In vivo flexion and kinematics of the knee after TKA: comparison of a conventional and a high flexion cruciate-retaining TKA design
- First Online:
- Cite this article as:
- Suggs, J.F., Kwon, YM., Durbhakula, S.M. et al. Knee Surg Sports Traumatol Arthrosc (2009) 17: 150. doi:10.1007/s00167-008-0637-4
- 324 Downloads
This study investigated the in vivo 6DOF knee kinematics and tibiofemoral contact location after total knee arthroplasty using a conventional and a high flexion cruciate retaining component (15 NexGen CR, 11 NexGen CR-Flex). Each patient performed a single-leg lunge while being imaged by a dual fluoroscopic imaging system. Data were analyzed at hyperextension, 0° to 90° in 15° intervals, and at maximum flexion. The average maximum weight-bearing flexion for all the CR patients was 110.1° ± 13.4°, and for all the CR-Flex patients was 108.2° ± 13.2°. No difference was seen in the maximum flexion achieved by the patients, and the kinematics demonstrated by the groups was similar. However, at high flexion, the tibiofemoral articulating surfaces were more conforming in the CR-Flex design than the CR design, suggesting that the use of the high flexion component improved the tibiofemoral contact environment at high flexion in patients who could achieve high flexion.
KeywordsTotal knee arthroplasty High flexion In vivo kinematics Biomechanics Knee
Restoration of the full range of knee flexion after total knee arthroplasty (TKA) is important to patients in maintaining various life style activities, such as sports, gardening, stair ascent/descent, and taking a bath [9, 12, 21]. It is believed that contemporary TKA patients are more active than patients are in the past and have a greater desire to participate in activities that require high flexion. Consequently, many new TKA components have been designed to better accommodate high knee flexion after surgery. It has been suggested that the mechanical environment experienced by the polyethylene insert at high flexion may be highly unfavorable and that participation in high flexion activities could accelerate wear of the polyethylene component [17, 20].
Several studies have evaluated high flexion TKA designs using either clinical examination or single-plane fluoroscopic techniques [2, 6, 7, 10, 11, 22]. These in vivo studies have only dealt with posterior-substituting designs. No study has reported on the biomechanics of high flexion posterior cruciate-retaining TKA designs. Further, no study has compared the in vivo contact biomechanics of high flexion TKAs with those of conventional TKA designs.
The objective of this study was to compare the in vivo kinematics of two cruciate retaining TKA designs, one conventional design (NexGen CR, Zimmer, Warsaw, IN) and one high flexion design (NexGen CR-Flex, Zimmer, Warsaw, IN). We hypothesized that the CR-Flex design would enhance knee flexion compared to the conventional CR design. Six degree-of-freedom kinematics was obtained from patients implanted with either the conventional component or the high flexion component using a dual fluoroscopic imaging system. Information on maximum knee flexion and the contact location between the femoral component and the polyethylene insert were also compared between the two designs.
Materials and methods
In vivo experimental design
Demographics for CR and CR-Flex groups
69.1 ± 10.9
66.6 ± 11.2
195.1 ± 31.0
193.1 ± 41.0
69.6 ± 3.0
68.5 ± 3.4
Postop. time (months)
15.6 ± 7.5
10.7 ± 3.3
Clinical results for CR and CR-Flex groups
Passive ROM (degree)
120.6 ± 11.1
117.7 ± 9.7
Max weight-bearing flexion (degree)
110.1 ± 13.4
108.2 ± 13.2
Weight-bearing ROM (degree)
118.5 ± 15.7
115.1 ± 17.0
Knee society knee score
91.4 ± 13.2
89.5 ± 11.5
Knee society functional score
84.9 ± 14.4
84.7 ± 15.0
The positions of the total knee components during the weight-bearing flexion were deduced with the use of a virtual dual fluoroscopic imaging system created in solid modeling software (Rhinoceros®, Robert McNeel and Associates, Seattle, WA), where the image intensifiers were represented by the acquired fluoroscopic images, and the X-ray sources were represented by two virtual cameras . Solid models of the TKA components were imported into the virtual fluoroscopic system. The component models were manipulated in 6DOF until they overlapped their silhouettes on both fluoroscopic images, as seen from their respective cameras. When the models overlapped their silhouettes, the in vivo pose at the time of image acquisition was recreated. By repeating this process for each pair of fluoroscopic images, the in vivo positions of the total knee components along the flexion path were represented by a series of 3D total knee models.
The tibiofemoral contact location was determined by calculating the centroid of the overlap between the femoral component and the polyethylene surfaces in the medial and lateral compartments . If no overlap was present, the point on the polyethylene surface nearest to the femoral condyle was used as the contact location. A previous study has shown that the imaging system has an accuracy of 0.16 mm for the femoral component and 0.13 mm for the tibial component, so lift-off was defined as the closest distance between the polyethylene and the femoral condyle being greater than 0.29 mm .
To quantitatively describe the contact locations, two coordinate systems were created for the articular contact in the medial and lateral compartments. The origins were midway between the anterior–posterior extremes of the polyethylene insert and 25% of the insert’s medial–lateral dimension from the medial–lateral extremes.
Component geometry and surgical technique
Patients in each group were averaged at hyperextension, in 15° intervals from 0° to 90° of flexion, and at maximum flexion of the implant . The reported data at hyperextension and maximum flexion only included patients who achieved greater than 3° of hyperextension or 100° of flexion, respectively. A student’s t test with Bonferroni correction was used to compare the maximum flexion, 6DOF kinematics data, and the contact locations in the medial and lateral compartments between the CR group and the CR-Flex group. Differences were considered significant when P < 0.05.
The passive ROM averaged 120.6° ± 11.1° for all the CR patients and 117.7° ± 9.7° for all the CR-Flex patients (Table 2). There was no significant difference between the two patient groups in passive ROM. During weight-bearing flexion, the average maximum flexion for all the CR patients was 110.1° ± 13.4°, and the average maximum flexion for all the CR-Flex patients was 108.2° ± 13.2°. There was no difference in maximum weight-bearing flexion between the two patient groups during the weight-bearing flexion.
6DOF kinematics of CR and CR-Flex TKA patients
The patient groups exhibited similar patterns of medial–lateral femoral translation. In the CR patients, the femoral component moved laterally from 0.5 ± 0.7 mm at hyperextension to −0.8 ± 0.9 mm at 45° of flexion and then medially to 0.1 ± 1.7 mm at maximum flexion. In the CR-Flex group, the femoral component moved laterally from −0.1 ± 0.6 mm at full extension to −1.5 ± 0.9 mm at 45° of flexion and then medially to 0.5 ± 1.6 mm at maximum flexion. No statistically significant differences were found between the groups.
Tibiofemoral contact kinematics of CR and CR-Flex patients
In the medial–lateral direction, the lateral compartment contact of the CR group gradually moved laterally from −3.5 ± 3.6 mm at hyperextension to −7.9 ± 6.3 mm at maximum flexion. For the CR-Flex patients, the contact also moved laterally from −3.5 ± 8.1 mm at hyperextension to −5.9 ± 7.9 mm at maximum flexion. There was no difference in lateral compartment contact location between the two patient groups.
In the medial compartment, the contact location in the anterior–posterior direction remained relatively constant with flexion until maximum flexion for both groups. In the CR group, the medial compartment contact occurred at −2.0 ± 4.0 mm throughout the flexion range until maximum flexion, where the contact moved to −5.4 ± 9.1 mm. The medial compartment contact in the CR-Flex group remained at −2.0 ± 3.5 mm through early and mid-flexion and reached −4.2 ± 6.9 mm at maximum flexion. There was no difference between the groups in the AP location of the medial compartment contact.
The medial compartment contact location was also relatively constant in the medial–lateral direction throughout the entire flexion range for both groups. In the CR group, the contact remained around 4.5 ± 4.5 mm throughout the range of flexion. In the CR-Flex group, the contact was around 2.3 ± 5.0 mm. Again, no difference was found between the CR and CR-Flex knees.
Observation of tibiofemoral contact patterns of CR and CR-Flex patients
Lift-off occurred at maximum flexion in five knees in the CR group and three knees in the CR-Flex group. In the CR group, there were three knees with lift-off only in the lateral compartment, one knee with lift-off only in the medial compartment, and one knee with lift-off in both compartments. In the CR-Flex group, there was one knee with lift-off only in the lateral compartment, one knee with lift-off in the medial compartment, and one knee with lift-off in both compartments. The average maximum flexion for these eight knees with lift-off was 112.8 ± 13.7, which was 5° greater but not statistically different from that of knees with no lift-off (107.7 ± 12.9).
Despite the debate over the need and efficacy of high flexion components [17, 18, 19, 20], many new components have been used clinically with the aim of enhancing the flexion capability of the knee after TKA . However, previous studies have only compared high flexion TKA designs to conventional designs in a posterior substituting knee [2, 6, 7, 10, 11, 22]. This study investigated either the 6DOF knee kinematics of patients after TKA using a conventional cruciate retaining component (NexGen, CR) or a high flexion cruciate retaining component (NexGen, CR-Flex).
In this study, patients with specially designed high flexion components behaved similarly kinematically to those with conventional implants. There was no difference in posterior femoral translation throughout the entire flexion range. For the CR patients, the tibiofemoral contact moved 4.5 mm posteriorly in the medial compartment and 13.0 mm in the lateral compartment during the weight-bearing lunge, indicating a “medial pivot” motion of the knee during flexion. For the CR-Flex patients, the tibiofemoral contact moved 5.1 mm posteriorly in the medial compartment and 8.1 mm in the lateral compartment. There were no dramatic differences in the contact positions during knee flexion between the two patient groups. The CR-Flex knees showed approximately 3° greater internal tibial rotation than the CR knees throughout the flexion range, although this difference was not statistically significant.
Besides the similarity in kinematics, the two patient groups had similar flexion under both passive and weight-bearing conditions. However, the tibiofemoral contact behavior was different between the components at high flexion angles (>120°). Figure 6 showed the tibiofemoral contact patterns of a CR patient and a CR-Flex patient at 130° of flexion. At this flexion angle, the condylar tip of the conventional CR TKA was in contact with the polyethylene surface. This could cause a stress concentration on the polyethylene surface and lead to increased wear in patients who attain high flexion. However, at the same flexion angle, the articulating surface of the CR-Flex component was much more conforming compared to the conventional CR design. The increased conformity would help to reduce any potential high stresses experienced by the polyethylene at high flexion. This improvement in contact can be explained by the thicker posterior femoral condyle of the CR-Flex design (Fig. 2). The increased thickness of the femoral condyle allows for a larger radius of curvature at higher flexion angles, which translates into more conforming surfaces between femoral and polyethylene components at high flexion. Therefore, this high flexion total knee design seems to have improved the articular contact mechanics when the knee is able to achieve high flexion. This observation supports a previous prediction based on radiographs at full flexion, which suggested that the high flex designs had better contact area .
It should be noted that the data obtained in the present study for the conventional implant is similar to published data for other cruciate-retaining components. Previous studies have reported passive maximum flexion values between 100° and 120° [1, 3, 4, 5, 25]. The current study found an average weight-bearing maximum flexion of 110° and a mean passive ROM around 120°, which are within the range of the data reported in literature.
In the literature, most studies on high flexion TKA patients consist of Asian cohorts and focus on the passive range of motion of PS TKA designs [6, 7, 10, 11, 22, 23]. There are inconsistent conclusions when comparing the flexion capability of patients with conventional implants and high flexion implants. For example, Bin et al.  compared 90 conventional LPS knees to 90 matched LPS-Flex (Zimmer) knees at 1 year postoperatively. They found the LPS-Flex knees to have more ROM (129.8° ± 5.2°) than the conventional knees (124.3° ± 9.2°). Huang et al.  also found LPS-Flex knees to have about 10° more flexion than LPS knees at 2 years follow-up. Gupta et al.  compared a conventional rotating platform posterior stabilized design (P.F.C. Sigma RP, Depuy) to high flexion version of the same component (P.F.C. Sigma RP-F). They reported that the patients with a high flexion design gained significantly more ROM from preop to postop (110°–125°) than the patients with the conventional design (110°–116°). Kim et al.  compared LPS-Flex to LPS in 50 bilateral patients and did not find a difference in ROM between the components (139 vs. 136°). Seon et al.  compared LPS-Flex to a mobile bearing CR design (e-motion, B. Braun-Aesculap) and found no difference in maximum flexion (131 vs. 129°). Few studies have reported the maximum flexion of PS TKAs during weight-bearing flexion. In a recent report, we found a maximum weight-bearing flexion of 113° in a group of patients replaced with a LPS-Flex component . The patients with CR or CR Flex components investigated in this study were found to have a similar maximal flexion compared to those patients with LPS Flex components.
One limitation of the current study is that the condition of the polyethylene surface that could not be directly analyzed due to the in vivo nature of the experiment. As contact behavior was revealed different between the two-implant designs at high flexion, it would be clinically interesting to examine the wear modes and patterns of their polyethylene components. In the future, this can be studied using retrieved polyethylene components from revision patients who used one of the two implants.
In conclusion, the kinematics of the CR-Flex patients analyzed in this study was similar to those of the patients with a conventional CR component. No difference was seen in the maximum flexion achieved by the patients, and the kinematics demonstrated by the groups was comparable. Use of this high flexion component did appear to improve tibiofemoral conformity at high flexion in patients that could achieve high flexion. Further analysis is necessary to determine if the longevity of the polyethylene is indeed improved with a high flexion component.
This study was supported by a research grant from Zimmer Inc. The guidance of Dr. Harry Rubash and Dr. Andrew Freiberg and the technical assistance of Elizabeth Desouza and Angela Moynihan were greatly appreciated. The experiments performed in the course of this study comply with the laws of the United States.