SECTION II: ORIGINAL ARTICLES
Numerous authors3,10,18 have advised relative undercorrection of alignment of the knee in medial unicompartmental arthroplasty to avoid progressive osteoarthritis (OA) of the opposite compartment. The results of knee alignment on the long-term outcome of unicompartmental arthroplasty has received attention in some reports.9,18 To our knowledge, however, no study of measurement of leg alignment with the hip-knee-ankle angle after unicompartmental arthroplasty has been reported and therefore there is little evidence that overcorrection causes progression of OA in the opposite compartment. Furthermore, there is a possibility that correction of the deformity achieved during surgery does not remain constant in the long term and that varus deformity tends to recur as observed after high tibial osteotomy.6
Wear of the polyethylene (PE) tibial implant is the dominant issue in unicompartmental arthroplasty and has been linked to failure of the PE through resultant loosening.7,11,12,17 Wear of the PE is dependent on the sliding distance of the femoral condyle as it articulates across the tibial implant and on the load carried by the interacting surfaces. The sliding distance depends on the type and duration of the activities done by the patient but the load applied to the joint in vivo depends on the hip-knee-ankle angle. Therefore there is a risk that undercorrection of the deformity increases wear of the PE.
The aim of this study was to identify, from a series of patients with long-term followup, the influence of postoperative alignment on wear of the PE, on degenerative changes in the opposite compartment, and on the risk of recurrence of the deformity.
MATERIALS AND METHODS
There were 156 medial prostheses implanted between 1978 and 1988. The implant used was the Lotus Mark 1 unicompartmental prosthesis developed by the GUEPAR Group (Howmedica, Benoist Girard, Herouville Saint Clair, France),5 with a flat PE tibial component and a resurfacing femoral component. All of these implants were sterilized by the same method (ethylene oxide gas). To remove the possible confounding effect of the absence of the anterior cruciate ligament (ACL) on wear and loosening, only knees with an ACL were included in this study. Routine AP and lateral, and skyline radiographs with the patient weightbearing were taken preoperatively. Patients with joint space narrowing on the opposite femorotibial compartment or on the preoperative patellar skyline view were excluded from receiving unicompartmental arthroplasty. All patients had complete joint space narrowing on the medial femorotibial compartment. Among these 156 prostheses, 82 were implanted in 72 patients who died without revision surgery before 10 years followup, and eight patients were lost to followup before 10 years. Sixteen knees in 16 patients had been revised. The most recent followup was done a minimum of 10 years after the index arthroplasty. Fifty-eight knees (46 patients) had no revision at the most recent followup (average, 15 years; range, 10–20 years). The thickness of the PE component was 6 mm in two knees, 9 mm in 32 knees, 12 mm in 23 knees, and 15 mm in one knee. These 46 patients returned for clinical and radiographic followups. The knee scores determined preoperatively with the system of the Hospital for Special Surgery13 were 57 ± 13 points. The mean age of the patients at surgery was 70 years (range, 43–83 years). The rating systems of the Hospital for Special Surgery and the Knee Society8 were used to determine postoperative scores. The latest radiographic evaluation, done for the 58 knees in the patients who still were alive at the most recent followup, was done to obtain information on the relationship of alignment to the wear rate of the PE, to the progression of arthritis in the unresurfaced compartment, and to the risk of recurrence of the deformity.
Preoperative alignment (hip-knee-ankle angle) was measured (Fig 1) on radiographs6 of the entire limb taken with the patient weightbearing; all patients were asked to stand with the anterior part of the knees facing forward and the posterior part of the knees facing the film. A cassette holding long radiographs was placed behind the patient. Both lower extremities were included in one radiograph. The AP radiograph of the entire lower limb taken with the patient weightbearing was obtained with the xray beam centered at the knees at a distance of 3 months and the foot oriented in the direction of the xray beam. The hip-knee-ankle angle was formed by the angle between the line joining the center of the femoral head to the center of the knee and the line joining the center of the knee to the center of the ankle. These axes usually form a straight line (180°). Therefore, with a varus deformity the angle is less than 180° and when there is a valgus deformity, it is greater than 180°. The average preoperative hip-knee-ankle angle was 173° (range, 152°-179°). The amount of preoperative varus deformity was not a contraindication at this time. The amount of preoperative deformity considered to be relieved by surgery was the varus deformity attributable to loss of cartilage and bone on the femur and on the tibia.
Long films of the entire limb (hip-knee-ankle angle) also were obtained during the first postoperative month. At 1 month, the postoperative average hip-knee-ankle angle was 177° (range, 157°-190°). In 10 of these 58 medial implants, the preoperative deformity was overcorrected to valgus (Group A) with an average postoperative hip-knee-ankle angle of 183° (range, 181°-186°). The other 50 knees had an average postoperative hip-knee-ankle angle of 173° (range, 157°-179°). Forty implants had slight varus (Group B) with a postoperative hip-knee-ankle angle between 180° and 170° and eight had severe undercorrection with varus (Group C) greater than 10° postoperatively (hip-knee-ankle angle < 170°).
The hip-knee-ankle angle was evaluated at the most recent followup by a radiograph of the entire limb, taken with the patient weightbearing, which was obtained using the same conditions as the postoperative radiographs of the entire limb. Limb rotation was matched among preoperative and postoperative radiographs and radiographs of the entire limb from the latest followup by using the position of the foot. The recurrence of the deformity was calculated as the difference between the postoperative hip-knee-ankle angle and the same angle at the most recent followup. Joint space was measured in millimeters in the opposite compartment on standard AP radiographs taken with the patient weightbearing. These radiographs were obtained before the operation, early in the postoperative period, and usually at each year interval thereafter until the last followup or the most recent followup. Fluoroscopy was used to position the knee for the radiograph taken with the patient standing to obtain a radiograph beam parallel to the tibial implant (Fig 2). Polyethylene wear was measured on films with reference to the metal marker in the tibial component (Fig 3), using a magnifier that enlarges the distance by 20-fold. Penetration of the femoral component in the PE was measured on radiographs. We are aware that this penetration is not the true wear and that this penetration represents creep and wear of the PE. For implants that were revised, wear was measured directly on specimens with a high precision dial gauge.
The outcome measures were decrease of joint space in the opposite compartment, wear of the PE, and recurrence of the deformity. Multiple linear regression analysis was used to determine differences in alignment of the knee among outcome measures, while adjusting for any confounding effect of age, weight, gender, thickness, and position of the implants. The Spearman test was used to determine possible correlations between the recurrence of the deformity, the rate of penetration of the femoral component into the PE, the amount of subsidence of the tibial implant, and the postoperative mechanical axis. The Mann-Whitney U test was used to identify the significance of differences between groups. The chi square test was used to identify trends within groups with categoric variables. Significance was defined as p < 0.05. The hip-knee-ankle angle was measured twice by two observers with the same goniometer on each radiograph of the entire limb. A two-way random effect (variance components) ANOVA was used to determine the reliability of measurements. The impact of this variation attributable to observer error (based on the inversion of the tolerance limit calculations) was that one could be 95% confident that the observer measurement would be less than 2.7°. The precision of our measurements for wear of the PE was assessed by comparison with the measurements taken on retrieved components with a high precision dial gauge (Mitutoyo/MTI, Paris nord II, Roissy, France) graduated to 0.003 mm. The relative error of the radiographic technique was determined by calculating the absolute difference between the true (dial gauge measured) wear and the radiologic estimation of wear and then by dividing this difference by the true wear. For the implants retrieved, the error for estimation of linear penetration was 24.5% (minimum, 12.6%; maximum, 63.7%). The greatest difference occurred in the smallest depths of wear, where the radiographic estimation tended to overestimate the true wear.
Assessment at the most recent followup was done on 58 knees in 46 patients. Among the 58 knees, 10 were in Group A, 40 were in Group B, and eight were in Group C (Table 1). Using multiple linear regression analysis, alignment of the knee had a significant influence on the knee scores (p = 0.01), on the rate of wear of the cartilage in the opposite compartment (p = 0.01), on the rate of wear of the PE (p = 0.03), and on the recurrence of the deformity (p = 0.02).
Alignment of the knee influenced the knee score at the most recent followup. Of these 58 knees that were not revised and that could be examined at least 10 years after surgery (average, 15 years; maximum, 20 years), according to the scale of the Knee Society,8 the mean clinical score was 184 ± 10 points (95 ± 9 points for the knee score and 89 ± 9 points for the functional score) (Table 1). The best scores were obtained in the 32 knees of Group B (chi square test with Yates correction; p = 0.01 for a score greater than 184 points).
Valgus overcorrection increased the risk of lateral degeneration. The radiographic progression of OA in the opposite femorotibial compartment of the knee (Table 1) occurred more frequently and more severely among patients who had valgus overcorrection of the deformity (six of the 10 knees of Group A, with an average rate of wear of 0.23 mm per year for these six knees) than among patients of Group B and Group C (eight of 48 knees, with an average rate of wear of 0.12 mm for these eight knees); the difference was significant (respectively p = 0.04 with the chi square test with Yates correction for the number of knees, and p = 0.03 with the Mann-Whitney U test for wear). For these 14 knees with joint space narrowing in the opposite compartment, the mean Knee Society functional score at the most recent followup was significantly lower than the mean score for the knees with a normal joint space on the contralateral compartment (68 ± 8 points compared with 86 ± 8 points (p < 0.05) with the Mann-Whitney U test).
Varus undercorrection increased wear of PE and recurrence of the deformity. Between the 1-month postoperative examination and the latest followup (10–20 years), changes in alignment and wear of the PE occurred in the 48 undercorrected knees (Group B and Group C) in patients who were followed up for that length of time (Table 1). A recurrence of varus of an average of 5° (minimum, 2°; maximum, 8°) was observed between the hip-knee-ankle angle measured 1 month postoperatively and at the most recent followup. There was a significant correlation (Spearman test; Rs = 0.380; p = 0.01) between the postoperative undercorrection in varus at 1 month and the recurrent varus recorded at the most recent followup. There was an association between the rate of penetration of the femoral component in the PE of the tibial component and the amount of varus deformity still present on the postoperative hip-knee-ankle angle (Spearman test; Rs = 0.264; p < 0.01). Wear rate of PE and recurrence of the deformity were significantly (respectively p = 0.02, p = 0.03) higher in Group C than in Group B.
Our study showed that alignment influences progression of OA in the opposite compartment and wear in the tibial implant. Previous studies which attempted to assess the influence of alignment9,10,18 were limited because of their short followup and the absence of accurate measurement of alignment of the knee. Our study has several weaknesses inherent to a retrospective analysis. Wear was not measured directly on implants but only evaluated with radiographs. The penetration of the femoral condyle into the PE was the result of creep and wear. However, we thought that this measurement was appropriate because it allowed evaluation of the remaining thickness of the implant. We used only one parameter (joint space narrowing) for the progression of OA in the opposite compartment, but this joint space narrowing reflects the loss of articular cartilage in the joint and is considered to be a more reliable marker of OA than osteophytes or subchondral sclerosis. The measurement of the hip-knee-ankle angle only was an evaluation of the coronal plane alignment and other factors can influence wear in both femorotibial compartments (weight of the patient, age, position of the implants). For this reason, we used multiple linear regression analysis to evaluate the influence of alignment in the coronal plane.
Despite these limitations, our study provides valuable information. Progression of OA in the contralateral compartment of the knee was an important cause of failure in some reports of unicompartmental arthroplasty.4,15,16 Our series confirms that OA in the contralateral compartment can occur after surgery where there has been overcorrection of the preexisting varus deformity.1,2 Because overcorrection of a preexisting varus deformity is a significant cause of OA in the opposite compartment, careful selection of patients and deliberate attempts to avoid overcorrection are necessary. Low progressive joint space narrowing in the opposite compartment was observed in the long-term followup in some undercorrected knees. The rate of reduction of the joint space (< 0.15 mm per year) in these knees is similar to the rate of reduction of joint space observed in normal knees during a long-term followup14 and perhaps only is attributable to changes in cartilage in relation to the patient’s progression in age.
Severe undercorrection of the deformity was a cause of increased wear in the PE. Because the PE of the tibial implant was flat, the prosthesis that we used carried a risk of increased PE wear because of its incongruous design when a severe varus undercorrection of the deformity was kept at surgery. Because we did not do ligamentous release, postoperative alignment of the knee was dependent on the amount of passive correction possible at the time of surgery and on the amount of preoperative varus deformity.
A slight varus postoperative deformity seems to decrease wear in the PE. The problem inherent in this angular deformity is that it may be difficult to tell how much deformity will remain after the procedure. Postoperative alignment16,18 is dependent on the thickness of the tibial implant, on the level of resection of the tibia, on the ligamentous balance, and on the preoperative deformity. Keeping a slight varus postoperative undercorrection after medial unicompartmental arthroplasty, however, has one worrisome consequence. The correction of the deformity achieved during surgery did not remain constant and varus tended to recur because of wear of the PE. Therefore adverse effects of this slight undercorrection could be observed in the very long term, and continued observation will be necessary. An increased penetration of the femoral component into the PE was identified in our study as associated with undercorrection.
1. Barrett WP, Scott RD: Revision of failed unicondylar unicompartmental knee arthroplasty. J Bone Joint Surg 69A:1328–1335, 1987.
2. Choon HL, Rand JA: Revision of failed unicompartmental knee arthroplasty. Clin Orthop 287:193–201, 1993.
3. Goodfellow JW, Kershaw CJ, Benson MK, O’Connor JJ: The Oxford knee for unicompartmental osteoarthritis: The first 103 cases. J Bone Joint Surg 70B:692–701, 1988.
4. Grelsamer RP: Unicompartmental osteoarthrosis of the knee. J Bone Joint Surg 77A:278–292, 1995.
5. Hernigou P, Goutallier D: GUEPAR unicompartmental Lotus prosthesis for single-compartment femorotibial arthrosis: A five- to-nine-year follow-up study. Clin Orthop 230:186–195, 1988.
6. Hernigou P, Medevielle D, Debeyre J, Goutallier D: Proximal tibial osteotomy for osteoarthritis with varus deformity: A ten to thirteen-year follow-up study. J Bone Joint Surg 69A:332–354, 1987.
7. Insall J, Aglietti P: A five to seven-year follow-up of unicondylar arthroplasty. J Bone Joint Surg 62A:1329–1337, 1980.
8. Insall J, Dorr LD, Scott WN: Rationale of the Knee Society clinical rating system. Clin Orthop 248:13–14, 1989.
9. Kennedy WR, White RP: Unicompartmental arthroplasty of the knee: Postoperative alignment and its influence on overall results. Clin Orthop 221:278–285, 1987.
10. Klemme WR, Galvin EG, Petersen SA: Unicompartimental knee arthroplasty: Sequential radiographic and scintigraphic imaging with an average five-year follow-up. Clin Orthop 301:233–238, 1994.
11. Larsson SE, Larsson S, Lundkvist S: Unicompartmental knee arthroplasty: A prospective consecutive series followed for six to 11 years. Clin Orthop 232:174–181, 1988.
12. Lindstrand A, Stenström A, Lewold S: Multicenter study of unicompartmental knee revision: PCA, Marmor, and St Georg compared in 3,777 cases of arthrosis. Acta Orthop Scand 63:256–259, 1992.
13. Ranawat CS, Insall JN, Shine J: Duo-Condylar knee arthroplasty: Hospital for Special Surgery design. Clin Orthop 120:76–80, 1976.
14. Sahlström A, Johnell O, Redlund-Johnell I: The natural course of arthrosis of the knee. Clin Orthop 340:152–157, 1997.
15. Scott RD, Cobb AG, McQueary FG, Thornhill TS: Unicompartmental knee arthroplasty: Eight- to 12-year follow-up evaluation with survivorship analysis. Clin Orthop 271:96–100, 1991.
16. Stockelman RE, Pohl KP: The long-term efficacy of unicompartmental arthroplasty of the knee. Clin Orthop 271:88–95, 1991.
17. Swank M, Stulberg SD, Jiganti J, Machairas S: The natural history of unicompartmental arthroplasty: An eight-year follow-up study with survivorship analysis. Clin Orthop 286:130–142, 1993.
© 2004 Lippincott Williams & Wilkins, Inc.
18. Weale AE, Murray DW, Baines J, Newman JH: Radiological changes five years after unicompartmental knee replacement. J Bone Joint Surg 82B:996–1000, 2000.