Range of motion following total knee arthroplasty (TKA) is a crucial measure of clinical outcome. The purpose of this randomized, controlled study was to determine which factors are predictive of postoperative range of flexion. Fifty-six patients received either a standard or a high-flexion design NexGen Legacy Posterior-Stabilized TKA (Zimmer, Warsaw, Indiana). The relationship between preoperative flexion, intraoperative flexion, and range of flexion 1 year postoperatively was determined. The influence of soft tissue release and the type of femoral component was also investigated.
A significant correlation existed between preoperative flexion, intraoperative flexion, and maximum flexion 1 year postoperatively. Patients who had a preoperative range of flexion less than the mean range of flexion for the overall group gained flexion, whereas patients with a preoperative range of flexion greater than the mean range of flexion lost flexion. The degree of soft tissue release performed and the type of implant used had no influence on maximum flexion at 1 year.
The principal predictive factor of postoperative range of flexion, regardless of the degree of soft tissue release or implant design, is the preoperative and intraoperative range of flexion.
Pain relief and an improved level of function are the primary goals of total knee arthroplasty (TKA). Range of motion (ROM) is an important determinant of postoperative outcome, as it influences the ease of many functional activities. While surgeons can be confident that TKA can relieve pain,1 the relationship between surgery and an improved ROM is less established.
Many factors have been investigated to determine their influence on postoperative knee flexion. The strong correlation between preoperative and postoperative flexion is well established.2-6 The effect of factors such as age, body weight, and sex appears to be inconsistent,2-4 with poor standardization cited as a common problem. It has also been suggested that patients with good preoperative ROM will benefit little from surgery in terms of improved flexion and may find that their ROM is compromised. Parsley et al5 described a “migration towards a mean” from a broad spectrum of preoperative ROM. Similar findings were described by Lee et al,6 who demonstrated that the final postoperative ROM for a group of patients with poor preoperative flexion could be best predicted by intraoperative flexion rather than a preoperative value. Similar studies focusing on the relationship between intraoperative flexion and final ROM are lacking.
The effect of optimal soft tissue release and improved implant design are factors that may have a bearing on the ROM achieved both immediately and at 1 year postoperatively. A more extensive soft tissue release may be beneficial and help restore flexion in patients with a poor preoperative ROM. High-flexion TKA designs are available and are being targeted toward patients who have a good preoperative ROM.
The goals of this study were to determine the relationship between preoperative flexion, intraoperative flexion, and range of flexion 1 year postoperatively and to investigate the effect of soft tissue release and femoral component design on intraoperative flexion and range of flexion 1 year following TKA. Our hypothesis was that the final ROM was dependent on preoperative and intraoperative flexion, regardless of the degree of soft tissue release or implant design.
Materials and Methods
Patients were recruited from the waiting lists of 3 consultants (R.N.). The requirement for recruitment was the presence of unilateral osteoarthritis of the knee. Patients were excluded if they had inflammatory arthritis, osteoarthritis of the hip causing pain or restricted mobility, a foot or ankle disorder that limited walking, dementia, or a neurological disorder that affected mobility. Computer-generated randomization was used to produce cards instructing standard or high-flexion, which were placed in sealed envelopes and opened in sequence. An envelope was opened immediately before proceeding with surgery.
Clinical data were recorded preoperatively, intraoperatively, and 1 year postoperatively. An independent clinical research fellow (M.V.) who had no direct involvement with in-patient care and was blinded to the type of implant used and the degree of soft tissue release measured the active range of flexion and extension preoperatively and the final range of flexion 1 year postoperatively. Active peak flexion and extension of the knee was measured with the patient sitting on a plinth using a handheld goniometer with the arms aligned along the long axes of the femur and tibia. The ROM was calculated by subtracting the range of extension from the range of flexion.
Intraoperative flexion was measured with the patient supine and anesthetized, before application of the tourniquet. Measurements were taken pre- and immediately postoperatively using the drop test technique (maximum knee flexion under gravity without a thigh tourniquet) using a handheld goniometer (Physio Med Services, Glossop, United Kingdom). The range of intraoperative flexion was measured preoperatively and postoperatively. The change in intraoperative flexion was calculated by subtracting the preoperative drop test from the postoperative drop test.
Each patient underwent TKA through a standard medial parapatellar approach, with eversion of the patella. Soft tissue balancing was performed using a sequence of tissue release as described by Ritter et al2 for medial, lateral, and posterior structures. The extent of the soft tissue release was recorded, and a simple scoring system2 was used to compare the degree of soft tissue release in all patients. The degree of medial release ranged from 1 for deep ligament only to 4 for a release of the deep and superficial medial collateral ligaments, pes anserinus tendon, and semimembranosus tendon. The degree of lateral release ranged from 1 for the release of the iliotibial band only to 4 for the release of the iliotibial band, lateral capsule, and arcuate ligament; popliteus tendon; and lateral collateral ligament. The degree of posterior release was graded 1 for osteophytes only, 2 for the release of the posterior capsule only, and 3 for posterior capsule and osteophytes.
A NexGen Legacy Posterior-Stabilized (LPS) fixed-bearing TKA system (Zimmer, Warsaw, Indiana) was used with no resurfacing of the patella in all patients. Tibial preparation was identical in both designs. Femoral preparation using the high-flexion design involved more bone being removed from the posterior aspect of both femoral condyles to accommodate the extra implant thickness in this area (Figure 1). The high-flexion design has a smaller femoral radius of curvature and thicker posterior condyle. In theory, the smaller femoral radii of curvature increases the contact area between the posterior femoral condyle and the tibial insert. In addition, the NexGen LPS Flex has a modified cam/post mechanism and an anterior cutout slope in the polyethylene insert to allow increased jump distance while avoiding dislocation at deep flexion angles.
Figure 1: Design features of the high-flexion knee, detailing the extra bone resection from the posterior condyles.
The postoperative protocol for both patient groups was identical, with a standard care plan of early weight bearing and ROM exercises. Discharge from the hospital was dictated by each patient’s mobility and home circumstances, rather than by range of knee movement. Outpatient physiotherapy was arranged depending on individual patient requirements but was not used routinely.
Pearson’s correlation coefficient was calculated between the different measures of knee flexion. Multivariate analysis of variance was used to analyze the differences between the various grading of soft tissue release. Post-hoc Tukey tests were applied to investigate significant differences between the groups. Analysis of variance was also used to compare the standard and high-flexion designs and was repeated for a subgroup of patients who had a preoperative active flexion of >120°. A P value <.05 was regarded as statistically significant. All statistical calculations were performed using SPSS version 15 (SPSS Inc, Chicago, Illinois).
One hundred twenty-three patients with osteoarthritis were invited to take part in the study, of whom 71 (58%) agreed to participate. Fifteen patients were removed from the study, because 11 did not receive a unilateral TKA at the participating hospital. Four were unable to attend the follow-up assessment; 1 because of psychiatric illness, 1 because of an infected knee joint, 1 who was waiting for a hip replacement, and 1 who died before the final assessment could be performed. The final study group comprised 56 patients: 28 (12 men and 16 women) who underwent a standard-design TKA and 28 (17 men and 11 women) who underwent a high-flexion TKA.
The mean range of preoperative flexion for the overall group was 107° (range, 75°-135°) and for 1-year postoperative flexion was 108° (range, 65°-132°). The preoperative drop test measured intraoperatively was 115° (range, 80°-145°) and the postoperative drop test was 124° (range, 90°-150°). The mean range of preoperative extension for the overall group was 5° (range, 0°-15°) and for 1-year postoperative extension was 2° (range, 0°-5°). These numbers represent the degrees short of full extension (0°).
Patients who had a preoperative range of flexion less than the mean range of flexion (107°) of the overall group gained flexion, whereas patients with a preoperative range of flexion greater than the mean range of flexion lost flexion. This difference was statistically significant (P<.001).
Correlation analyses of preoperative flexion, pre- and postoperative drop tests, and 1-year postoperative range of flexion for the overall group revealed a significant relationship (r=0.598; P<.001) between preoperative and 1-year postoperative flexion (Figure 2). There was also a significant correlation between the preoperative drop test (Figure 3) and 1-year postoperative flexion (r=0.52; P<.001) and the postoperative drop test (Figure 4) and 1-year postoperative flexion (r=0.399; P<.001). The correlation between preoperative flexion and preoperative drop test was r=0.54 and P<.001, and between preoperative flexion and postoperative drop test was r=0.438 and P<.001. However, there was no correlation between the change in drop test and the change in flexion 1 year postoperatively (r=-0.054; P=.694).
Figure 2: Preoperative flexion vs range of flexion 1 year postoperatively. For clarity, the axes are labelled at 20° increments.
Figure 3: Preoperative drop test and range of flexion 1 year postoperatively. For clarity, the axes are labelled at 20° increments.
Figure 4: Postoperative drop test and range of flexion 1 year postoperatively. For clarity, the axes are labeled at 20° increments.
Intraoperative data on the type and extent of medial, lateral, and posterior soft tissue releases were recorded. Forty-nine patients underwent a medial release and 40 patients underwent a posterior release. The most common medial release performed was release of deep and superficial medial collateral ligaments (46%), and the most common posterior release performed was removal of osteophytes only (64%). Due to low patient numbers (7 patients), the effect of lateral soft tissue release was not investigated.
Table 1 shows preoperative flexion and postoperative change of flexion for the first 3 grades of medial release. Only 2 patients underwent grade 4 medial release. There was a tendency toward a lower preoperative drop test and a greater improvement in the drop test for those with a more extensive medial release, but this trend was not statistically significant.
The different grades of posterior release are shown in Table 2. Patients who underwent a posterior release where the osteophytes and posterior capsule were removed tended to have greater improvement in active ROM 1 year postoperatively compared to the other groups, but this difference was not statistically significant. The Spearman’s correlation coefficient between the different measures of knee flexion and extension and a total combined score for medial and posterior release was calculated. There was a significant correlation between the combined soft tissue release score with the change in active extension 1 year postoperatively (r=0.338; P=.011).
A comparison between patients who received a standard TKA and those who received the high-flexion design is shown in Table 3. There was no significant difference between the 2 designs in the change in intraoperative flexion (postoperative drop test minus preoperative drop test) and maximum flexion 1 year postoperatively.
A subgroup of patients who had a preoperative flexion of >120° were analyzed (Table 4). Patients in both groups were not significantly different (P<.900). Although there was an increase in the drop test, patients in both groups had less active flexion 1 year postoperatively. There was no significant difference between the 2 designs for this subgroup of patients with a high preoperative range of flexion. Despite both groups of patients achieving similar gains intraoperatively, patients who had a preoperative flexion <120° gained flexion, whereas patients who had a preoperative flexion >120° lost flexion.
The principal findings of this study were that preoperative and intraoperative flexion are the best predictors of final ROM, regardless of the degree of soft tissue release and implant design.
The ability to predict whether patients will regain or improve on their existing preoperative ROM has been investigated by numerous authors.2-7 By far the largest of these studies was conducted by Ritter et al.2 They retrospectively reviewed 3066 patients who underwent primary TKA at 1 institution over 15 years. They found that the principal predictive factor of postoperative ROM was preoperative ROM, and that achieving a good intraoperative ROM improved the likelihood that a good postoperative ROM would be maintained.
We found no relationship between the change in intraoperative flexion (postoperative drop test minus preoperative drop test) and the change in final flexion 1 year postoperatively. This can be explained by the fact that during anesthesia, the preoperative drop test was greater than the preoperative active flexion, and the resulting increase in postoperative drop test was significantly greater. However, the gain achieved in the drop test postoperatively was not sustained at 1 year due to the complex interaction of dynamic muscle forces and soft tissue constraints, as well as patient motivation, rehabilitation, and psychosocial status, all of which have no influence when the patient is anesthetized during measurement of the drop test. Under passive nonweight-bearing conditions, the knee seeks the course of least resistance and may not reflect normal weight-bearing articulated motion.8 The relationship between intraoperative flexion and final flexion provides the surgeon with greater insight into the final ROM that can be expected 1 year following TKA. This would aid in the counseling of patients on the expected results postoperatively and tailoring postoperative physiotherapy regimes.
The degree of medial or posterior soft tissue release performed did not affect flexion 1 year postoperatively. There was a tendency in patients with a lower preoperative drop test to achieve a greater improvement in the drop test with a more extensive medial release. In addition, patients who underwent a more extensive posterior release tended to have a greater improvement in the range of flexion 1 year postoperatively. Interestingly, a correlation analysis between the degree of combined medial and posterior soft tissue release and the change in extension revealed a moderate correlation. This is almost certainly due to the correction of flexion contractures.
These findings are similar to previous studies that have investigated the effects of soft tissue release and its influence on postoperative flexion.2,7 Ritter et al2 showed that the removal of posterior osteophytes was related to the greatest increase in postoperative flexion in the group of patients with a varus tibiofemoral alignment preoperatively. Harvey et al7 found that the release of the collateral ligament was commonly required in knees with a significant flexion deformity, but such extensive releases had no influence on the final range of flexion.
The design of the femoral component had no effect on the intraoperative flexion or the final range of flexion. The change in intraoperative flexion results suggests that the theoretical advantage of increasing the thickness of the posterior condyles to reduce impingement in flexion does not seem to be born out in practice, with no improvement in the final range of flexion 1 year postoperatively. This is because few of the patients had preoperative flexion values >120°, and therefore impingement between the back of the femoral component and the poly bearing was not an issue.
When separated into 2 groups according to whether the preoperative ROM was >120° or <120°, the gain in passive flexion intraoperatively was similar for both implants. However, this gain was not sustained in patients who had a preoperative flexion >120° who lost flexion 1 year postoperatively, regardless of the type of implant used. A possible explanation for this is that there may have been a failure to take advantage of the existing ROM by emphasizing it postoperatively, as physiotherapy was not used in all patients after discharge from the hospital.
Huang et al9 suggested that some patients with poor preoperative flexion may benefit from a high-flexion TKA. However, their nonrandomized study only included patients according to their desire for high-flexion activities postoperatively. Those with significant comorbidities were excluded from the study group. This selection bias toward a more active and motivated population may have contributed to the higher flexion achieved in that group. Other studies have recommended that high-flexion TKA should be reserved for patients with a high preoperative flexion3,10,11; however, this is at the expense of looseness, instability, and excessive polyethylene wear.12
Our results indicate that patients with a preoperative flexion <120° who underwent a high-flexion TKA gain flexion; however, this gain is not significantly greater than patients who underwent the standard-design TKA. Patients with a preoperative flexion >120° lose flexion 1 year postoperatively regardless of the type of implant used. This suggests that the type of implant had no effect on final range of flexion 1 year postoperatively, whether it was used in patients with a low or high preoperative range of flexion.
One drawback of our study was the relatively low number of patients who agreed to participate in the limited time frame of the study. This prevented a more extensive analysis of soft tissue release on range of flexion for the different implant designs. There was no significant difference in the demographic data between those who participated and those who chose not to. Further research is currently underway that will incorporate data from multiple institutions to validate the effect of individual soft tissue release and implant design on ROM following TKA.
- Heck DA, Robinson RL, Partridge CM, Lubitz RM, Freund DA. Patient outcomes after knee replacement. Clin Orthop Relat Res. 1998; (356):93-110.
- 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(7):1278-1285.
- Anouchi YS, McShane M, Kelly F Jr, Elting J, Stiehl J. Range of motion in total knee replacement. Clin Orthop Relat Res. 1996; (331):87-92.
- Ritter MA, Stringer EA. Predictive range of motion after total knee replacement. Clin Orthop Relat Res. 1979; (143):115-119.
- Parsley BS, Engh GA, Dwyer KA. Preoperative flexion. Does it influence postoperative flexion after posterior-cruciate-retaining total knee arthroplasty? Clin Orthop Relat Res. 1992; (275):204-210.
- Lee DC, Kim DH, Scott RD, Suthers K. Intraoperative flexion against gravity as an indication of ultimate range of motion in individual cases after total knee arthroplasty. J Arthroplasty. 1998; 13(5):500-503.
- Harvey IA, Barry K, Kirby SP, Johnson R, Elloy MA. Factors affecting the range of movement of total knee arthroplasty. J Bone Joint Surg Br. 1993; 75(6):950-955.
- Hsieh HH, Walker PS. Stabilizing mechanisms of the loaded and unloaded knee joint. J Bone Joint Surg Am. 1976; 58(1):87-93.
- Huang HT, Su JY, Wang GJ. The early result of high-flex total knee arthroplasty: a minimum of 2 years of follow-up. J Arthroplasty. 2005; 20(5):674-679.
- Akagi M, Nakamura T, Matsusue Y, Ueo T, Nishijyo K, Ohnishi E. The Bisurface total knee replacement: a unique design for flexion. Four-to-nine-year follow-up study. J Bone Joint Surg Am. 2000; 82(11):1626-1633.
- Allen DG, Beers C, Trammell R. Postoperative evaluation of the Nexgen Legacy posterior stabilized LPS flex implants. Paper presented at: La Societe Internationale De Chirugie Orthopedie et de Traumatologie/La Societe Internationale de Recherche Orthopedie et de Traumatologie XXII World Congress; August 28-31, 2002; San Diego, California.
- Ranawat CS. Design may be counterproductive for optimizing flexion after TKR. Clin Orthop Relat Res. 2003; (416):174-176.
Mr Ahmed and Drs Gray and Nutton are from the Department of Trauma and Orthopedics, the Royal Infirmary of Edinburgh, and Dr van der Linden is from the School of Health Sciences, Queen Margaret University, Edinburgh, Scotland.
Mr Ahmed and Drs Gray, van der Linden, and Nutton have no relevant financial relationships to disclose.
Correspondence should be addressed to: Issaq Ahmed, BEng, MBCHB, MRCS, Department of Trauma and Orthopedics, The Royal Infirmary of Edinburgh, Little France, Edinburgh, lothian, EH16 4 United Kingdom (firstname.lastname@example.org).