|Home | About | Journals | Submit | Contact Us | Français|
We report the medium-term outcomes of the Rotaglide mobile bearing total knee arthroplasty (RTK). Between 1994 and 1999, 357 RTK prostheses were implanted at our institution. Of 150 knees attending for follow-up, none had needed revision. Mean American Knee Society Score (AKSS), Oxford knee score and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score at 8 years were 153.6, 23.1 and 16.5, respectively. Radiological analysis revealed no prosthesis with signs of impending failure requiring revision. Survivorship was 100% in those attending. However, we are aware of two failures requiring revision, both of which were due to infection. We have no reported failures requiring revision due to aseptic loosening. This may be a result of the implant design. Limitations of the study include loss to follow-up and radiological analysis. The RTK gives good to excellent medium-term results and we support its continued use.
Nous rapportons les résultats à moyen terme de la prothèse à plateau mobile de type Rotaglide. Entre 1994 et 1999, 357 prothèses ont été implantées dans notre établissement. Sur 150 genoux pris en compte dans le suivi aucun n’a nécessité de révision. Les scores AKSS, Oxford et Womac à 8 ans ont été respectivement de153.6, 23.1 et de 16.5. L’analyse radiologique n’a pas montré de signes de faillite nécessitant une révision. La courbe de survie est de 100% sur cette période. Nous surveillons deux échecs qui ont nécessité une révision dont tous les deux sont secondaires à une infection. Nous n’avons pas non plus rapporté d’échecs nécessitant une révision pour descellement aseptique. Ces bons résultats sont secondaires au design de l’implant. Les limites de cette étude sont les perdus de vue de l’analyse radiologique. Nous pouvons conclure que la prothèse totale du genou de type Rotaglide RTK permet de bons et d’excellents résultats à moyen terme et nous sommes encouragés à poursuivre son utilisation.
We have used the original unmodified Rotaglide total knee arthroplasty (RTK) (Corin Medical, Cirencester, UK) since April 1994. In the literature to date only two studies have reported results for the RTK with follow-up of longer than five years [14, 15]. One of these publications was limited to a small series of bilateral total knee arthroplasties comparing the RTK mobile bearing to a fixed bearing total knee arthroplasty (TKA). In addition there are three studies showing short-term results with follow-up of between one and five years [3, 9, 10]. In total, approximately 3,000 RTKs have been implanted worldwide, with about 1,000 of these in the UK. In addition, a slightly modified RTK+ prosthesis has been implanted in approximately 27,000 knees with 6,000 of these in the UK.
The aim of this study was to report medium-term (5 to 8 year) follow-up of the RTK using a concise reporting format which allows comparison to future long-term studies .
We included in the study patients who had an RTK at our institution from April 1994 to January 1999. Patient details were obtained from the operation register and from a computer database. Five consultants and 11 registrars performed 108 and 42 operations, respectively. Usually, the patient had a combined spinal and general anaesthetic. Patients were routinely mobilised on the first post-operative day. Patellar resurfacing was not routinely performed, this decision being left to the operating surgeon. Drains, if inserted, were only left in for 24–48 hours. The diagnosis was osteoarthritis in 143 knees (95%), rheumatoid arthritis in five knees (3%) and inflammatory arthritis in two knees (1%). The average hospital inpatient stay was 11 days (range: 4–107).
No other prosthesis was used for primary TKA during this period and all prostheses were the original RTK design. A modified ‘Rotaglide plus’ design was introduced to the market during this period by Corin Medical, but we did not use this.
In the period studied, 357 RTK prostheses were implanted into 317 consecutive patients (357 knees). A total of 52 patients (57 knees) had died and 132 (150) were lost to clinical and radiographic follow-up. This left 133 patients (150 knees) who attended for assessment.
All surviving patients were contacted and invited to attend a special clinic which was exclusively for RTK patients within this study. All attending patients were assessed clinically and radiographically.
The patients were clinically assessed by one of two authors neither of whom had performed any of the primary RTK arthroplasties. Both were Senior House Officers at the time of the study. The first three patients assessed by each observer were also seen by the other observer to try and reduce inter-observer bias. There was no statistically significant difference between observers in any of the clinical scores obtained.
The clinical outcomes were assessed using three scoring systems: the American Knee Society Score (AKSS) , Oxford knee score (OKS)  and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score . Radiographic analysis was performed by a Consultant Musculoskeletal Radiologist (JB) using the system described by Ewald . This conforms to the concise reporting format suggested by Poss et al. .
Kaplan-Meier survivorship analysis was performed using the following endpoints: revision for any cause, revision for aseptic loosening of the femoral component, revision for aseptic loosening of the tibial component and revision for aseptic loosening of the patellar component.
At follow-up there were 52 males (59 knees) and 81 females (91 knees) with a mean age at operation of 70 years (56–91) and mean age at follow-up of 80 years (66–100). There were 73 left-sided and 77 right-sided RTKs. Patellar resurfacing had been performed in 64 of 150 knees (42.7%).
All 150 knees (100%) were still in situ at follow-up. The mean time to follow-up was 79 months (range: 60–96, median: 84 months). We are aware that there were two revisions needed in patients who were lost to follow-up. Both of these were revised for infection between two and three years after the primary RTK. One had a successful two-stage revision procedure and the other underwent a revision which failed due to recurrent infection. This patient’s knee was subsequently arthrodesed.
The mean AKSS scores were 164.1 at five years, 158.3 at six years, 159.3 at seven years and 153.6 at eight years (median: 175, 169, 159 and 153, range: 96–200, 71–200, 108–200 and 83–200, respectively) (Fig. 1). The mean Oxford knee scores were 19.9 at five years, 20.7 at six years, 21.8 at seven years and 23.1 at eight years (median: 17, 17.5, 20 and 21, range: 12–36, 12–44, 12–39 and 13–46, respectively) (Fig. 2). The mean WOMAC scores were 11.8 at five years, 11.4 at six years, 11 at seven years and 16.5 at eight years (median: 8, 4, 6 and 13, range: 0–46, 0–63, 0–60 and 0–70, respectively) (Fig. 3).
At follow-up we obtained radiographs of 64 knees. Radiographic analysis revealed 49 components with 0–4 mm of total radiolucency, 15 with 5–9 mm and none with 10 mm or more (Table 1).
The mean femoral anteroposterior (AP) valgus angle was 96.4° (range: 90–108°) and mean lateral flexion angle 4.3° (range: −10 to 10°). The mean tibial AP position was 88.0° (range: 80–95°). Therefore, the mean tibial position was 2° varus. The mean tibial posterior slope was 8.1° (range: 0–18°). Overall, the tibiofemoral angle was 4.4° valgus.
Survivorship analysis revealed an overall survival of 100% at eight years for this cohort (Table 2). This represents the best case scenario. We are aware of two revisions for infection and the worst case survivorship would be 50%.
Our results support the continued use of the Rotaglide total knee arthroplasty prosthesis.
The limitations of the study include the high loss to follow-up (150 knees, 50%) and low number of patients with radiographs available for analysis at the time of the study. Only the two revisions mentioned were recorded. Both of these were due to deep infection and both were revised between two and three years. Interrogation of the department’s revision database (Manchester Orthopaedic Database) picked up these two revisions, but did not reveal any more of the 150 missing knees as having been revised at the time of the study.
The Manchester Orthopaedic Database has previously been shown to be highly accurate at our institution and the method of data entry has not changed since previous studies [1, 12]. We also conducted a search of 42 sets of notes from patients who had been lost to follow-up due to non-attendance. None of these were found to have been revised.
In light of the above we are of the opinion that the survivorship of the RTK knee is likely to be close to the 100% calculated by the best case scenario. If the known two revisions are the only failures then the survivorship would be 99% at 5–8 years. We do not have any evidence to support the worst case scenario being likely; however, without 100% attendance at follow-up this is impossible to prove.
Radiographic analysis reveals that no components had greater than 10 mm of radiolucency indicating impending failure . Furthermore, there was no statistically significant difference in clinical scores between those with 0–4 mm and 5–9 mm of radiolucency.
The RTK compares favourably with other total knee arthroplasties at medium-term follow-up. At a similar time of follow-up survivorship of the Insall Bernstein II knee arthroplasty is 93.7–95.8% , the Nexgen 95.9–99.5% , the PFC 95.6–98.2%  and Kinemax over 96% . Our study shows survivorship of the order of 99% with no revisions for aseptic loosening. This may be due to the tolerance to wear created by the design of the prosthesis.
The Rotaglide total knee arthroplasty offers patients good to excellent outcomes at 5–8 years. We support its continued use.