Genu recurvatum is an unusual presentation before total knee replacement. Before surgery, recurvatum may be associated with valgus deformity, rheumatoid arthritis, or extensor mechanism weakness or paralysis. 5 The presence of this deformity may raise concern about its recurrence and achieving a stable, well-functioning total knee arthroplasty after surgery.
The current study retrospectively reviewed the results of total knee replacement in patients with genu recurvatum to determine whether this deformity recurs or has an adverse effect on the outcome of a knee arthroplasty.
MATERIALS AND METHODS
Between June 1987 and March 1997, 5736 primary total knee arthroplasties were done at the authors’ institution. From this patient base, a consecutive series of 60 total knee replacements were done in 56 patients with at least a 5° hyperextension deformity before surgery. One patient with myopathy (postpolio syndrome) and two patients with posterior stabilized implants were excluded. The average age of the remaining 53 patients at surgery was 70 years. There were 29 women and 24 men. Four patients had simultaneous bilateral total knee arthroplasties. The diagnosis primarily was osteoarthritis (98%). Thus, no patients with neuromuscular disease or inflammatory arthropathy were included in this study. The remaining demographic data are shown in Table 1.
Patients were evaluated according to the Knee Society clinical 4 and radiographic 1 evaluation systems before surgery and at each followup. Joint line positioning was intact and thus not removed because it was considered to be functional (providing stability) at the time of surgery. Posterior cruciate ligament recession was done as measured on the radiographs taken before surgery and immediately after surgery according to the method described by Figgie et al. 2 Briefly, the lateral radiograph was used to measure the perpendicular distance between the tibial tubercle and the weightbearing surface of the tibial plateau. Standard 100-cm anteroposterior (AP) radiographs taken with the patient standing and lateral radiographs obtained with the patient in a nonweightbearing position were taken before surgery and at each followup. Patients were included in this study if their preoperative (primary total knee arthroplasty) examination revealed at least a 5° hyperextension deformity. In all patients (before and after surgery), this measurement was obtained by the attending orthopaedic surgeon with the patient in the supine position on the examination table with the assistance of a goniometer. Because lateral radiographs routinely were taken with the patient’s knee in flexion, rather than extension, the physical examination, rather than radiographs, were used to measure hyperextension. At the authors’ institution, such data are recorded immediately for computer compilation. Hyperextension always is recorded as positive and contracture as negative. For the current study, these measurements were rounded to the nearest 5° because smaller estimates were thought to be within measurable error. Patients were not included if the preoperative recurvatum measured less than 5°.
All knees were approached through a standard medial parapatellar incision. A posterior cruciate-retaining prosthesis was implanted in all patients. In 49 (86%) knees an anatomic graduated component prosthesis (AGC, Biomet, Warsaw, IN) was used. In eight (14%) knees a posterior cruciate condylar prosthesis (Howmedica, East Rutherford, NJ) was implanted. The choice of implant was based on availability. All tibial and patellar components were cemented. In all patients, the posterior cruciate ligament was intact and thus not removed because it was considered to be functional (providing stability) at the time of surgery. Posterior cruciate ligament recession was done as needed. The condition of the anterior cruciate ligament at the time of surgery also was observed. It was judged to be normal (15 knees, 26%), present but lax (eight knees, 14%), a nonfunctioning band (nine knees, 16%), or absent (25 knees, 44%). Intraoperative ligament releases also were recorded at the time of surgery. Instability in the coronal plane was measured intraoperatively after prosthetic implantation. This instability was estimated to be zero, 1 to 5 mm, or 5 to 10 mm.
Ambulation was initiated on the first postoperative day, and range of motion (ROM) was initiated on the second postoperative day. All patients were allowed immediate full weightbearing. Clinical and radiographic evaluations were done at each followup commencing 8 weeks after surgery and then at 6 months and 1, 3, 5, 7, and 10 years. The average followup was 4.5 years (range, 3–10 years).
Statistical analysis was done using paired t test, linear regression, or Poisson regression for comparing continuous sets of data and chi squared or Pearson correlation when evaluating noncontinuous data (Statistical Analysis System, Cary, NC).
Preoperative hyperextension averaged 11° (range, 5°–20°). Forty-seven (82%) patients had hyperextension deformity of at least 10° before surgery. Anteroposterior alignment averaged 4.9° valgus. Table 2 shows the preoperative Knee Society knee, function, and pain scores, which averaged 41 points, 41 points, and 13 points, respectively.
Table 3 shows a comparison between preoperative hyperextension and preoperative AP alignment. Although the average valgus alignment was slightly greater in the knees with 10° recurvatum (6.1°), the differences were not statistically significant for Pearson correlation (p = 0.4253) or linear regression (p = 0.3470).
Intraoperatively, the hyperextension deformity was corrected in all but one knee (56 knees, 98%). Fifty-four knees had full extension intraoperatively after prosthetic implantation (95%). Two (3%) knees had a 10° contracture at the end of the operation. A 20° hyperextension deformity persisted in one (2%) knee intraoperatively. At the surgeons’ discretion it was decided not to attempt to correct the deformity with repositioning of the components.
Table 4 shows the number of medial ligamentous, lateral ligamentous, posterior, and lateral retinacular releases as a function of hyperextension deformity. Linear regression (continuous data) analysis revealed no significant relationship between the degree of preoperative recurvatum and medial ligamentous release (p = 0.6013), lateral ligamentous release (p = 0.2612), and posterior releases (p = 0.2186). Poisson regression (categorical) also revealed no relation between the need for lateral patellar (retinacular) release and extension deformity (p = 0.2052). In addition, medial ligament releases (p = 0.1698), lateral ligamentous releases (p = 0.7255), posterior releases (p = 0.7192), and lateral retinacular releases (p = 0.9738) were unrelated to the severity of the preoperative valgus (AP) alignment.
Six (10%) knees had mild medial instability of 5 mm or less and four (7%) knees had medial instability between 5 and 10 mm after implantation of the prostheses. In all knees the remaining instability was thought to be acceptable and not sufficient enough to require additional ligament balancing. No lateral, posterior, or posterolateral instability was seen at the end of the operation. The tibiofemoral joint line position was raised an average of 0.6 mm as a result of the total knee arthroplasty. In 38 (66%) knees, no measurable change was seen in the joint line position between the radiographs obtained before and after surgery. In four (7%) knees, the joint line was lowered 2 to 3 mm, and in 15 (26%) knees the joint line was raised between 2 and 5 mm.
Knee Society knee, function, and pain scores improved to 81 points, 78 points, and 43 points, respectively, at final followup (Table 5). At the most recent followup, flexion averaged 113° (range, 80°–135°). Only two (3.5%) knees had a hyperextension deformity at the most recent followup. In both knees, 10° hyperextension was observed. Fifty-two (91%) knees had neutral extension (0°). Of the two patients who had a hyperextension deformity after surgery (10° each), one patient had 20° hyperextension and the other patient had 0° hyperextension intraoperatively after prosthetic implantation. Both knees had 5 to 10 mm residual medial instability at the operation. Extension of the entire study group averaged 0° at final followup.
At final followup, no knee replacement was revised for any reason. A progressive radiolucent line greater than 1 mm in tibial Zone 7 was identified in one knee. No other progressive radiolucent lines were observed.
Although the presence of a nonfunctioning anterior cruciate ligament was associated statistically with a greater degree of recurvatum before surgery (p = 0.0420), the intraoperative condition of the anterior cruciate ligament was not related to the final postoperative extension (p = 0.1563). Intraoperative extension (p = 0.0001) and final postoperative extension (p = 0.0001) were significantly less when compared with the preoperative hyperextension deformity (paired t test). However, final extension was statistically greater (as much as 10°) at final followup in knees with residual medial instability (p = 0.0067, paired t test). Although only two knees had a recurvatum deformity at final followup (both with residual medial instability), an increase in extension at final followup was observed in three of the 10 knees with residual medial instability. These three knees had residual medial instability between 5° and 10°.
Because of associated muscular weakness and paralysis, genu recurvatum has been considered a relative contraindication to total knee replacement. Although in the current study there were no patients with neuromuscular disease or inflammatory arthropathy, specific attention should be given to evaluation of the quadriceps, hamstrings, and gastrocnemius complex when patients present with this deformity because the recurvatum may recur with a surface replacement arthroplasty. 5
Several authors have described techniques to correct genu recurvatum at the time of knee replacement arthroplasty. Insall 3 suggested that operative correction may be obtained easily by underresection of bone ends and use of relatively thicker femoral or tibial components. In no knees in the current study was the distal femoral or proximal tibial underresected. In addition, as evidenced by the joint line not changing in 39 of 60 (65%) knees in patients in the current study and the average change in the joint line being less than 1 mm, the use of thicker components was not required to achieve surgical correction. Krackow 6 and Krackow and Weiss 7 suggested that posterior capsular plication or proximal and posterior transfer of the collateral ligaments may be required to achieve surgical correction of the hyperextension deformity. The use of these techniques was not required intraoperatively in any of the patients in the current study to correct hyperextension.
Only two (3.5%) knees had a residual flexion contracture (10° each) intraoperatively. At final followup, the final extension in these two knees did increase. In addition, two knees that had full extension intraoperatively lost extension at final followup, one knee had a 5° contracture and the other knee had a 10° contracture. Other authors have reported that flexion contractures that remain at surgery improve with time. 8–10 Yet in the current study, in three of 10 knees that had residual medial instability at the time of surgery, final knee extension increased statistically (as much as 10°) at final followup. Two of these three knees had a flexion contracture intraoperatively. This finding seems to stress the importance of correcting any residual medial instability at surgery to avoid any potential recurrence of the genu recurvatum.
In the current study, 30 of the 57 knees required lateral retinacular release. During this study, femoral component rotation was not measured at the time of surgery. Most likely, the high lateral release rate was attributable to femoral component positioning. Currently, the authors externally rotate the femoral component between 0° and 5°, which has diminished substantially the need for lateral retinacular release.
Fifty-three patients with genu recurvatum had total knee replacement without underresection of bone, posterior capsular plication, or ligament transfer. The posterior cruciate ligament was retained in all patients. Only two knees had residual hyperextension deformity at final followup, both of which had between 5 and 10 mm medial instability after prosthetic implantation.
In the absence of neuromuscular disease, hyperextension deformities tend not to recur after total knee replacement. Genu recurvatum can be treated successfully without a posterior stabilized total knee replacement with retention of the posterior cruciate ligament. Care should be taken to avoid even mild degrees of residual instability in the coronal plane at surgery because this is associated with increased extension, including hyperextension. The presence of genu recurvatum before surgery does not preclude a well-functioning total knee replacement. Total knee replacement is not contraindicated in patients with genu recurvatum.
1. Edwald FC: The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin Orthop 248: 9–12, 1989.
2. Figgie HE, Goldberg VM, Kingsbury GH, et al: The influence of tibial-patellofemoral location on function of the knee in patients with the posterior stabilized condylar knee prosthesis. J Bone Joint Surg 68A: 1035–1040, 1986.
3. Insall JN: Surgical Techniques and Instrumentation in Total Knee Arthroplasty. In Insall JN, Windsor RE, Scott WN, Kelly MA, Aglietti P (eds). Surgery of the Knee. New York, Churchill Livingstone 739–804, 1993.
4. Insall JN, Dorr LD, Scott RD, et al: Rationale of the Knee Society clinical rating system. Clin Orthop 248: 13–14, 1989.
5. Insall JN, Haas SB: Complications of Total Knee Arthroplasty. In Insall JN, Windsor RE, Scott WN, Kelly MA, Aglietti P (eds). Surgery of the Knee. New York, Churchill Livingstone 891–934, 1993.
6. Krackow KA: The Technique of Total Knee Arthroplasty. St Louis, CV Mosby 1990.
7. Krackow KA, Weiss A-PC: Recurvatum deformity complicating performance of total knee arthroplasty: A brief note. J Bone Joint Surg 72A: 268–271, 1990.
8. McPherson EJ, Cushner FD, Schiff CF, et al: Natural history of uncorrected flexion contractures following total knee arthroplasty. J Arthroplasty 9: 499–502, 1994.
9. Ritter MA, Stringer EA: Predictive range of motion after total knee replacement. Clin Orthop 143: 115–119, 1979.
10. Tanzer M, Miller J: The natural history of flexion contracture in total knee arthroplasty: A prospective study. Clin Orthop 248: 129–134, 1989.