The Knee. The Twin Peg Oxford partial knee replacement: The first 100 cases. Stephen H. White a,, Sharon Roberts a, Peter W.

Transcription

1 The Knee 19 (2012) Contents lists available at ScienceDirect The Knee The Twin Peg Oxford partial knee replacement: The first 100 cases Stephen H. White a,, Sharon Roberts a, Peter W. Jones b a Department of Orthopaedic Surgery, The Robert Jones and Agnes Hunt Orthopaedic and District NHS Trust, Oswestry, Shropshire. SY10 7AG, United Kingdom b Department of Medical Statistics, Keele University, Keele, Staffordshire. ST5 5BG, United Kingdom article info abstract Article history: Received 18 October 2010 Received in revised form 14 December 2010 Accepted 15 December 2010 Keywords: Unicompartmental knee replacement Oxford knee replacement Twin Peg Hiflex knee Anteromedial osteoarthritis We present the clinical results of the first 100 patients who received the Twin Peg Oxford partial knee replacement which has a 15 extra femoral surface for contact in deep flexion, and has two pegs for more secure fixation. We measured the clinical outcome 2 years after the medial unicompartmental arthroplasty using patient and surgeon derived outcome measures. The results showed a mean Oxford Knee Score of 41, a mean American Knee Society Objective Score of 93 and a Functional Score of 84, a mean range of motion of 130 and a high satisfaction rate. Results were significantly better in male patients. There were no deaths, infections, dislocations, fractures or revisions. There were no pathological radiolucencies suggestive of early loosening. We conclude that the Twin Peg Oxford partial knee replacement shows excellent clinical and radiological results at 2 years. For surgeons who have concern over the risk of femoral loosening with the single peg Oxford knee, or seek an improved surface area of contact in full flexion, this implant offers an excellent alternative Elsevier B.V. All rights reserved. 1. Introduction The Twin Peg Oxford partial knee is a modification of the Oxford knee designed with two pegs to improve femoral fixation and a longer articular surface to reduce contact stresses in deep flexion (Fig. 1). The Oxford knee has proved highly successful in many centres [1 3], but some have found no better pain relief and function of unicompartmental knee replacements compared to total knee replacement and a higher proportion of failures due to early femoral loosening [5]. We had experienced problems with early femoral loosening and were keen to use a more secure two peg version and so encouraged the manufacturers to provide such a device [6,7]. In 2003 a-two peg femoral prosthesis was made available having been primarily designed for patients in the Far East where it is common for flexion to reach 165 after unicompartmental knee replacement. With such extreme flexion, there is abnormal loading and the potential for polyethylene wear due to from reduced femoro-tibial contact, raising the risk of femoral component loosening. The new femoral component incorporating the extra peg increases the arc by 15, and can be inserted 15 to the long axis of the femur, thereby adding more contact with the bearing at high flexion angles. The implant initially called the Hiflex knee, is identical and now called the Twin Peg Oxford partial knee (Biomet UK Ltd). Because of our concern over femoral loosening we decided to discontinue using the single peg component and instead Corresponding author. Tel.: ; fax: address: Georgi.Norris@rjah.nhs.uk (S.H. White). use the two peg implant. With the approval of our new procedures committee and our patients' informed consent we changed to this Twin Peg version in In this paper we describe the audit of clinical outcome at 2 years after implantation. 2. Patients and methods 2.1. Patients and outcome measures The first 100 patients who underwent Twin Peg Oxford partial knee replacement were studied in our two year outcome clinic. Because some patients had staged bilateral knee replacements there were 108 knees. The diagnoses were osteochondritis dissecans (2), post traumatic arthritis (3) but the majority of cases (95) had anteromedial osteoarthritis [8]. All patients exhibited clinically correctable varus deformities, anteromedial tibial erosions on lateral radiographs, and at surgery had intact anterior cruciate ligaments and satisfactory lateral compartments. At the 2 year outcome clinic patients completed the Oxford Knee Score (OKS), the American Knee Society Objective score (AKS), and the American Knee Society Function Score (AFS). In addition, we recorded the Charnley fitness classification and the range of motion of each knee [9]. A satisfaction questionnaire form was completed and a form to record any complications that had occurred during the first 2 years. We were unable to follow up 13 patients. One distant patient had been transferred to another consultant with symptoms suggestive of patellar tendinitis. Four patients had died of unrelated causes, and /$ see front matter 2010 Elsevier B.V. All rights reserved. doi: /j.knee

2 S.H. White et al. / The Knee 19 (2012) Operative technique Fig. 1. Comparison of the Phase III Oxford partial knee (left) and the Twin Peg Oxford partial knee (right) which is implanted in 105 of flexion, and has a longer arc of metal for contact in deep flexion. eight patients did not attend mainly because of distance or other medical conditions; the last entry in the medical records of these 12 patients showed no complications, revisions or further surgery. The post operative radiographs were analysed to see whether at 2 years there were radiolucencies of the femoral component interface on the lateral radiograph at the posterior flexion facet or at the anterior extension facet (see Fig. 2). We were unable to obtain screened views at our hospital. However, our radiographers made every effort to align the lateral radiograph to the femoral component and the AP radiograph to be in line with the tibial component, repeating the film if necessary. Both AP and lateral radiographs were viewed for femoral or tibial cementophytes. The angle of insertion of the femoral component was measured by comparing a line drawn along the shaft of the femur on the lateral radiograph and the line along the long peg of the femoral component. We defined the standard phase III femoral angle as 90 and the Twin Peg as 105 or a variation of that angle. The AP radiograph was viewed for lateral compartment arthritis (narrowing of the lateral compartment joint space below 5 mm) [10]. We positioned the patient as for a total knee replacement but with the sandbag supporting the foot at a knee flexion angle of approximately 105. A reduced invasive incision allowed both the surgeon and assistant to see inside the joint with ease. Diathermy was used to mark the likely contact points of the femur with the tibia throughout flexion. First we resected the tibial plateau. Our standard technique was to use the normal femoral drill guide positioned at a greater angle of knee flexion than the usual 90, estimated at 105. The extramedullary femoral drill guide was aligned in the coronal plane towards the centre of the hip which was estimated as two finger breadths medial to the anterior superior iliac spine. The femoral drill holes were made to align with the contact marks on the femoral surface. To prevent impingement, more bone and cartilage anterior to the milled femoral surface required removal using chisels because the Twin Peg femoral component is 15 longer than the phase III implant. To measure the extension gap, first the feeler gauge was removed and the knee extended to approximately 20. The difference in millimetres between the 20 extension gap and the 105 flexion gap determined the size of the spigot which was required to ream more of the distal femur. The usual stages were followed, and after cementing the components the appropriate anatomical meniscal bearing was implanted Statistical analysis Measures were summarised by using mean and standard deviation in most cases, median and range are also quoted where appropriate. Statistical analysis was carried out on the data from the second operation on staged bilateral cases when functional outcome was involved since the statistical tests used assume that the measurements are independent. The means in two groups were compared by using t tests and in more than two groups by using analysis of variance (ANOVA) with a Bonferroni adjustment for multiple testing of pairs at the 5% level. Association between measurements was measured by using Pearson's correlation. Where necessary, non parametric alternatives were used. 3. Results 3.1. Complications No patients suffered a complication ofdeath, pulmonary embolus, fracture, dislocation, deep infection or medial collateral ligament division. There were lesser complications relating to surgery as follows. One patient had symptoms suggestive of patellar tendinitis, two patients had superficial wound infections which responded to oral antibiotic therapy, one patient suffered post operative pneumonia and one patient a kidney infection. All of those patients recovered well with antibiotics. There were two cases ofunexplained medial knee pain. No patients underwent revision surgery or were listed for revision Clinical outcome measures At 2 years the mean OKS was 40.8 (SD 7.1), as shown in Fig. 3. The mean AKS was 92.8 (SD 12.1) and the mean AFS 83.8 (SD 13.5). The AFS was significantly lower (pb0.001) for patients in Group C than for patients in the A and BB groups (Table 1). Of the 87 patients who responded to the satisfaction questionnaire 84 (97%) were satisfied, two were dissatisfied and one was undecided. Range of motion. No patient had a fixed flexion deformity. The median range of flexion was 130 (range ), indeed 83% of patients had at least this range of movement. Bearing size. The outcome of the replacement according to bearing size showed no significant correlation with the OKS, AKS or AFS outcome measures. Patients under the age of 50 compared to patients greater than or equal to that age showed no significant difference in outcome. There was no significant difference comparing left or right knee replacements. Gender differences were apparent (see Table 1). Women achieved an OKS of 38.3±8 and men an OKS of 42.8±5.8 (pb0.01). The gender differences in AKS (pb0.001), and AFS (pb0.01) were also highly significant Radiographic results Fig. 2. Lateral radiograph of a patient with the Twin Peg Oxford knee. This patient had an OKS of 44, an AKS of 100 and an AFS of 100. We observed no radiolucencies of the cement bone interfaces of the femoral or the tibial components 2 mms or greater in this series. Observations were made however of

3 38 S.H. White et al. / The Knee 19 (2012) Fig. 3. Distribution of Oxford knee scores in the 87 patients for whom it could be calculated. minor radiolucencies with regard to the femoral component in four cases and minor retention of cement in three cases (Table 2). These patients were asymptomatic. The OKS and AKS scores were excellent for the two patients who had a lateral joint space gap of less than 5 mm. Femoral angle. The mean angle of implantation of the femoral component was ±5.7. There was no significant correlation of the angle with the OKS, AKS or the AFS. 4. Discussion These results show a high quality of outcome with a low incidence of complications using the Twin Peg femoral component. The subjective assessment gave a mean Oxford knee score of 40.8, and a high American Knee Society objective rating of These results compare favourably with the best published outcome of the phase III prosthesis [3], and are better than the OKS of 36.5 of our previous results with the phase II prosthesis and our OKS of 38 using the phase III prosthesis [6,7]. The Twin Peg femoral component is positioned in greater flexion than the phase III Oxford single peg component and our patients achieved a remarkable degree of knee flexion with a median range of 130. There was a 97% satisfaction rate and no severe complications. Our results also compare favourably with national standards. New benchmarks have been set for knee replacement surgery following a study of 10,000 randomly selected cases at least 1 year after surgery in the UK National Joint Registrar [12]. The OKS of 40.8 with the Twin Peg knee is considerably higher than the national average for knee replacement of 35, and our patients 97% satisfaction rate is higher than the UK average of 82% and the Swedish Arthroplasty Registry average of 81% [13]. Table 1 Outcome of the knee replacement at 2 years according to Charnley classification of mobility, and by gender. Results expressed as mean +/ one standard deviation. *includesunclassified cases. Charnley classification Number of patients OKS AKS AFS Range of motion Mean Mean Mean Mean All patients 87* 40.8 (7.1) 92.8 (12.1) 83.7 (19.9) 130 (7.1) A (7.0) 95.1 (11.5) 93.1 (13.2) 132 (6.5) BB (4.5) 96.0 (6.5) 89.3 (17.2) 132 (5.6) B (4.7) 94.3 (11.0) 83.6 (16.0) 133 (4.2) C (7.8) 88.2 (14.5) 70.7 (21.4) 127 (5.9) Men (5.8) 97.1 (4.9) 89.5 (15.3) 132 (7.0) Women (8.0) 87.6 (16.0) 76.3 (22.8) 129 (7.1) The radiological assessment of radiolucencies needs to be interpreted with caution as we did not have the benefit of screened views. However, we observed on our standard views satisfactory bone cement interfaces without early signs of femoral or tibial loosening at 2 years. With respect to the femur there were no lucencies around the femoral pegs or at the distal facet. Only five patients had detectable radiolucencies and these were at the posterior bone cement interface and with a maximal width of 1.2 mm. Goodfellow et al. [10] recognised the physiological lucent line as almost always less than 2 mm thick and defined by a thin radiodense bone plate; the pathological lesion is thicker, and the margins of the radiolucent zone are characteristically ill defined. Our radiolucent lines corresponded to this physiological type. A recent cadaveric study has shown that even with screened views accurate radiographic imaging of the posterior facet is usually imperfect [11]. Although without screened views we will have underestimated radiolucencies at some interfaces, the viewing of the anterior and posterior interfaces of the femoral pegs, being cylindrical, would not be affected by minor variations in rotation of the X-ray beam. With the advantage of modern PACS magnification we can confirm that no radiolucencies were seen along the femoral pegs to indicate loosening. Clarius et al. [11] have emphasised the importance of cementation of the femoral peg because it is difficult to achieve the same degree of interdigitation at the posterior facet. A surprise finding was a difference in outcome according to gender. Although female patients gained excellent relief of pain and improvement of function with a mean OKS score of 38.3, male patients had Table 2 Characteristics of cases with radiographic abnormalities. Satisfaction Charnley classification OKS AKS AFS ROM Femoral wedge sclerotic Yes BB mm Femoral wedge sclerotic Yes C mm Femoral wedge sclerotic Yes BB mm Femoral parallel posterior sclerotic margin 1.2 mm Yes C Posterior cement retention Yes C Posterior cement retention Yes A Medial tibial cementophytes Yes BB Lateral joint space 3.45 mm Yes BB Lateral joint space 2.6 mm Yes C

4 S.H. White et al. / The Knee 19 (2012) significantly better results with a mean OKS score of 42.8 (pb0.01). Significant differences were also apparent for the AKS and AFS scores (see Table 1). To our knowledge this is the first time a gender difference has been reported for the Oxford knee. The difference may therefore be particular to this Twin Peg version, however, other workers have found gender differences in outcome for some other knee prostheses [14]. Interestingly, data from the UK National Joint Register shows that women are less often satisfied with their total knee replacement than men and have lower Oxford knee scores [12]. The reasons for this are as yet unclear but one possible explanation may be due to variation in the anatomy of male and female knees, for differences are being increasingly recognised [15]. Chau et al. [16] have recently demonstrated a 10 point drop in the OKS with overhang of the tibial component by 3 mms or more. It became apparent in our work that some women's knees were too narrow for the standard tibial component. By carefully recessing the intercondylar wall the tibial component can usually be accommodated without serious overhang but it is important not to encroach too far on to the footprint of the anterior cruciate ligament. A narrower version of the tibial component may be the answer. A strength of our study is the analysis of a sizeable number of patients from a number of different perspectives. It is known that instruments for measuring outcome of knee replacements each have their own short comings so we have used a variety of measures, all of which show a high quality of outcome [17]. Because the AKS score has a flexion limit of 125 we made an additional record using a long arm goniometer to record the total range [18]. No patient in our study was left with a fixed flexion deformity, and we observed a very high median flexion of 130 (range 100 to 140 ) with many patients achieving full flexion. However, we are also aware of the deficiencies of our study. There is a possibility that our improved results with the Twin Peg knee are not accounted for by the design changes but simply due to improved surgical technique. Surgical expertise has been found to be an important determinant of outcome of the Oxford knee [19]. In any event we consider the improved surface contact stresses at high flexion and enhanced fixation compared to the phase III knee strongly recommend adoption of this Twin Peg design. Another deficiency was the lack of sufficient pre-operative scores to make before and after comparisons; also, inevitably some patients were lost to follow-up. Nevertheless, their medical records did not indicate worse outcomes than those who did attend for full analysis. In our previous experience loosening tends to be apparent early. We reported the outcome of 28 patients with the Oxford Phase II implant at a mean of 2.3 years follow-up and detected 3 cases of loosening using similar clinical and radiographic criteria to those used in our current study [6]. Early failure, in particular loosening has also been reported in a larger study [20]. Similarly, in the SKAR report of 2004, the most common cause for failure of UKA was loosening of a component [5]. Lewold et al. [4] found that components tended to loosen early, the mean time to revision from this cause was 26 months. Although longer term studies will be required to detect failure due to loosening, there is some reassurance from our most recent figures in the National Joint Registry for England and Wales, which coincidentally has been monitoring knees since 2003, the same year that we introduced the Twin Peg Oxford knee. In November 2010 the Patient Time Incidence Rate (PTIR) for revision of our 229 Oxford unicompartmental knee replacements is 0.21%, which is well below the National revision rate of 9.1% at 5 years [21]. Our patients achieved a high degree of knee flexion, a median and mean of 130 but is this due to the design of the prosthesis or not? The phase II Oxford knee replacement as reported in 1988 [22] achieved a mean angle of maximum flexion of 105. Nowadays, patients expect a much greater range of flexion and this has already been achieved with the Oxford phase III implant which has a mean range of 133 [3]. Our results with a mean of 130 show that the Twin Peg Oxford knee confers no greater flexion than this. We therefore conclude that surgeons should not expect better flexion using the Twin Peg knee instead of the conventional Oxford Phase III component. The advantage of the Twin Peg Oxford knee lies not in greater flexion range but in greater congruency between the femoral component and the meniscal bearing at high flexion (see Fig. 1). This should be beneficial for stress distribution and surface wear but long term studies will be needed to confirm this. It would be possible to provide a modified femoral drill guide at the 105 instead of 90 but we were happy to use the 90 standard guide and estimate the 105 insertion angle. Reassuringly, we found no correlation between the femoral angle of insertion as measured on the radiographs and the clinical outcome. Thus, we conclude that this technical point is not critical. There is increasing interest in partial knee replacement as some centres are reporting superior results with unicompartmental rather than total knee arthroplasties in terms of lesser complications, more rapid rehabilitation, a greater range of movement and a lower revision rate [3,23]. Surgeons who have until now preferred total knee replacement may therefore find the Twin Peg Oxford knee an attractive option. Firstly, the set up on the operation table is virtually identical to that of total knee replacement, all that is required to change to a total knee is to adjust the knee flexion angle from 105 to 90 by moving the sandbag. Secondly, the operative incision does not require special training in minimally invasive exposure since it conforms to the lower part of a standard, anterior incision and can be easily extended proximally for a total knee replacement. Our exposure gives a good opportunity to assess the state of the rest of the knee joint and our results show a successful outcome using this approach. This is not to say that a minimally-invasive approach could not be used; it may well be that smaller incisions will achieve still more successful results than our own. 5. Conclusions In conclusion, we have demonstrated that the Twin Peg Oxford knee can be safely implanted and provide excellent pain relief and function 2 years after surgery. Nearly normal flexion is to be expected and patient satisfaction is high. Future research is required concerning gender difference in outcome; a narrower tibial component may be required in some cases. Research will also be needed to confirm whether the theoretical benefits of improved stress at high flexion with the Twin Peg knee is translated into better long term survivorship of the prosthesis. 6. Conflict of interest statement None of the authors of this work have any conflict of interest. References [1] Svard UC, Price AJ. Oxford medial unicompartmental knee arthroplasty. A survival analysis of an independent series. J Bone Joint Surg Br 2001;83-B: [2] Keys GW, Ul-Abiddin Z, Toh EM. Analysis of first forty Oxford medial unicompartmental knee replacements from a small district hospital in UK. Knee 2004;11: [3] Pandit H, Jenkins C, Barker K, Dodd CAF, Murray DW. The Oxford medial unicompartmental knee replacement using a minimally-invasive approach. J Bone Joint Surg Br 2006;88-B: [4] Lewold S, Goodman S, Knutson K, Robertson O, Lidgren L. Oxford meniscal bearing versus the Marmor knee in unicompartmental arthroplasty arthrosis. A Swedish multicentre survival study. J Arthroplasty 1995;10: [5] Lidgren L, Knutson K, Robertson O. Swedish Knee Arthroplasty Register. Annual Report Lund; [6] Weale AE, Halabi OA, Jones PW, White SH. Perceptions of outcomes after unicompartmental and total knee replacement. Clin Orthop Relat Res 2001;382: [7] Luscombe KL, Lim J, Jones PW, White SH. Minimal invasive Oxford medial unicompartmental knee arthroplasty: a note of caution. Int Orthop 2007;31:

5 40 S.H. White et al. / The Knee 19 (2012) [8] White SH, Ludkowski PF, Goodfellow JW. Anteromedial osteoarthritis of the knee. J Bone Joint Surg Br 1991;73-B: [9] Charnley J. The long-term results of low-friction arthroplasty of the hip performed as a primary intervention. J Bone Joint Surg Br 1972;54B: [10] Goodfellow J, O'Connor J, Dodd C, Murray D. Unicompartmental arthroplasty with the Oxford knee. Postoperative management and radiography. Oxford University Press; p [11] Clarius M, Mohr G, Jaeger S, Seeger JB, Bitsch RG. Femoral fixation pattern in cemented Oxford unicompartmental knee arthroplasty an experimental cadaver study. Knee 2010;17: [12] Baker PN, van der Muelen JH, Lewsey J, Gregg PJ. The role of pain and function in determining patient satisfaction after total knee replacement: data from the National Joint Registry for England and Wales. J Bone Joint Surg Br 2007;89-B: [13] Robertsson O, Dunbar M, Pehrsson T, Knutson K, Lidgren L. Patient satisfaction after knee arthroplasty: a report 27, 372 knees operated on between 1981 and 1995 in Sweden. Acta Orthop Scand 2000;71: [14] Singh JA, Gabriel S, Lewallen D. The impact of gender, age, and pre operative pain severity on pain after total knee arthroplasty. Clin Orthop Relat Res 2008;466(11): [15] Varadarajan KM, Gill TJ, Freiberg AA, Rubash HE, Li G. Gender differences in trochlear groove orientation and rotational kinematics of human knees. J Orthop Res 2009;27(7): [16] Chau R, Gulati A, Pandit H, Beard DJ, Price AJ, Dodd CA, Gill HS, Murray DW. Tibial component overhang following unicompartmental knee replacement does it matter? Knee 2009;16: [17] Dawson J, Carr A. Outcomes evaluation in orthopaedics. J Bone Joint Surg Br 2001;83-B: [18] Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop 1989;248:13 4. [19] Goodfellow J, O'Connor J, Dodd C, Murray D. Unicompartmental arthroplasty with the Oxford knee. The surgeon as a variable. Oxford University Press; p [20] Saldanha KAN, Keys GW, Svard UCG, White SH, Rao C. Revision of Oxford medial unicompartmental knee arthroplasty to total knee arthroplasty results of a multicentre study. Knee 2007;14: [21] National Joint Registry for England and Wales. 7th Annual Report 2010; ; p (print). [22] Goodfellow JW, Kershaw CJ, Benson MKDA, O'Connor JJ. The Oxford knee for unicompartmental osteoarthritis. J Bone Joint Surg Br 1998;70-B: [23] Newman J, Pydisetty RV, Ackroyd C. Unicompartmental or total knee replacement: the 15 year results of a prospective randomised controlled trial. J Bone Joint Surg Br 2009;91-B:52 7.