Long-term results of total knee arthroplasty with single-radius versus multi-radius posterior-stabilized prostheses

Total knee arthroplasty (TKA) excellent operation rate has reached 90% [1], and the revision rate has been less than 4% [2]. However, the satisfaction rate of patients with TKA has only been 75–80%, and 10% of patients still experience anterior knee pain (AKP) or patellofemoral complications [3].

The first theory of the knee rotation center was proposed by Frankel et al. [4] in 1971. They described knee flexion occurring around a varying transverse axis, with the instantaneous rotation center of the femoral posterior condyle forming a “J curve” [4, 5]. The multi-radius (MR) femoral prosthesis was theoretically designed, and the trans-epicondylar axis (TEA) which was perpendicular to the mechanical alignment (MA) of the lower extremity was widely used as a skeletal marker for the axis of flexion and rotation [6]. The multi-radius femoral prosthesis was designed based on this theory and was widely utilized since the 1980s. For example, in the sigma PFC system, the femoral implant had three segment progressive radius at the sagittal plane, and in the recent MR prostheses, the Attune system has been a continuously progressive radius like a spiral cord which may provide a more fluent flexion movement [7].

However, in 1993, Hollister et al. [8] demonstrated that the actual flexion-extension axis (FEA) of the knee sometimes did not coincide with the epicondylar axis and had a more distal and posterior axis. The locus of the rotation center between 10 and 120° can be regarded as a single spherical radius [8, 9]. The single-radius (SR) femoral prosthesis was designed according to this theory. For example, the first SR prosthesis, the Scorpio single-radius TKA (Stryker, Mahwah, NJ), was produced in 1996 [10]. The second stage was the Scorpio NRG system, and it is now the latest stage with high-flexion system in the Triathlon TKA prosthesis [11, 12].

The SR prosthesis provided a more distal FEA to have a longer extensor moment arm [13], maintaining stabilization during the middle segment range of motion (ROM), thereby reducing the paradoxical anterior femoral movement [14], alleviating anterior knee pain (AKP), and providing a better patellofemoral trajectory [11]. Although SR prosthesis designs had theoretical advantages, there has been lack of sufficient quality studies. Most of these studies included no more than 50 patients in each group or followed up for no longer than 3 years [15‐18], as shown in Table 5. Finally, some studies suffered from methodological shortcomings.

The aim of this retrospective cohort study was to compare the clinical results between SR and MR posterior-stabilized (PS) prostheses. We hypothesized that both two prostheses can provide satisfactory results and the SR prostheses provide similar functional, radiological, and survival results to the MR prostheses in TKA.

The retrospective cohort study was approved by the Institutional Review Board of West China Hospital, Sichuan University (ID number: 2012-268). The work was registered in the Chinese Clinical Trial Registry (ID number: ChiCTR1800016129). Informed consent was obtained from all patients and their relatives. Information about deceased participants was collected from their relatives.

This retrospective cohort study included a consecutive series of 327 patients who underwent TKA by one surgery group in West China Hospital between October 2006 and October 2007. We included patients who underwent primary unilateral TKA with SR (Stryker Scorpio NRG) or MR (DePuy sigma PFC) prostheses. The two prostheses were the most utilized prostheses during that time period. We declared no additional information about the implants to patients, so which implant to utilize was irrelevant to characteristics of the patients or severity of the symptoms and was only chosen by patients themselves for their own desires.

We excluded patients with revision TKA, patients diagnosed with inflammatory arthritis, patients who underwent bilateral TKA, patients who required cruciate-retain (CR) or any other types of SR or MR prosthesis, patients who required cones or augments for severe bone defects, and patients who could not give informed consent. Therefore, 65 patients were excluded from the study for any reasons (Fig. 1). At the final 10-year follow-up, 42 patients were lost to follow-up, underwent revision, or died (Table 3). Accordingly, the final follow-up was conducted on 220 patients, among whom 106 received SR prostheses and 114 patients received MR prostheses. Baseline data from the two groups are listed in Table 1. There were no significant differences in terms of mean age, gender, BMI, or diagnoses of disease.

We compared two types of posterior-stabilized prostheses: SR (Scorpio NRG, Stryker Orthopedics, Mahwah, NJ, USA) and MR (PFC, DePuy Orthopedics, IN, USA). We hypothesized that both the two prostheses can provide satisfactory results and the SR prostheses provide similar functional, radiological, and survival results to the MR prostheses. We discovered that the SR prostheses gave less anterior knee pain than did the MR prostheses, while other clinical and functional outcomes did not show any significant differences. Radiological results, complications, and survival of the prostheses were also similar.

We overviewed recent literature of the comparison between SR and MR prostheses and compared our current study to them. As is shown in Table 5, many of the previous studies had conflicting outcomes, none of them provided longer than 5 years follow-up [15‐18, 22]. Although our study was a retrospective cohort study, it was the longest study which included at least 10-year follow-up and examined clinical, functional, radiological, and survival outcomes.

Although TKA had wonderful results in treating severe osteoarthritis, nearly 25% of patients were dissatisfied after TKA, and up to 30% had anterior knee pain or painless noise [1]. MR prostheses were designed based on the J curve theory and have been the most widely utilized prostheses since the 1980s [6, 7]. SR prostheses were designed based on the theory that the actual FEA of the posterior femur condyles was considered a single-radius axis of a cylinder for flexion and extension of the knee through 10–120°. SR prostheses provide a more posterior flexion axis and an increased extensor moment arm to relieve the tension of the quadriceps. Additionally, SR designs maintain middle range stabilization and reduce the paradoxical anterior femoral movement to alleviate AKP.

Our study reported that the SR prostheses gave less anterior knee pain than did the MR prostheses. This is in line with Hamilton et al. [11], who compared 75 SR (Triathlon) with 90 MR prostheses, reporting that the SR designs reduced the worst daily pain (P = 0.003) over 3 years of follow-up; however, Oxford Knee Scores (P = 0.09) and timed functional performance tasks (P = 0.23) did not reach statistical significance. Similarly, Palmer et al. [23] reported that 66.3% of patients with the SR prostheses experienced no pain, less than 54.4% with the MR prostheses. Anterior knee pain had related factors in some of the implant designs. Overstuffing of the knee joint, patella impingement, and instability of the knee may result to pain [24]. Similar sizes of the femoral implants were utilized in both groups, and patellar resurfacing was not utilized in neither patient in our study. Theoretically, SR provided a lateral flexion axis so the crepitus between patella and femoral implants may be less than MR. SR increased extensor moment arm to relieve the tension of the quadriceps as to led to better quadriceps efficiency. In the MR femoral implants, the length of the ligament changes at mid-flexion as the momentary axis changes from a long one to a short one, which may cause in instability, while the SR femoral design has a fixed main axis which maintains the tension of the collateral ligaments during the movement. All of these may explain our follow-up discovery.

SR had theoretical advantages that were likely to translate into satisfactory improved clinical outcomes. We did not observe significant improvement in terms of other clinical, radiological, and survival results when comparing the SR with the MR prostheses in TKA. Several possible reasons could explain this.

First, the clinical scales traditionally used to generate orthopedic results such as HSS and KSS score may not be sensitive enough to illuminate differences in prostheses designs [25]. Mahoney et al. [10] compared 83 patients with SR (Scorpio) to 101 patients with MR prostheses and reported that significantly more patients in the SR group were able to rise from a 16-in. chair without using their arms starting at 6 weeks and the difference was maintained through 2 years; however, there were no differences in the KSS scores between the two groups at 2 years. Jo et al. [14] compared 50 patients with SR (Scorpio NRG) and 50 patients with MR, using a navigation system to measure stability during 0°, 30°, 60°, and 90° of flexion, and reported that the SR group had better intraoperative stability, especially improved midrange stability from 60 to 90°, but no significant differences in the HSS scores or WOMAC scores. Larsen et al. [16] compared 16 patients with SR (Scorpio) to 16 patients with MR prostheses, using computer navigation for gait analysis, and reported that the SR implant had better kinematic properties at 1-year follow-up than did the MR prostheses. MR prostheses remained more extended and had decreased power absorption during weight acceptance than did the SR prostheses. However, both surgical groups had similar KSS for knee and functional scores at the 1-year follow-up. Therefore, any advantages of the SR prostheses in terms of muscular recovery, component stability, or gait cycle may not be accurately reflected in the conventional function scales.

Second, soft tissue and gap balance are influenced by many factors, including different types of osteotomies and the effect of releasing the tight medial-lateral structures. Therefore, the stability of the knee could not be completely guaranteed by the SR prosthesis. Stoddard et al. [15] compared eight patients with SR (Triathlon) to eight patients with MR prostheses using a computer navigation system to measure the stability and reported that significant differences were not found between the types of femur design and that mid-range instability may have been related to unrecognized ligament laxity. Therefore, considering the complexity and multi-step process of the soft tissue balance, SR prostheses did not guarantee the stability of the knee at all angles.

Our study had several limitations. First, it was a retrospective cohort study such that selection bias would be expected. Nevertheless, our study reported long-term follow-up with a large population of consecutive patients. Second, more accurate tests are required to increase the sensitivity of the clinical and functional results.

Both SR and MR posterior-stabilized prostheses can lead to satisfactory outcomes. The SR prosthesis design gave less anterior knee pain than did the MR prostheses. The two prostheses showed no differences in terms of clinical scales, radiological results, satisfaction rate, or survival results at long-term follow-up, requiring more accurate measurements.