In Far East, including Japan and the Middle East, daily activities are frequently carried out on the floor. Deep flexion of the knee joint is therefore very important in these societies. Some patients who underwent total knee arthroplasty (TKA) in these countries often perform deep flexion activity, such as squatting, cross-leg sitting and kneeling. However it is still unknown that deep flexion activity affects long term durability after TKA. The purpose of this study was to examine the correlation between deep flexion and long term durability.

Between December 1989 and May 1997, 507 total knee arthroplasties were carried out in 371 patients using the Bi-Surface Knee System (Japan Medical Material, Osaka, Japan) at two institutions and routine rehabilitation program continued for one to two months after TKA. One patient who underwent simultaneous bilateral TKA was excluded because of pulmonary embolism within one month. The other 505 knees (370 patients) were divided into two groups according to the range of flexion after our routine rehabilitation program; one group (Group A: 207 knees) consisted of more than 135 degrees flexion knees and the other group (Group B: 298 knees) consists of less than 135 degrees flexion knees. Patients whose follow-up period was less than 10 years were excluded from this clinical evaluation. Range of flexion was measured preoperatively, at the time after routine rehabilitation program, and at the latest follow-up. Knee function was evaluated on the basis of Knee Society knee score and functional score preoperatively and at the latest follow-up. Kaplan-Meier survivorship analysis was performed with revision for any operation as the end point.

In Group A, the mean preoperative range of flexion was 133.0±16.3 degrees, and at the time after routine rehabilitation program, this improved to 139.7±5.1 degrees. This angle maintained to 136.2±14.3 at the latest follow-up. In Group B, the mean preoperative range of flexion was 111.6±20.4 degrees, and at the time after routine rehabilitation program, this improved to 114.5±13.6 degrees. This angle maintained to 118.2±17.8 at the latest follow-up. The Knee Society knee score and functional score was improved from 43.0±16.9 points and 39.0±20.2 points preoperatively to 95.1±5.8 points and 51.8±21.2 points at the latest follow-up, respectively in Group A. The Knee Society knee score and functional score was improved from 37.1±16.7 points and 31.9±18.4 points preoperatively to 92.5±8.7 points and 53.1±26.1 points at the latest follow-up, respectively in Group B. Kaplan-Meier survivorship at 10-year was 95.5% in Group A and 96.2% in Group B with any operation as the end point. The survivorship between Group A and Group B was not statistically significant.

Good range of flexion was maintained and Knee society score was excellent after a long time follow-up for the patients who achieved deep flexion after TKA. Deep flexion was proved not to affect long term durability in this Bi-Surface Knee System.

Binary Surface type knee prosthesis (bisurface knee) has successfully been utilized in total knee arthroplasty (TKA) in order to improve flexional motion, especially, deep flexion. Binary surface means that the knee prosthesis has two different bearing structures, that is, normal condylar surfaces and ball-socket structure. The ball and the socket are placed between the condylar surfaces of the femoral component and the tibial insert, respectively. Two different designs of bisurface knee have been proposed so far and only one model called KU has been utilized in clinical applications. The other model called CFK is still under development and characterized to have a post-cam structure to stabilize the knee motion. These bisurface knees are expected to attain deep flexional motion and therefore, it is important to understand their safety and durability at high flexion angles. In the present study, the finite element analysis (FEA) is conducted to characterize the mechanics of the bisurface knees under deep knee flexion. Risk assessment of the bisurface knees are then performed based on the FEA results.

Detailed 3D-FEA models are constructed using CAD data and deep knee flexion corresponding to a squatting motion is reproduced by using spring models and proper boundary conditions. The spring models attached to the tibial component are used to express the mechanical effects of soft tissues. Internal rotational motion is also considered with the flexional motion. The femoral and the tibial components are assumed to be rigid and the tibial insert made of UHMWPE is an elastic-plastic solid having a nonlinear constitutive relation determined from experiments. The femoral component is rotated continuously from 0° to 135° to express the flexional motion and the tibial component is also rotated to express internal rotation.

The equivalent stress of the condylar surface of the new CFK model is almost equivalent to that of the KU model during flexion from 0° to 90°, however, the stress values are different at the angles higher than 90°. At higher angles of flexion than 90°, the bearing surface of the KU consists of the condylar and the socket surfaces, while the bearing surface of the CFK consists of the socket surface only. Therefore, the CFK exhibits higher stress than the KU at these high angles. The ball-socket bearing system enables these bisurface knees to be adapted to deep flexional motion. The CFK is trying to achieve higher flexion angles than the KU by employing the modified ball-socket bearing structure, however, higher stress concentration on the socket surface of the CFK may hasten degradation of the tibial insert. It is also found that the stress concentration on the socket surfaces increase with increase of the internal rotation angle and therefore, the risk of damage of the tibial insert becomes higher with internal rotation.

In summary, 3D dynamic FEA is utilized to make a risk assessment of the bisurface knees and the computational results suggest that the design of the ball-socket structure is one of the most important factors to determine the safety and durability of the knees.

The objective of this study is to determine the in-vivo knee joint kinematics of patients having specially designed knee prosthesis (Bi-surface) at sitting sedentary (seiza) state.

An increase in the demand for TKA has required improvement in the durability and flexibility of knee prostheses. One of the representative cases which have improved knee flexion is Bi-surface knee. Bi-surface knee has two joint surfaces; one for weight bearing and the other for flexion motion which has a unique ball-and-socket joint. This knee prosthesis, having been applied for two decades, has not yet been precisely analyzed how the femoral and tibial components are articulating at deep knee flexion.

Since there is no practical method to measure directly prosthetic kinematics in-vivo; we applied indirect techniques, pattern matching method to the Bi-surface patients. The method has been originated by Banks and Hodge (1964), and we have improved it in order to obtain higher and more reliable accuracies.

The number of subjects examined by X-ray apparatus was 18 knees of 14 patients (3 male and 11 female) who could attain the seiza. Patients were asked to sit at seiza state and their Bi-surface knees were X-ray photographed from lateral side. We focused if the internal rotation was shown at maximum flexion as commonly shown for a normal knee. We also represented the CAD models with the same position/orientation as the data from the pattern matching, thereby investigating the contact states between the ball and socket by viewing them from the desired direction.

The following results were introduced. The mean maximum flexion angle was 144.1° (SD=5.3°), and the mean internal rotation angle at maximum flexion was 15.2° (SD=6.6). The maximum flexion angle among all subjects was 153.3° and internal rotation was 19.5° at that flexion angle. The number of subjects which had (a) contact point(s) on the tibio-femoral and/or ball-socket surface(s) was 5 knees (2 knees had contact point on both the ball-and-socket and the tibio-femoral lateral surfaces, 3 knees had only on either surface) and the other 13 knees had a slight gap between two components.

Correlation was found between the value of the maximum flexion angle and the value of internal rotation angle at that flexion; the subjects of larger maximum flexion angle also demonstrated larger internal rotation angle. This suggests that at deep knee flexion, the tibial internal rotation may play an important role after TKA as a normal knee does. By checking the CAD representations, we found that the tibio-femoral and ball-socket surfaces were separate for most subjects at seiza state. Although serious impingements were not found, it was suggested the risk of subluxation when a patient rises up.

The limitation of our study is that we used simple still X-ray pictures. In order to assess kinematics for ascending from seiza state, kinematic analyses from fluoroscopic images are needed.