotations calculated from maximum and minimum values of each knee averaged 9 ± 3°. The total amount of valgus/varus rotations calculated from maximum and minimum values of each knee averaged 3 ± 1°. The range of motion calculated from maximum flexion and extension of each knee averaged 113 ± 19°. In static positions, the medial femoral condylar AP position averaged −4 ± 4 mm at maximum lunge and −5 ± 4 mm at maximum kneeling, while the lateral AP position averaged −10 ± 3 mm at maximum lunge and −9 ± 4 mm at maximum kneeling. The femoral external rotation averaged 8 ± 5° at maximum lunge and 6 ± 5° at maximum kneeling. The valgus/varus angle averaged 1 ± 1° at maximum lunge and 0 ± 1° at maximum kneeling. The skeletal flexion angles at standing, maximum lunge, and maximum kneeling averaged 5 ± 7°, 120 ± 14°, and 123 ± 14°, respectively.
530
T. Watanabe et al. / The Knee 22 (2015) 527–534 respectively, Fig. 3a). The well-balanced knees also had slightly anterior shift of lateral femoral condyle against lateral and medial tight group (p b 0.001, p = 0.045, respectively, Fig. 3b), with similar amount of translation. Resulting from these condylar translations, the well-balanced group exhibited constant femoral external rotation with similar amount compared with other two groups (p b 0.001, Fig. 3c). The medial tight group also demonstrated stable rotation with greater angle, whereas the lateral tight group showed decrease of femoral external rotation in deep flexion. No significant pair-wise difference at each knee angle was found between the three groups in medial and lateral femoral condylar translation and in femoral external rotation. In valgus/varus angle, the wellbalanced group exhibited more neutral valgus rotation and there were significant differences among the three groups (p b 0.001, Fig. 3d). Some pair-wise differences were detected through 70° to 95° knee flexion positions between the medial and lateral tight groups (p b 0.05). With regard to clinical range of motion, there were correlations between gap difference of 135° − 0° and preoperative flexion (r = 0.465, p = 0.004) and postoperative flexion (r = 0.479, p = 0.003) (Table 4). No correlation was demonstrated between joint tilting angle and preoperative and postoperative knee extension and flexion (Table 4). Also, there were no significant preoperative or postoperative extension or flexion differences among three tilting groups (Table 3). Clinical assessments were significantly improved after surgery, except the flexion angle (Table 1). There were strong positive correlations between preoperative and postoperative extension (r = 0.606, p b 0.001), and preoperative and postoperative flexion (r = 0.585, p b 0.001). The F subgroup showed the same significant findings as the entire study group relating preoperative and postoperative clinical data (Table 1). In correlations between gap differences and tilt, and postoperative knee kinematics, the F subgroup demonstrated more significant relations than those in the original group: The original group showed eight significant correlations between gap differences of 135° − 0°, 120° − 90°, 135° − 90° and kinematic values, while the F subgroup showed 13 significant correlations between these gap differences and kinematic parameters (Table 5). Similar to the original study group, the F subgroup showed no significant differences comparing kinematic values and clinical range of motion for the three tilting groups. In correlations between gap differences and tilt, and clinical range of motion, the F subgroup showed correlations between gap differences 120° − 0° and 120° − 90°, and preoperative flexion, in addition to the correlations observed in the original group. In kinematic pathway comparison the F subgroup exhibited the same significant differences, except the F subgroup showed no significant differences in medial condylar AP translation among three groups. The OA subgroup showed the same significant differences in preoperative and postoperative clinical data as the entire study group (Table 1). The OA subgroup demonstrated fewer correlations between gap differences and tilt, and postoperative knee kinematics than the entire study group: The whole study group showed eight significant correlations between gap differences of 135° − 0°, 120° − 90°, 135° − 90° and kinematic values, while the OA subgroup showed six significant correlations (Table 5). Similar to the whole study group, the OA subgroup showed no significant differences for comparisons of kinematic values and clinical range of motion among three tilting groups. In correlations between gap differences and tilt, and clinical range of motion, the OA subgroup showed the same significant differences as the whole study group. The OA subgroup exhibited the same significant differences in kinematic pathway comparisons, except the OA subgroup showed no significant differences in medial condylar AP translation among three groups.
2.5. Assessment of correlations and group comparisons The correlations between intraoperative measurements and postoperative kinematics were evaluated. The knee kinematics was also compared among three tilting groups. Furthermore, preoperative and postoperative clinical measurements were compared, and the correlations between intraoperative measurements and preoperative and postoperative range of motion were assessed to find out clinical relevance. We carried out subgroup analyses for the 36 female knees (32 OA knees and four RA knees, F subgroup) and 38 OA knees (32 female knees and six male knees, OA subgroup). Male and RA subgroups would contain only eight knees and six knees, respectively, which are inadequate for statistical analysis in this study. Paired Student's t-test was used to compare pre- and post-operative clinical measurements. Pearson's test was applied for correlation analyses. Analysis of variance (ANOVA) with post hoc Tukey test was used for tilting group comparisons. The probability value (p) below 0.05 was considered significant. Results are presented as mean ± standard deviation. 3. Results Several correlations are demonstrated between gap difference of 135° with respect to the zero degree reference and postoperative knee kinematics (Table 2). Gap difference of 135° − 0° showed positive correlations with total posterior translation of lateral femoral condyle (r = 0.336, p = 0.042) and total femoral external rotation (r = 0.488, p = 0.002, Fig. 2a) during squatting. Gap difference of 135° − 0° also demonstrated negative correlation with AP position of lateral femoral condyle at maximum kneeling (r = −0.510, p = 0.001, Fig. 2b), which has the same meaning as the positive correlation in squatting, because, in kneeling, negative value indicates posterior position of the femoral condyle, while in squatting, we took up the total amount of translations which are always positive values. Maximum knee flexion also had positive correlations with this gap difference (r = 0.355, p = 0.031). Other correlations were observed between deep flexion gap differences with respect to the 90° reference and postoperative knee kinematics (Table 2). Gap difference of 120° − 90° exhibited positive correlations with femoral external rotation at maximum lunge (r = 0.353, p = 0.019). Gap difference of 120° − 90° also showed negative correlation with AP position of lateral femoral condyle at maximum kneeling (r = −0.308, p = 0.042) and positive correlation with femoral external rotation at maximum kneeling (r = 0.492, p = 0.001) Gap difference of 135° − 90° had positive correlations with femoral external rotation at maximum lunge (r = 0.447, p = 0.006). No other correlation was found between joint gap differences and postoperative knee kinematics. No correlation was demonstrated between joint tilting angle and postoperative knee kinematics (Table 2). Also, there were no significant kinematic value differences among three tilting groups in squatting, lunge and kneeling activities (Table 3). However, if we look at the kinematic pathway during squatting, the well-balanced group showed significantly different kinematics compared with other two groups. The well-balanced knees demonstrated less anterior translation of medial femoral condyle in mid- to deep flexion (p = 0.004, p = 0.001 against lateral and medial tight group,
4. Discussion We investigated the relationships between intraoperative gap measurements and weight-bearing kinematics in 44 knees with primary
Table 2 Correlations between gap differences and tilt, and postoperative knee kinematics of 44 knees. Activities
Kinematics
Squatting
Total medal condylar AP translation (mm) Total lateral condylar AP translation (mm) Total femoral external rotation (°) Total valgus rotation (°) Flexion arc (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Maximum extension (°)
Lunge
Kneeling
Standing
p, r value (correlation) GD10-0
GD30-0
GD60-0
GD90-0
GD120-0
GD135-0
GD120-90
GD135-90
Tilt
0.936 0.114 0.063 0.655 0.104 0.907 0.783 0.672 0.666 0.855 0.120 0.316 0.430 0.441 0.189 0.482
0.939 0.819 0.657 0.859 0.350 0.904 0.892 0.948 0.493 0.694 0.311 0.377 0.125 0.891 0.190 0.884
0.756 0.840 0.919 0.410 0.622 0.707 0.676 0.564 0.863 0.830 0.207 0.548 0.086 0.402 0.285 0.614
0.696 0.670 0.530 0.504 0.283 0.571 0.894 0.767 0.968 0.456 0.219 0.743 0.179 0.352 0.181 0.995
0.776 0.171 0.067 0.886 0.133 0.516 0.402 0.177 0.300 0.621 0.582 0.273 0.413 0.192 0.152 0.709
0.788 0.042, r = 0.336 0.002, r = 0.488 0.987 0.169 0.878 0.100 0.112 0.279 0.257 0.243 0.001, r = −0.510 0.081 0.405 0.031, r = 0.355 0.488
0.827 0.232 0.136 0.400 0.715 0.073 0.350 0.019, r = 0.353 0.138 0.631 0.251 0.042, r = −0.308 0.001, r = 0.492 0.747 0.903 0.617
0.335 0.719 0.316 0.880 0.733 0.277 0.328 0.051 0.063 0.881 0.304 0.195 0.006, r = 0.447 0.618 1.000 0.728
0.162 0.451 0.426 0.342 0.154 0.485 0.969 0.667 0.546 0.272 0.242 0.527 0.663 0.106 0.084 0.976
GD: gap difference between two knee flexion angles, AP: anteroposterior.
T. Watanabe et al. / The Knee 22 (2015) 527–534
531
Fig. 2. (a) Gap difference of 135° minus zero degree showed positive correlations with total femoral external rotation (r = 0.488, p = 0.002) during squatting. (b) Gap difference of 135° minus zero degree demonstrated negative correlation with anteroposterior position of lateral femoral condyle in maximum kneeling (r = −0.510, p = 0.001).
PS-TKA in order to establish objective relationships between intraoperative variables and postoperative knee kinematics. Intraoperative joint gaps usually are measured with the patella everted or laterally shifted, which tightens lateral structures and reduces the effect of the extensor mechanism in deep flexion. This may preclude accurate evaluation of joint gaps and ML soft-tissue balance [5,8,20]. Also, intraoperative joint gaps generally are measured in extension and 90° flexion, which provides limited information in mid- and deep flexion. We measured joint gaps with the patella reduced and in smaller flexion intervals and examined the effect on postoperative kinematics. Consistent with our hypotheses, joint gap differences in deep flexion with respect to the zero degree or 90° references correlated with the total amount of posterior lateral condylar translation and femoral external rotation during squatting, femoral external rotation during lunge, posterior lateral translation, femoral external rotation, and maximum knee flexion during kneeling. The gap difference of 135° with respect to the zero degree reference also positively correlated preoperative and postoperative flexion. ML tilting did not have significant correlation with postoperative kinematics. Various kinematic values were not significantly different among three tilting group, either. However, dynamic kinematics during squatting demonstrated the differences of femoral condylar AP translation and femoral external rotation between the well-balanced knees and the other unbalanced knees. The valgus rotations were different
among three groups with the balanced knees in the most neutral rotation. There are a few limitations in this study. First, many factors including surgeries, postoperative treatment, and patients' conditions could affect arthroplasty outcomes, so focusing solely on the relationships between intraoperative laxity measurements and postoperative kinematics is challenging. However, we believe a single-surgeon series is appropriate to assess these relationships. For this study, all surgeries were performed by a single surgeon with the same procedures: surgical approach, bone cutting, soft tissue balancing, and other maneuvers. Having eliminated variation in surgical techniques to the extent possible [21], we focused on the effect of intraoperative joint gaps and ML soft tissue balance on postoperative kinematics. Second, our subject cohort consisted primarily of women with osteoarthritis. Our findings may not generalize to men or patients with rheumatoid arthritis. Also, we studied PS arthroplasties, where PCL resection enlarges the flexion gap and the post-cam mechanism controlled knee kinematics [22]. Arthroplasties that retain the PCL but lack the post-cam mechanism may exhibit different characteristics. Previous studies have shown larger joint gaps at 90° or greater flexion predicted larger postoperative knee flexion [20,23,24]. Different studies have reported greater posterior femoral translation has a positive correlation with greater maximum knee flexion [14,15]. Our results
Table 3 Comparison of kinematic values and clinical range of motion among three tilting groups.
Squatting
Lunge
Kneeling
Standing Clinical range of motion
AP: anteroposterior.
Total medal condylar AP translation (mm) Total lateral condylar AP translation (mm) Total femoral external rotation (°) Total valgus rotation (°) Flexion arc (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm) Femoral external rotation (°) Valgus rotation (°) Maximum flexion (°) Maximum extension (°) Preoperative extension Preoperative flexion Postoperative extension Postoperative flexion
Lateral tight (n = 14)
Well-balanced (n = 13)
Medial tight (n = 17)
p-Value (ANOVA)
7±2 8±3 10 ± 3 3±1 114 ± 12 −3 ± 4 −10 ± 3 9±6 1±1 120 ± 10 −5 ± 4 −9 ± 3 5±6 1±1 127 ± 9 −4 ± 6 −10 ± 10 123 ± 15 −1 ± 2 128 ± 11
8±1 9±4 9±3 3±1 117 ± 25 −5 ± 3 −10 ± 4 7±4 1±1 123 ± 17 −6 ± 3 −9 ± 4 4±4 1±1 125 ± 16 −10 ± 9 −5 ± 11 126 ± 20 −1 ± 3 130 ± 19
7±2 8±3 9±3 3±1 108 ± 19 −3 ± 3 −10 ± 4 9±6 0±1 117 ± 15 −4 ± 4 −9 ± 4 7±5 0±2 119 ± 16 −5 ± 5 −6 ± 8 122 ± 24 −1 ± 2 125 ± 15
0.304 0.838 0.896 0.559 0.266 0.886 0.824 0.449 0.412 0.480 0.274 0.894 0.380 0.213 0.331 0.096 0.419 0.836 0.905 0.630
532
T. Watanabe et al. / The Knee 22 (2015) 527–534
Fig. 3. (a) The well-balanced knees demonstrated less anterior translation of medial femoral condyle in mid- to deep flexion (p = 0.004, p = 0.001 against lateral and medial tight group, respectively). (b) The well-balanced knees also had slightly anterior shift of lateral femoral condyle against lateral and medial tight group (p b 0.001, p = 0.045, respectively), with similar amount of translation. (c) The well-balanced group exhibited constant femoral external rotation with similar amount compared with other two groups (p b 0.001 against lateral and medial tight group). (d) The well-balanced group showed more neutral valgus rotation and there were significant differences among the three groups (p b 0.001). Some pair-wise differences were detected at 70° to 95° knee flexion positions between the medial and lateral tight groups (p b 0.05).
are consistent with these previous studies. Our study also demonstrated correlations between joint gap differences in deep flexion with respect to the zero degree and 90° references and total lateral posterior translations, femoral external rotation, and lateral condylar AP positions. These relationships between joint gaps and postoperative kinematics have not been well discussed. Greater deep flexion gaps, compared to gaps at zero degree and 90°, are indicative of extensor mechanism flexibility. In very deep flexion, e.g. 135°, joints with a still-flexible extensor mechanism could show greater knee flexion and more natural kinematics, including femoral posterior translation and external rotation. It is well known that preoperative range of flexion correlates postoperative flexion angle [25,26]. Preoperative soft tissue condition around the knee including extensor mechanism should affect postoperative knee flexion. Current study demonstrated positive correlations between
preoperative and postoperative range of flexion, and joint gap differences in deep flexion. Large deep flexion gap differences with respect to the zero degree and 90° references would reflect more extensible soft tissue which is closely related to the postoperative better range of flexion. ML soft tissue balance is an important factor for successful TKA [1,2]. Failure to achieve adequate ML balance is a predictor of poor outcomes [27]. Sasanuma et al. reported intraoperative soft tissue balance after releases averaged 2.1° in extension and − 1.6° at 90° flexion [28]. We found an average 0.0° ML tilt in this study, consistent with previous reports [9,28]. Perfect balance is difficult to achieve [3], and it has been questioned whether rectangular gaps are ideal since normal gaps are trapezoidal [29]. In current study, we did not detect any correlation between ML balance in the range of several degrees and knee kinematics
Table 4 Correlations between gap differences and tilt, and clinical range of motion. Clinical range of motion
Preoperative extension Preoperative flexion Postoperative extension Postoperative flexion
p, r value (correlation) GD10-0
GD30-0
GD60-0
GD90-0
GD120-0
GD135-0
GD120-90
GD135-90
Tilt
0.138 0.540 0.597 0.255
0.185 0.711 0.951 0.510
0.100 0.798 0.342 0.491
0.472 0.359 0.624 0.278
0.799 0.055 0.579 0.133
0.698 0.004, r = 0.465 0.563 0.003, r = 0.479
0.443 0.251 0.125 0.725
0.613 0.144 0.902 0.285
0.302 0.983 0.921 0.505
GD: gap difference between two knee flexion angles.
T. Watanabe et al. / The Knee 22 (2015) 527–534
533
Table 5 Correlations between gap differences and tilt, and postoperative knee kinematics of two subgroups. Activities
Kinematics
p, r value (correlation) Female (n = 36)
Squatting
Lunge
Kneeling
Standing
Osteoarthritis (n = 38)
GD120-0
GD135-0
GD120-90
GD135-90
GD120-0
GD135-0
GD120-90
GD135-90
Total medal condylar AP translation (mm) Total lateral condylar AP translation (mm)
0.760 0.306
0.914 0.088
0.365 0.424
0.974 0.182
0.492 0.093
0.511 0.367
0.208 0.991
Total femoral external rotation (°)
0.058
0.314
0.059
0.332
0.819 0.099 0.594
0.898 0.930 0.317
0.853 0.164 0.440
0.003, r = 0.519 0.807 0.303 0.778
0.081
Total valgus rotation (°) Flexion arc (°) Medial condylar AP translation (mm)
0.003, r = 0.517 0.810 0.176 0.963
0.700 0.029, r = 0.364 0.025, r = 0.374 0.469 0.250 0.092
0.932 0.526 0.219
Lateral condylar AP translation (mm)
0.232
0.051
0.135
0.315
0.134
0.515 0.954 0.028, r = 0.356 0.375
Femoral external rotation (°)
0.206
0.165
0.087
0.090
0.623
0.566
0.274
0.394
0.005, r = 0.444 0.127
0.064
Valgus rotation (°) Maximum flexion (°) Medial condylar AP translation (mm) Lateral condylar AP translation (mm)
0.423 0.678 0.162
0.663 0.792 0.286
0.317 0.160 0.100
0.597 0.268 0.449
Femoral external rotation (°)
0.225
Valgus rotation (°) Maximum flexion (°)
0.929 0.127
0.019, r = 0.418 0.795 0.659
Maximum extension (°)
0.722
0.226 0.289 b0.001, r = −0.630 0.040, r = 0.371 0.848 0.034, r = 0.381 0.473
0.037, r = 0.376 0.034, r = −0.382 0.851 0.456 0.103
0.642
0.028, r = −0.365 0.002, r = 0.499 0.194 0.991 0.374 0.028, r = −0.365 0.002, r = 0.495 0.604 0.789
0.012, r = 0.445 0.943 0.911
0.306
0.362 0.376 0.006, r = −0.487 0.107
0.203 0.195
0.312 0.086
0.001, r = 0.509 0.625 0.498
0.053
0.940
0.878
0.380
0.885
0.522
0.103
GD: gap difference between two knee flexion angles, AP: anteroposterior.
in various activities. However, some kinematic differences among three tilting groups were demonstrated. Normal and osteoarthritic knee kinematics show continuous and smooth femoral external rotation with knee flexion [30,31]. Femoral external rotation and Condylar posterior translation are the greatest in normal knees, followed by osteoarthritic knees, and the least in knees with an arthroplasty. In this study, the well-balanced group showed less medial condylar anterior translation in mid-flexion and linear femoral external rotation with similar amount compared with the other two groups. The valgus/varus rotation reflected the ML soft tissue balance with the well-balanced knees in the most neutral. Female knees showed results similar to the entire study group, with additional significant relations between gap differences and postoperative knee kinematics, and between gap differences and clinical range of motion. Female joints generally exhibit more flexibility than male joints [32], which would affect joint kinematics. Greater soft tissue elasticity in female knees could enhance the influence of joint gap on the postoperative knee kinematics. The OA subgroup knees exhibited results similar to the whole study group, but fewer correlations between gap differences and postoperative knee kinematics. Both OA and RA knees exhibit greater AP laxity, restricted femoral axial rotation, and an increased valgus/varus rotation compared to normal knees, although AP laxity decreases with disease severity only in OA knees [33]. Unfortunately, our study cohort had too few RA knees to permit direct comparison of OA and RA knees, so conclusions relating disease and knee mechanics cannot be made from our data. Our study demonstrated gap differences of 135° − 0° was correlated with the total posterior translation of lateral femoral condyle and femoral external rotation during squatting, lateral condylar AP position and maximum knee flexion in kneeling. Gap differences of 120° − 90° and 135° − 90° also exhibited similar correlations. Gap difference of 135° − 0° was also correlated with preoperative and postoperative knee flexion. ML tilting affect postoperative kinematic pathway during squatting. Well-balanced knees showed less anterior translation of medial femoral condyle in mid- to deep flexion, consistent femoral external rotation, and the most neutral valgus/varus rotation. These
findings suggest adequate intraoperative joint gap differences beyond 90° flexion with respect to the zero degree or 90° references are important to obtain greater lateral femoral translation, femoral external rotation, and better postoperative flexion. Also, ML imbalance negatively affects knee kinematic pathway during weight-bearing activity, indicating the importance of soft tissue balancing.
Acknowledgments We thank Hirotsugu Muratsu MD, for the advice on using the intraoperative joint gap tensor developed at Kobe University.
References [1] Kurosaka M, Yoshiya S, Mizuno K, Yamamoto T. Maximizing flexion after total knee arthroplasty: the need and the pitfalls. J Arthroplasty 2002;17:59–62. [2] Scuderi G. The stiff total knee arthroplasty causality and solution. J Arthroplasty 2005;20:23–6. [3] Griffin FM, Insall JN, Scuderi GR. Accuracy of soft tissue balancing in total knee arthroplasty. J Arthroplasty 2000;15:970–3. [4] Sambatakakis A, Attfield SF, Newton G. Quantification of soft-tissue imbalance in condylar knee arthroplasty. J Biomed Eng 1993;15:339–43. [5] Matsumoto T, Muratsu H, Tsumura N, Mizuno K, Kuroda R, Yoshiya S, et al. Joint gap kinematics in posterior-stabilized total knee arthroplasty measured by a new tensor with the navigation system. J Biomech Eng 2006;128:867–71. [6] Winemaker MJ. Perfect balance in total knee arthroplasty: the elusive compromise. J Arthroplasty 2002;17:2–10. [7] Wallace AL, Harris ML, Walsh WR, Bruce WJ. Intraoperative assessment of tibiofemoral contact stresses in total knee arthroplasty. J Arthroplasty 1998;13: 923–7. [8] Yoshino N, Watanabe N, Watanabe Y, Fukuda Y, Takai S. Measurement of joint gap load in patella everted and reset position during total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2009;17:484–90. [9] Moro-oka TA, Shiraishi H, Iwamoto Y, Banks SA. Modified gap-balancing technique in total knee arthroplasty: evaluation of the post-operative coronal laxity. Knee Surg Sports Traumatol Arthrosc 2010;18:375–80. [10] Banks SA, Harman MK, Bellemans J, Hodge WA. Making sense of knee arthroplasty kinematics: news you can use. J Bone Joint Surg Am 2003;85-A(Suppl. 4):64–72. [11] Mikashima Y, Tomatsu T, Horikoshi M, Nakatani T, Saito S, Momohara S, et al. In vivo deep-flexion kinematics in patients with posterior-cruciate retaining and anteriorcruciate substituting total knee arthroplasty. Clin Biomech (Bristol, Avon) 2010;25: 83–7.
534
T. Watanabe et al. / The Knee 22 (2015) 527–534
[12] Okamoto N, Breslauer L, Hedley AK, Mizuta H, Banks SA. In vivo knee kinematics in patients with bilateral total knee arthroplasty of 2 designs. J Arthroplasty 2011;26: 914–8. [13] Cates HE, Komistek RD, Mahfouz MR, Schmidt MA, Anderle M. In vivo comparison of knee kinematics for subjects having either a posterior stabilized or cruciate retaining high-flexion total knee arthroplasty. J Arthroplasty 2008;23:1057–67. [14] Banks S, Bellemans J, Nozaki H, Whiteside LA, Harman M, Hodge WA. Knee motions during maximum flexion in fixed and mobile-bearing arthroplasties. Clin Orthop Relat Res 2003:131–8. [15] Victor J, Banks S, Bellemans J. Kinematics of posterior cruciate ligament-retaining and -substituting total knee arthroplasty: a prospective randomised outcome study. J Bone Joint Surg (Br) 2005;87:646–55. [16] Itokazu M, Masuda K, Wada E, Ohno T, Yoshida M, Takatu T. Influence of anteroposterior and mediolateral instability on range of motion after total knee arthroplasty: an ultrasonographic study. Orthopedics 2000;23:49–52. [17] Warren PJ, Olanlokun TK, Cobb AG, Walker PS, Iverson BF. Laxity and function in knee replacements. A comparative study of three prosthetic designs. Clin Orthop Relat Res 1994:200–8. [18] Banks SA, Hodge WA. Accurate measurement of three-dimensional knee replacement kinematics using single-plane fluoroscopy. IEEE Trans Biomed Eng 1996;43:638–49. [19] Banks SA, Harman MK, Hodge WA. Mechanism of anterior impingement damage in total knee arthroplasty. J Bone Joint Surg Am 2002;84-A(Suppl. 2):37–42. [20] Matsumoto T, Mizuno K, Muratsu H, Tsumura N, Fukase N, Kubo S, et al. Influence of intra-operative joint gap on post-operative flexion angle in osteoarthritis patients undergoing posterior-stabilized total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2007;15:1013–8. [21] Nozaki H, Banks SA, Suguro T, Hodge WA. Observations of femoral rollback in cruciate-retaining knee arthroplasty. Clin Orthop Relat Res 2002:308–14. [22] Yagishita K, Muneta T, Ikeda H. Step-by-step measurements of soft tissue balancing during total knee arthroplasty for patients with varus knees. J Arthroplasty 2003;18: 313–20. [23] Higuchi H, Hatayama K, Shimizu M, Kobayashi A, Kobayashi T, Takagishi K. Relationship between joint gap difference and range of motion in total knee arthroplasty: a
[24]
[25] [26]
[27]
[28] [29] [30]
[31]
[32]
[33]
prospective randomised study between different platforms. Int Orthop 2009;33: 997–1000. Takayama K, Matsumoto T, Kubo S, Muratsu H, Ishida K, Matsushita T, et al. Influence of intra-operative joint gaps on post-operative flexion angle in posterior cruciateretaining total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2012;20: 532–7. Kawamura H, Bourne RB. Factors affecting range of flexion after total knee arthroplasty. J Orthop Sci 2001;6:248–52. Ritter MA, Harty LD, Davis KE, Meding JB, Berend ME. Predicting range of motion after total knee arthroplasty. Clustering, log-linear regression, and regression tree analysis. J Bone Joint Surg Am 2003;85-A:1278–85. Wasielewski RC, Galat DD, Komistek RD. Correlation of compartment pressure data from an intraoperative sensing device with postoperative fluoroscopic kinematic results in TKA patients. J Biomech 2005;38:333–9. Sasanuma H, Sekiya H, Takatoku K, Takada H, Sugimoto N. Evaluation of soft-tissue balance during total knee arthroplasty. J Orthop Surg (Hong Kong) 2010;18:26–30. Tokuhara Y, Kadoya Y, Nakagawa S, Kobayashi A, Takaoka K. The flexion gap in normal knees. An MRI study. J Bone Joint Surg (Br) 2004;86:1133–6. Moro-oka TA, Hamai S, Miura H, Shimoto T, Higaki H, Fregly BJ, et al. Dynamic activity dependence of in vivo normal knee kinematics. J Orthop Res 2008;26: 428–34. Hamai S, Moro-oka TA, Miura H, Shimoto T, Higaki H, Fregly BJ, et al. Knee kinematics in medial osteoarthritis during in vivo weight-bearing activities. J Orthop Res 2009; 27:1555–61. Shultz SJ, Shimokochi Y, Nguyen AD, Schmitz RJ, Beynnon BD, Perrin DH. Measurement of varus-valgus and internal–external rotational knee laxities in vivo—part II: relationship with anterior–posterior and general joint laxity in males and females. J Orthop Res 2007;25:989–96. Wada M, Imura S, Baba H, Shimada S. Knee laxity in patients with osteoarthritis and rheumatoid arthritis. Br J Rheumatol 1996;35:560–3.