The predictable pain relief and functional improvement after primary total knee replacement have been well-established with many studies reported during the past 15 to 20 years. 1,2,4–8,10–19 The long-term outcome of one design of total knee replacement has been a difficult clinical question to answer because numerous changes and improvements in implant design, materials, and fixation have led to relatively few long-term studies of the survival of a total knee replacement.
Numerous factors have been found in association with early failure of total knee replacement including abnormal postoperative varus alignment, 1,14 excessive polyethylene stress and wear, failure of metal backing, thin polyethylene, poorly manufactured polyethylene, and ligament imbalances.
There has been much debate during the past decade on the relative advantages and disadvantages of posterior cruciate retention versus substituting designs of total knee replacement. Posterior cruciate ligament retention may afford increased range of motion (ROM) and improved stair climbing ability while enhancing the biomechanics of femoral rollback during knee flexion. It is thought that posterior cruciate ligament retention may better balance the flexion gap and maintain the position of the joint line. Posterior cruciate ligament substituting designs have been hypothesized to prevent posterior subluxation of the tibia and additionally enhance rollback during deep knee flexion. The increased conformity required in posterior cruciate ligament substituting designs may decrease the stresses in the polyethylene and at the bone-cement interface. 3 There seems to be no difference in the long-term results of either design if the procedure is done well.
Since 1983, the authors have implanted a cemented total condylar design of knee replacement (Anatomic Graduated Components [AGC], Biomet, Inc, Warsaw, IN) in patients undergoing primary total knee replacement. This is a posterior cruciate-retaining design with a one-piece tibial component. The tibial component was implanted with a CoCr baseplate. The articular surface is a nonmodular direct compression molded polyethylene component with a nonconforming articulation. The femoral component is universal and does not have specific geometry for right and left. Three hundred forty-eight of the femoral components were uncemented and the remainder were cemented. All tibial and patellar components were cemented. The patella is resurfaced routinely. A metal-backed polyethylene component formerly was implanted in 296 knees, and now an all-polyethylene insert is used routinely. The purpose of the current study is to describe the outcome of 4583 AGC total knee replacements with up to 15-years followup.
MATERIALS AND METHODS
A retrospective review of the records for all patients who had primary total knee replacement with the AGC since 1983 at the authors’ institution were evaluated. The surgery was performed by four surgeons using similar techniques. Demographic data were collected as to age, gender, indication for surgery, complications, and clinical and radiographic failures.
The geometry of the femoral and tibia components has not changed since the authors started performing the operation in 1983. Both components were cemented except for 348 femoral components. Survival data were determined for patient and implant survival. Postoperative complications, revisions, functional outcome with pain scores, and knee scores were evaluated. Lateral and patellar view radiographs and anteroposterior (AP) radiographs obtained with the patient standing were taken at 1-to 2-year intervals and at the most recent clinic visit. The radiographs were reviewed for signs of aseptic loosening, osteolysis, and radiolucent lines.
The Kaplan-Meier survival analysis 9 was used in a standard form in which the patients lost to followup were excluded from the study and were considered to have a successful result at their last visit. There was a best case scenario in which it was assumed that all patients who were lost to followup were doing well throughout the end of the study and a worst case scenario in which it was assumed that all the patients lost to followup had failed results at their last clinic visit. Failure was defined as loosening, revision of any component, or both. A prosthesis was determined to be loose if it had migrated or there was a change in alignment.
Statistics were performed with Kaplan-Meier 9 survival curves and log rank test to determine factors associated with failure. A p value < 0.05 was selected to determine statistical significance.
There were 4913 AGC total knee replacements performed between September 1983 and December 1996. Exclusion criteria for this study included infection and revision surgery. Sixty-two knees were excluded because of infection for a rate of 1.3%. There were 58 revision total knee replacements. Two hundred ten knees were lost to followup (4.2%). This left 4583 total knee replacements performed on 3054 patients for this series. There were 407 deaths (8.9%) during the followup period.
The diagnosis was osteoarthritis in 87% of patients. The average age of the patients at the time of total knee replacement was 70.4 years (+/− 8.9 years; range, 18–93 years). Sixty percent of the patients were women. All patients had femoral, tibial, and patellar components inserted. There were 296 metal-backed patellar components and 4287 all-polyethylene patellar components, both of which were cemented.
Survival of Components
The survival data and the number of total knee replacements and failures based on time and component are shown in Table 1. Log rank test analysis revealed that survival of any of the components was not related to age (p = 0.39), gender (p = 0.73), or weight as determined by body mass index (p = 0.28). There were 2646, 780, and 30 knees available at 5, 10, and 15 years, respectively (Fig 1).
Femoral Component Failure
Six femoral components were loose and all were revised (0.1%). The mean time to failure of the femoral component was 4.6 years with a range of 0.5 to 7.6 years. All loose femoral components were observed by 5 years postoperatively.
Tibial Component Failure
Twenty-one tibial components were loose and 19 were revised (0.4%). The mean time to failure of the tibial component was 4.5 years with a range of 0.6 to 10.9 years. All loose tibial components were observed by 3 years postoperatively. There were three ligament instabilities, two medial metal baseplate fractures, and the remainder of the components were loose at the bone-cement interface and failed in varus.
All-Polyethylene Patella Component Failure
There were 180 (4.2 %) all-polyethylene patellar components that were loose. There were only 15 revisions. The mean time to failure was 2.6 years ±1.75 years. Lateral retinacular release was significantly associated with loosening of the patella (p < 0.001).
Metal-Backed Patellar Component Failure
Of 296 total knee replacements, there were 28 (9.5 %) loose metal-backed patellas. The mean time to loosening was 6.8 years with a range 1.4 to 12.2 years. All 28 metal-backed patellas were revised.
Knee Scores and Function
The overall knee score at most recent followup averaged 81 points with a range of 26 to 96 points. The average followup was 10.7 years.
Range of Motion
The average postoperative ROM of 110° did not change dramatically for 15 years.
Total knee replacement has become one of the most reliable and predictable procedures in orthopaedic surgery. 1,2,4–8,10–19 Total knee replacement has been shown to restore patient function and relieve the pain and deformity that results from knee arthrosis. The success of a total knee replacement has been determined by many factors including pain relief, functional outcome, ROM, radiographic loosening, or component revision. There have been numerous changes in the total condylar type of total knee arthroplasty during the past 20 years. These modifications include alterations in component geometry, the addition of sizing options and modularity, more accurate instrumentation for correcting deformity, increased understanding of the role of rotational alignment of the components, and improved cement techniques. Each of these modifications has positively influenced the long-term outcome of total knee replacement. These changes, however, make large long-term studies with one design of total knee replacement difficult to perform.
There has been much debate regarding the benefits and limitations of posterior cruciate ligament retention during total knee replacement. In theory, designs that have nearly flat tibial surfaces in the mediolateral plane and retain the posterior cruciate ligament have higher peak stresses in the polyethylene and at the bone-cement interface than those designs with more conformity. 3 Bartel et al 3 suggested that because the maximum principal stress, the contact stress, and the maximum shear stress all increase with decreasing conformity, then stresses in less conforming total knee replacement designs would be increased. 3 These stresses may exceed the polyethylene yield strength and lead to early failure of these designs. This was seen clearly in the current series of all-polyethylene tibial components of the AGC design. The mechanism of failure was fatigue in the medial plateau caused by overload of the cancellous bone leading to collapse and component subsidence or fatigue failure. Metal backing of the tibial component with this design dramatically reduces the tibial failure rates as seen in the current series in which a one-piece tibial component is implanted with polyethylene compression molded directly onto a metal endoskeleton. TABLE 2.
There is a theoretical concern that the increased constraint and conformity of the posterior-stabilized design may result in increased loosening rates with longer followup. It is interesting to examine the time to failure of the AGC. In the current series, the authors found that tibial, femoral, and all-polyethylene implants that were going to fail through aseptic loosening did so on average less than 5 years after implantation. These early failures most likely represent surgical errors of improper soft tissue balancing at the time of implantation and not a long-term failure of the prosthesis as seen with the mechanism of polyethylene wear. Despite the increased stresses in the polyethylene and in the bone-cement interface, the AGC design has proved to be durable with time with minimal polyethylene wear and low failure rates.
In 1993, Ranawat et al 12 reported the 15-year clinical survivorship of the total condylar knee arthroplasty with posterior cruciate ligament sacrifice and early cement technique to be 94.1%. In this series, with the end point defined as radiographic signs of loosening or revision of the components, the survival was 90.9%. There were no failures before 8 years and three late failures in 69 at-risk patients after 12 years. 12 Falatyn et al 7 reported similar 10-year followup results with the cemented total condylar total knee replacement and posterior cruciate ligament sacrifice to be 84.1% when the end point was defined as loose or revised components.
Font-Rodriguez et al 8 reported the results of 2629 total knee replacements during a 22-year period. With a posterior-stabilized design and a metal-backed tibial component, the 14-year survivorship was 98.1%. Of 2036 knee replacements, one tibial component and seven femoral components were revised for loosening. Colizza et al 5 reported the results of a posterior-stabilized prosthesis, with a metal-backed tibial component. At 11-years followup, there was no loosening in the tibial components and minimal polyethylene wear. Two femoral components were revised for loosening. The best case scenario produced a failure rate of 96.4% and a worst case scenario produced a survival rate of 92.6%. In a minimum 10-year followup study, Aglietti et al 1 reported survivorship results of 60 total knee replacements of the Insall-Burstein posterior-stabilized design to be 92% at 10 years. The specifics of time to failure were not reported directly. There was minimal wear in this series of compression molded one-piece tibial components. 1 The long-term outcome regarding loosening and revision in this series was excellent. 1 Increased rates of loosening were seen in knees with varus positioning reported as limb alignment of 4° valgus of less. Varus positioning has been shown to increase failure rates in another series. 15
In a series reporting the survivorship results of the PFC total knee replacement system with a minimum of 10 years followup, Schai et al 17 reported no loosening of tibial or femoral components but found wear-related problems in 8% of patients. They also changed from metal-backed patellar components to all-polyethylene components and reduced the need for patellar revisions. This was a similar finding to the need for revision of loose patellar components observed in the current series. The 10-year survival of the posterior cruciate condylar total knee replacement was reported to be 94.7%13 and for the posterior cruciate ligament-retaining total condylar total knee replacement was 96.8%15 The kinematic condylar prosthesis with posterior cruciate ligament retention has been reported to have estimated survival rates at 10 years of 96%9 and 92%. 19 Ritter et al 16 previously reported a 98% survival rate for the AGC at 10 years in a multicenter study. 16 The majority of the late failures were radiographic loosening in the patella component for which the patient did not require revision surgery.
The current study shows that the patella remains the most frequent source of failure of a total knee replacement. There was a striking contrast between the outcome of a failed all-polyethylene patellar component and a metal-backed patellar component. Of the loose or failed all-polyethylene components, only a few were revised. This compares with the metal-backed patellar component in which 100% were revised. This difference in the need for revision may be explained by a difference in pain, mechanical symptoms, and the metallosis that results from the failed metal-backed patellar component. During the later time of this study, there was a significant advancement in the understanding of the importance of rotational alignment of the femoral and tibial components to the tracking of the knee. Proper tibial and femoral component rotation leads to improved patellofemoral mechanics and tracking. This has led to a dramatic reduction in the need for lateral retinacular release and ultimately the incidence of avascular necrosis of the patella and patellar failure.
The current study revealed that the AGC posterior cruciate-retaining total knee replacement has excellent longevity with a clinical survivorship greater than 98% at 10 to 15 years. The AGC, being a nonconstrained design having nearly flat-on-flat geometry, and retaining the posterior cruciate ligament which should increase the stresses in the polyethylene and at the bone-cement interface, has proved to have minimal wear and excellent performance and longevity with time. The authors think this is a result of the direct compression molded polyethylene articulation, the one-piece configuration that incorporates metal backing on the tibial component and eliminates back-sided tibial component polyethylene wear, nonconstraint of any of the articulating surfaces, and proper ligamentous balancing of the collateral and posterior cruciate ligaments.
1. Aglietti P, Buzzi R, De Felice R, et al: The Insall-Burstein total knee replacement in osteoarthritis. J Arthroplasty 14: 560–565, 1999.
2. Bankston AB, Keating EM, Ranawat C, et al: Comparison of polyethylene wear in machined versus molded polyethylene. Clin Orthop 317: 37–43, 1995.
3. Bartel DL, Bicknell VL, Ithaca MS, et al: The effect of conformity, thickness, and material on stresses in ultra-high molecular weight components for total joint replacement. J Bone Joint Surg 68A: 1041–1051, 1986.
4. Callahan CM, Drake BG, Heck DA, et al: Patient outcomes following tricompartmental total knee replacement. JAMA 271: 1349–1357, 1994.
5. Colizza WA, Insall JN, Scuderi GR: The posterior stabilized total knee prosthesis. J Bone Joint Surg 77A: 1713–1720, 1995.
6. Ewald FC, Wright RJ, Poss R, et al: Kinematic total knee arthroplasty. J Arthroplasty 14 (4): 473–480, 1999.
7. Falatyn S, Lachiewicz PF, Wilson FC: Survivorship analysis of cemented total condylar knee arthroplasty. Clin Orthop 317: 178–184, 1995.
8. Font-Rodriguez DE, Scuderi GR, Insall JN: Survivorship of cemented total knee arthroplasty. Clin Orthop 345: 79–86, 1997.
9. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457–481, 1953.
10. Malkani AL, Rand JA, Bryan RS, et al: Total knee arthroplasty with the kinematic condylar prosthesis. J Bone Joint Surg 77A: 423–431, 1995.
11. Ranawat CS, Flynn Jr WF, Deshmukh RG: Impact of modern technique on long-term results of total condylar knee arthroplasty. Clin Orthop 309: 131–135, 1994.
12. Ranawat CS, Flynn Jr WF, Saddler S, et al: Long-term results of the total condylar knee arthroplasty. Clin Orthop 286: 94–102, 1993.
13. Ritter MA, Campbell E, Faris PM, et al: Long-term survival analysis of the posterior cruciate condylar total knee arthroplasty. J Arthroplasty 14: 293–296, 1989.
14. Ritter MA, Faris PM, Keating EM, et al: Postoperative alignment of total knee replacement: Its effect on survival. Clin Orthop 299: 153–156, 1994.
15. Ritter MA, Herbst SA, Keating EM, et al: Long-term survival analysis of a posterior cruciate-retaining total condylar total knee arthroplasty. Clin Orthop 309: 136–145, 1994.
16. Ritter MA, Worland R, Saliski J, et al: Flat-on-flat, nonconstrained, compression molded polyethylene total knee replacement. Clin Orthop 321: 79–85, 1995.
17. Schai PA, Thornhill TS, Scott RD: Total knee arthroplasty with the PFC system. J Bone Joint Surg 80B: 850–858, 1998.
18. Scuderi GR, Insall JN, Windsor RE, et al: Survivorship of cemented knee replacements. J Bone Joint Surg 71B: 798–803, 1989.
19. Weir DJ, Moran CG, Pinder IM: Kinematic condylar total knee arthroplasty. J Bone Joint Surg 78B: 907–911, 1996.
Merrill A. Ritter, MD; and John B. Meding, MD, Guest Editors© 2001 Lippincott Williams & Wilkins, Inc.