Unicondylar knee replacements: the Oxford perspective

Abstract

Anteromedial osteoarthritis is a common phenomenon in degenerative conditions of the knee. This limited wear pattern is contingent on a functioning anterior cruciate ligament and medial collateral ligament, and can be managed differently to tricompartmental or inflammatory arthritis. The UK National Institute for Health and Care Excellence (NICE) stipulate that patients with this pattern of disease should be offered the choice of either unicondylar (UKA) or total knee replacements, but many knee surgeons are either unfamiliar with UKA or do not readily include it in their practice. The Oxford UKA is an established and trusted prosthesis with excellent outcomes, yet cumulative marginal gains that improve outcomes further can be achieved by following the approach of high-volume centres. In this article we discuss the Oxford perspective on this subject, touching on the design philosophy, indications, perioperative pearls, day-case protocols, outcomes and tips for starting practice. The most common number of UKAs performed by knee surgeons a year is one; this article aims to provide readers with the understanding, confidence and tips to optimize UKA use within their practice and maximize patient outcomes.

Introduction

The lifetime risk of knee arthritis is 44.7%. Since the 1950s, surgeons have been aware of anteromedial osteoarthritis (AMOA), and efforts to address this with arthroplasty have evolved with time. Early prostheses included medial spacers and fixed bearing unicondylar knee replacements, but it was not until the 1970s that the concept of a mobile polyethylene bearing was conceived. Based on the meniscal work by Faribank, the mobile bearing aimed to emulate the intact native meniscus by not only being mobile, but by also conforming to the femoral condyle, increasing the congruency between the femoral and tibial components. It was proposed that a mobile bearing would not only reduce point loading, but also reduce the shear forces and resulting polyethylene wear responsible for the most common mode of failure of the fixed bearing implants available at that time. Initially this novel mobile bearing was implanted as a bicondylar knee replacement, and in 1982 the Oxford Partial Knee replacement was first used. The Oxford Unicondylar Knee Arthroplasty (UKA) has evolved, with changes to the prosthesis and instrumentation over time, but the fundamental design of a spherical femoral resurfacing and a flat tibial base plate articulating with a fully-congruent mobile bearing remains unchanged.

National joint registries (NJRs) across the world tell us that knee arthroplasty procedure numbers are increasing year-on-year. The proportion of all UKA is slowly increasing, but there remains a mismatch between the prevalence of AMOA and the number of unicondylar knee replacements performed, with many surgeons still preferring to treat these patients with total knee replacement (TKA). The 20 ^th National Joint Registry Report, published in 2023, demonstrates that only 13% of primary knee arthroplasty procedures were a UKA in 2022, with 85% being TKRs, a ratio of 1: 6.5. There are numerous hypothesized reasons for this, including confusion over the correct indications, concern regarding disease progression and, ultimately, concern about the risk of revision surgery rates. Kozinn and Scott published their paper on indications for all UKA in 1989, in which young age, increasing weight and level of activity were all included, unfavourably. In addition, evidence of early patellofemoral joint (PFJ) wear was a contraindication. These proposed contraindications are not applicable to the Oxford UKA, and there is strong evidence to support its use even when these are factors are present. The fixed bearing designs that Kozinn and Scott based their judgement on fail differently to mobile bearing designs: being more susceptible to point loading and shear forces, they demonstrate increased poly wear and implant loosening. The implant bearing interface of the Oxford UKA, however, is congruent, so wear is more akin to that seen in total hip replacements i.e. linear wear, specifically measuring 0.01 mm/year with modern ultra-high molecular weight polyethylene (UHMWPE).

It is clear that if only 13% of primary knee arthroplasties are UKA, then many surgeons will pass through training programmes with no, or very little experience of UKA, which will perpetuate the issue of patients receiving TKRs for AMOA. Around the world, it is accepted that education for surgeons who wish to specialize in treating knee arthritis must improve and should include training in UKA.

All the global joint registries have shown an increased risk of revision of UKA compared with total knee arthroplasty (TKA). A large amount of work has been undertaken to understand this phenomenon, and over time the revision rate for UKA has significantly improved. Large studies have demonstrated that surgeons’ thresholds for revision, as measured by Patient Reported Outcome Measure (PROMs), is lower for UKA than for TKR. Revision of TKR can be a large undertaking, and is seen as more complex than revising a UKA; therefore, both clinicians and patients alike tolerate a lower satisfaction score for TKA. Liddle et al. demonstrated in 2014 that revision rates between the two implant types is actually similar when controlling for variables. Liddle also demonstrated an increased likelihood of having an excellent outcome following an Oxford UKA compared to a TKR at 1 year, with reduced risks of dissatisfaction.

One of the criticisms often brought against the Oxford UKA is that the outcomes (patient satisfaction and rates of revision surgery) observed at the design centre are difficult to reproduce elsewhere, but several centres have been able to achieve comparable results. It is recognized, however, that a wider improvement in outcomes would not only benefit patients, but also healthcare institutions and the wider economy too. The Danish experience is an excellent example of how a national overhaul of practice in relation to Oxford UKAs has resulted in improved PROMs, day-case surgery rates and revision rates, similar to that of TKRs.

As with mastering any dextrous activity, repetition is key. The more familiar an individual is with the steps, the better the task is often performed. Minimum procedure numbers are becoming increasingly common in arthroplasty, to denote continued proficiency and appropriateness to continue operating. The most common number of Oxford UKAs performed by a surgeon each year however is one, with the second most common number performed annually being two. As demonstrated by Murray et al., the revision risk is highest for these low users, at between 4% and 5%, and optimized to 1–2% when UKA use comprises between 20% and 60% of one’s knee arthroplasty practice.

Other modifiable elements of practice are harder to demonstrate so poignantly, but a summary of ‘The Oxford Perspective’ surrounding the use of UKAs will provide a comprehensive review of the topic and our standards of practice. This article will elaborate on the design philosophy, indications, clinical work-up, intraoperative process, day-case and rehab protocols, PROM surveys, managing the painful UKA, and tips for starting practice.

Design philosophy

It is imperative to understand the design philosophy of the Oxford UKA. Knowledge in this area justifies one’s rationale for choosing this implant, identifying suitable patients and reaching the optimal percentage of practice to minimize revision rates.

An intact anterior cruciate ligament (ACL) is the key to this limited AMOA disease pattern, as it prevents the tibia from subluxing forward under the femur. As a result, the posterior tibial cartilage and subchondral bone remains preserved, despite chondral wear in the anterior portion. This can be seen macroscopically and also radiographically, on lateral radiographs, as demonstrated in Figures 1 and 2 . Not only does this wear pattern signify a clinically intact ACL, but it also preserves normal medial collateral ligament (MCL) length. In extension, the knee often exhibits a varus deformity, secondary to chondral loss and bone loss, with de-tensioning of the MCL; but in deep flexion, coronal alignment is restored by the preserved cartilage, and the MCL is returned to its normal functional length. It is a functional MCL that permits the kinematically-aligned resurfacing that the Oxford UKA delivers.

As presented in the introduction, the Oxford mobile bearing functions in a similar manner to the native meniscus. Conforming to the single radius of curvature of the femoral component, it is fully congruent and distributes load evenly throughout the range of flexion. As we know from other kinematic studies, the medial joint is less mobile than the lateral compartment, but the contact point does still vary during movement. The bearing’s flat tibial side facilitates this movement in the replaced knee, and as such, the joint is not artificially constrained.

With regards to the femoral component design, the single radius of curvature makes it easier to restore the isometric ligament balancing throughout the whole range of movement, compared to a poly-radial curve, as found in other implant designs. Furthermore, with the use of computed tomography (CT) for research and implant design, it has been demonstrated that the native medial femoral condyle is similarly a single radius of curvature, which is what led to the popularization of the concept of a ‘kinematically aligned’ knee replacement.

It is this restoration of the native kinematics and the preservation of ligaments that is responsible for the excellent outcomes achieved with Oxford UKAs; but in addition, this also reduces failure through wear. With shear forces markedly reduced compared to fixed bearing devices, the mode of failure is akin to that seen in total hip arthroplasty; namely, linear and volumetric poly wear. In the absence of impingement, wear is just 0.01 mm/year.

Indications

Indications for a mobile bearing Oxford UKA have remained broad and inclusive since the introduction of the technique in the 1980s, and there are good long-term data to support its use in all age groups. Hamilton et al. undertook a review of 1000 consecutive UKAs performed in Oxford: within this cohort, 68% would have been considered to have had ‘contraindications’ according to Kozinn and Scott’s criteria; however, there was no difference in outcomes in these patients compared to those who did not have the ‘contraindications’. In order to reach the optimal UKA usage within one’s arthroplasty practice (associated with the lowest revision levels), it is best to take an approach whereby all patients are considered for potential UKA, with unsuitable candidates being ruled out. An intuitive radiographic assessment tool has been devised in order to simplify decision-making when assessing suitability for Oxford UKA, and it is recommended that this is referred to when gaining familiarity with offering this procedure. This clinical decision-making aid is demonstrated in Figure 3 , and focuses on:

•

Medial bone-on-bone disease
•

A functionally intact ACL, with an anteromedial wear pattern
•

Full-thickness lateral cartilage, with a preserved lateral joint space on valgus stress radiographs
•

A functionally normal MCL, with restoration of the medial joint space on valgus stress radiographs, i.e. a correctable deformity
•

An acceptable patellofemoral compartment: absence of lateral patellar facet OA with bone loss, lateral subluxation or grooving.

It is possible to perform medial UKA in patients with ACL deficiency, and it is commonly stated by fixed bearing users that this is only technically feasible with a fixed bearing. It should be noted, however, that their appropriateness is only for ACL-deficient individuals with a clinically stable knee. An unstable ACL-deficient knee with altered kinematics is at significant risk of disease progression with any UKA, and for this cohort a TKR is more appropriate. The Oxford group have also previously reviewed the use of their UKAs in clinically stable but ACL-deficient patients, and concluded that primarily, it is the biological age, functional demands and primary symptoms that need to be considered when deciding whether UKA is suitable. A study of 50 cases undertaken in ACL-deficient individuals at the design centre demonstrated excellent outcomes, although the authors suggested that simultaneous or staged ACL reconstruction was more appropriate than UKA alone in the younger patient cohort.

Whilst age needs to be considered in all patients undergoing joint arthroplasty, there are not many other surgical options for younger patients with established bone-on-bone disease. Differentiating between those patients who are suitable for realignment procedures (offloading osteotomy) and arthroplasty is relatively clear, with most specialist knee surgeons now seeing the indications quite separate, with little overlap for debate. If the radiographic assessment clinical decision tool is used, then patients without bone-on-bone OA are unsuitable for UKA, whereas those with Kellgren Lawrence Grade 4 OA would have less predictable outcomes with osteotomy.

Clinic work-up

The outpatient clinic provides the first opportunity to consider UKA with one’s patients. Obtaining a thorough history, clinical examination and plain film radiographs are all that is usually required to determine suitability for a UKA. In Oxford, whilst salient history points include the duration, severity and impact of pain, the location of pain does not need to be medial. A study published in Knee in 2012 showed that the location of pain in degenerative disease does not accurately correlate with the compartment affected. Eliciting previous treatments or conditions, including lateral meniscectomy, rheumatological disease or knee instability, is also imperative, to identify true contraindications.

On examination, an assessment of correctability of the varus intra-articular deformity and the degree of any fixed flexion deformity is undertaken, in addition assessing MCL and ACL stability. General lower limb assessment for ulcers, pedal pulses and soft tissue quality is, of course, as relevant as ever when considering any arthroplasty.

Radiologically, there are a series of plain film radiographs that are standard practice in Oxford in order to apply the clinical decision-making tool. A weight-bearing AP radiograph of the knee is required, and/or a Rosenburg view, to demonstrate more subtle or focal degeneration. A lateral radiograph is key to assess the anteromedial wear pattern that indicates an intact ACL and a functional MCL. A patellar skyline view is needed to formally assess the PFJ and to ensure the absence of exclusion criteria features, such as lateral subluxation or corrugations of the patella. Stress radiographs can be extremely useful, and are also part of the radiological decision aid. The valgus stress view is important for two reasons: first, it allows one to assess whether any varus is correctable, with a functional MCL, and second, it shows whether the lateral compartment joint space remains preserved. Valgus stress views can be achieved either manually, by the clinician wearing lead gloves, or by placing the leg into a jig. It is important to reinforce that MRI is not part of our standard work-up, as in the majority of cases all the relevant information can be obtained from clinical assessment and plain film radiographs alone; however, MRI remains available when it is deemed that additional information is required.

When discussing the option of UKA and TKR with patients in clinic, shared decision-making is essential. It is one’s job as a surgeon to provide patients with the available information on how their condition can be managed. Given the potential benefits of UKA for patients, it is now mandated by NICE that anyone receiving arthroplasty specifically for AMOA should be offered UKA as an option. Whilst some patients wish to avoid surgery at all costs, others will have exhausted non-operative options such as weight loss, physiotherapy, analgesia (including corticosteroid injections), walking aids and orthotics. In Oxford, patients are advised of the advantages of UKA compared to TKR, for those for whom it is suitable, and these include:

•

Reduced postoperative risks and complications – work by Liddle et al. demonstrated that for a matched population, there was approximately 50% the risk of major complications with a UKA than a TKR: This included 50% the risk of myocardial infarction, cerebrovascular accident, venous thromboembolism and pulmonary embolism. Similarly, there was a reduction in the 90-day mortality, and there was half the infection risk with UKA.
•

Faster patient recovery – matched cohort data show a shorter length of hospital stay, better range of movement on discharge and at 6 weeks, and better functional scores at 6 weeks.
•

Better functional outcomes – Liddle’s group, again, have published on their matched cohort, and have shown that UKA patients are more likely to report excellent or good outcomes following surgery than patients undergoing TKR. Similarly, the randomized controlled Total or Partial Knee Arthroplasty Trial (TOPKAT) showed that 82% of UKA patients were satisfied with their UKA, compared to 77% with TKR, 95% of UKA patients reported that they were better following surgery, compared to 90% after TKR, and 91% of patients would have their UKA again, compared to 84% in the TKR group.

The main criticism that is raised against UKA is the higher revisions rates that are seen after UKR compared to TKR, as demonstrated by the NJR and by a number of publications. The reason for revision may, however, not be poorer clinical outcomes, according to Liddel’s study looking at 14,000 matched patients, but rather a lower threshold for revising a UKA than a TKR. For both the surgeon and the patient, the increased surgical complexity and increased complication profile involved with TKR revision makes revision TKR an unattractive prospect. The Oxford experience is that we do not have a higher revision rate for our UKAs compared to our TKRs, and we explain this to patients. As a group, we believe that this can be explained by the fact that revision rates are inversely proportional to surgeon numbers, in terms of usage percentage, and we are high users. In addition, our strict adherence to the inclusion criteria, clinical assessment and other steps, as described in this paper, maximize our outcomes. A common criticism of the Oxford UKA is that design centre outcomes cannot be replicated; however, our Danish colleagues have worked hard over the last 20 years to follow our approach, and in doing so have reduced their revision rates for UKR to near those of TKRs across the whole national registry.

Intraoperative considerations

The Oxford UKA technical guide is an excellent tool for walking surgeons through the steps of performing surgery, but intra-operative success is more than just operating. In Oxford, great attention is paid to preoperative positioning, limb preparation, appropriate prophylactic antibiotic infusion and analgesia. The Oxford leg holder ( Figure 4 ) is a specifically designed device to enable optimal position of the leg to perform an Oxford UKA.