The current design of the mobile-bearing Oxford, marketed as the Oxford Twin Peg Partial Knee, was initially introduced in 2003 in the United Kingdom and later implemented worldwide. An explanted device is shown. Improvements in the femoral design, effectively a “Phase 4,” consist of a more rotund undersurface to match the femoral bone cut, a twin-peg design to improve fixation and stability, and an additional 15° of femoral articular surface to increase contact in deep flexion
The first mobile-bearing UKA design was cleared by the US Food and Drug Administration for use in the United States in April 2004 with a physician training requirement prior to use. In June 2004, the FDA Orthopaedic Advisory Panel recommended the reclassification of mobile-bearing knee systems for general use; however, the FDA has not currently accepted this recommendation. It follows that less than 8% of all knee arthroplasties in the United States are unicompartmental procedures .
The mobile-bearing implant has also been investigated for use in the lateral compartment; however, the revision rate has been projected as high as 15% at 5 years postoperatively . The primary reason for this high failure rate is a high propensity for dislocation of the mobile bearing. This is believed to be the result of unique kinematics in the lateral compartment as the posterior femoral condyle translates a greater amount posteriorly in flexion compared to the medial femoral condyle. However, adjustments to the mobile-bearing design to maintain a full congruous spherical femoral articulation with a biconcave tibial plateau reduced the dislocation rate to 1.7% at 4 years in a new cohort of 101 patients . Nevertheless, due to this kinematic difference in the lateral compartment, a fixed-bearing implant has become the preferred type of lateral UKA implant .
The Oxford mobile-bearing UKA consists of a dual articulation between polyethylene insert and metallic femoral and tibial components. The conformity between the spherical femoral component and concave polyethylene bearing surface has been a design feature of the Oxford Partial Knee System present since its inception. Finite-element analysis predicts reduced contact stress due to an articular conforming design that distributes forces over a larger surface area. Ten-year in vivo measurements have demonstrated linear wear rates of 0.02 mm/year . Therefore, revisions for wear in long-term studies remain uncommon . However, impingement on retained osteophytes or cement particles remains a cause of not only bearing dislocation but also polyethylene wear . Kendrick et al. studied the impact of impingement on the polyethylene bearing as it related to wear rate. In a retrieval study of 47 Phase 1 and 2 bearings, the wear rate was 0.07 mm/year in the 31 bearings that demonstrated signs of impingement compared to 0.01 mm/year in those that did not . Moreover, the bearings demonstrating impingement affecting the articular surface had a penetration rate 2.5 times higher than those demonstrating extra-articular impingement. In general, the mobile-bearing design is believed to lessen the rate of polyethylene wear compared to fixed bearing designs in exchange for the risk of bearing dislocation .
This implant conformity and mobile-bearing design are also meant to decrease stress at the bone-cement interface. This has resulted in a small probability of aseptic component loosening, which has been estimated as low as 0.2% in some independent studies . This is believed to be an even smaller issue with the use of cementless components, which have been used in European countries since its release in 2004. In a randomized compared trial, fixation of cementless components was observed to be improved compared to cemented components as per lower amount of radiolucent lines . Currently, cementless components are under investigation in the United States, but are not currently approved for general use at the time of this writing.
The Microplasty instrumentation platform was subsequently introduced, designed to streamline the surgical procedure, and make it more efficient overall. The specific changes included a sizing spoon-stylus combination to decrease the need to recut the tibial plateau, an intramedullary femoral alignment guide, and a guide for reducing impingement. The spoon-based stylus references the posterior femoral condyle and removes 6.5 (3 “G-clamp”) to 7.5 mm (4 “G-clamp”) of tibial bone. The accuracy afforded by the spoons decreases the need for another resection of the tibial plateau as well as increases the likelihood of implanting smaller bearings (3 and 4 mm bearings). Femoral alignment in the Microplasty platform is performed via an intramedullary rod, whereas the Phase 3 instrumentation required visualization and adjustment of 6 separate variables. Removal of impinging osteophytes with Phase 3 instruments involved using an osteotome and then repeatedly checking for impingement in full-knee extension. The Microplasty guide for removing anterior osteophytes allows this step to be done once with no need to recheck impingement-free ROM.
We previously reported that the use of the new Microplasty instrumentation results in more accurate and reproducible femoral component placement . In another prior study, we analyzed whether the new Microplasty instrumentation improved efficiency and reduced operative time compared to the Phase 3 instrumentation . Patients in both groups were matched for gender, age, body mass index, preoperative ROM, and Knee Society pain and clinical scores. Operative time was defined as the time from skin incision until the final dressing was applied. Both groups were compared, and statistical significance was defined as p < 0.05. The mean operative time was significantly shorter with the Microplasty instrumentation (49 minutes) compared to the Phase 3 (58 minutes). Additionally, the standard deviation was significantly lower in the Microplasty group (14 minutes) versus the Phase 3 (17 minutes). The minimum and maximum operative times were also less in the Microplasty group compared with the Phase 3 (24–88 minutes versus 30–126 minutes).
The efficiencies of the Microplasty instrumentation resulted in an average of 9 minutes less per surgical case compared to Phase 3 instrumentation. This correlates to a 15% reduction in the time it takes to implant the Oxford mobile-bearing UKA. This 15% reduction in operating time should translate into the ability to perform more surgeries, decreased infection, decreased tourniquet use, and overall better experience for surgeon and patient alike.
Beginning in 1989, the classic article by Kozinn and Scott detailed contraindications to unicondylar arthroplasty procedures including both disease- and patient-specific criteria . They stated that patients exceeding an age of 60 years, weight of 180 pounds, or those extremely physically active heavy laborers were contraindicated for the procedure given an increased risk for mechanical loosening based on their anecdotal evidence. Disease-specific criteria, which included chondrocalcinosis on preoperative imaging or at the time of surgery and exposed subchondral bone within the patellofemoral joint, were identified as factors portending worse outcome. These principles stemmed from an unpublished study of 100 consecutive unicompartmental arthroplasty procedures performed by the authors with 10-year follow-up in which 13 failures from mechanical loosening were attributed to either surgical inadequacy, or patient-specific or disease-specific factors as categorized above. Other authors have echoed this sentiment when indicating patients for the procedure .
These historical patient indications severely restrict the number of patients considered as appropriate candidates for unicompartmental arthroplasty. One retrospective study of TKA cases declared that 21% of the cases may have been eligible for UKA based on disease-specific criteria, which included intact lateral cartilage, an intact ACL, no patellofemoral arthritis, ROM greater than 90°, and varus deformity less than 10° . Multiple investigations have aimed to refine appropriate indications for a unicompartmental arthroplasty with a mobile-bearing implant. In a prospective cohort of 1000 Oxford partial knee arthroplasties, Pandit et al. showed that the Oxford Phase 3 implant revision rate at 10-years was relatively similar for patients with one contraindication based on the Kozinn and Scott criteria as compared to those satisfying all criteria (2.4% vs. 4.0%) . The projected survival free of component revision from life-table analysis was higher in the contraindicated patients as compared to the ideal patients (97.0% vs. 93.6%). The causes of failure were different between these two groups as those ideal patients developed a higher rate of lateral compartment osteoarthritic progression, while the contraindicated patients suffered more mobile-bearing dislocations requiring revision surgery. This cohort of patients was comprised of 68% for whom the Kozinn and Scott principles would have contraindicated them for unicompartmental knee arthroplasty. In the updated study of this cohort, cumulative 15-year survival rate was not statistically different between those highly active male patients older than 60 years with weight greater than 180 pounds as compared with those patients without any of these contraindications (92.7% vs. 89.9%) . Furthermore, clinical outcomes as measured by Knee Society objective score , Oxford Knee score , and Tegner activity scale were similar or better in the Kozinn and Scott contraindicated patients. Further studies have demonstrated that age and activity do not compromise results of mobile-bearing unicompartmental arthroplasty, and these patients may be able to successfully attain a high level of activity postoperatively [31–33].
In contrast, Goodfellow and colleagues believed that treatment with a mobile-bearing unicompartmental arthroplasty should instead be applied in patients demonstrating the appropriate pathoanatomy independent of patient-specific factors. The specific applications included anteromedial osteoarthritis and spontaneous medial osteonecrosis of the knee. Anteromedial osteoarthritis (AMOA) is defined by medial compartment bone-on-bone joint space narrowing with intact posterior cartilage. In addition, the lateral compartment should contain full-thickness cartilage and both the anterior cruciate and medial collateral ligaments should be functional. This pathoanatomy manifests specific clinical signs and symptoms. Varus deformity is most noted in full extension due to the pattern of wear on the anterior portion of the tibial plateau and the inferior articular surface of the femoral condyle . This deformity is not fixed and can be corrected with a valgus stress at roughly 20° of flexion relaxing the posterior capsule. This is possible because joint space contact in flexion retains normal cartilage, therefore maintaining normal tension on the medial collateral ligament (MCL) and keeping its length constant. It is believed that the presence of functional cruciate ligaments correlates with the disease pattern observed as they maintain normal femoral roll-back in flexion.
Hence, unicompartmental arthroplasty should not be offered in cases with an impaired ACL. In some cases, the ACL may fail secondarily after the advent of anteromedial disease, causing a progressive erosion of the posterior cartilage and therefore a fixed varus deformity. In other instances where the medial compartment disease develops secondary to ACL rupture, the posterior cartilage is usually affected first due to anterior subluxation of the tibia. Still attempts have been made to reconstruct the torn ACL while performing unicompartmental arthroplasty with promising short-term results .
There has also been some confusion when it comes to application of unicompartmental arthroplasty in patients with anteromedial osteoarthritis who demonstrate arthritic changes in the patellofemoral joint. In a cohort of 677 patients, the Oxford group found that there was no relationship between implant survival at 15 years postoperatively and the presence of anterior knee pain preoperatively, nor with the degree of cartilage loss within the patellofemoral joint intraoperatively . The authors did document difficulty with stair descent in those patients treated with medial mobile-bearing unicompartmental arthroplasty demonstrating intraoperative evidence of full-thickness cartilage loss on the lateral aspect of the patella. Similarly, in a retrospective review of 100 consecutive Oxford medial unicompartmental arthroplasties with a minimum 8-year follow-up, patients with grade 3 change in the central and lateral aspect of the patellofemoral joint were found to have lower mean satisfaction with pain and function compared to the remainder of the cohort . Stair climbing ability was also significantly decreased in those patients with central and lateral lesions observed intraoperatively in the patellofemoral joint. For this reason, severe damage to the lateral side of the patellofemoral joint with bone loss and grooving is defined as a contraindication to the procedure; however, less severe damage to the lateral articulation, medial patellofemoral disease, and anterior knee pain should not be considered contraindications.
Rheumatoid arthritis is another contraindication to medial unicompartmental arthroplasty as the inflammatory process primarily affects the synovium, resulting in tricompartmental disease. Therefore, in patients with this underlying diagnosis, it is recommended that total knee arthroplasty be performed as there is a risk of rheumatoid progression when a unicompartmental arthroplasty is performed .
Based on these principles, a strict preoperative clinical evaluation should be implemented in order to determine the ideal candidate for medial mobile-bearing UKA . Clinically the patient should have varus malalignment in extension that corrects in flexion. Flexion contracture should not exceed approximately 15° and total range of motion should be greater than 100°. The ACL should be competent on clinical exam. Imaging should demonstrate significant loss of medial compartment joint space in either the anteroposterior weight-bearing view or the posteroanterior 45-degree flexion view. Lateral radiographs should display bony erosion of the anterior portion of the medial tibial plateau in contrast to an ACL-deficient knee in which the femoral condyle will be articulating with the posterior portion of the plateau, causing posterior erosion. A valgus stress view taken at 20-degrees of knee flexion should also be taken to confirm full-thickness cartilage within the lateral compartment and demonstrate a correctable deformity through a competent MCL. The patellofemoral joint should be imaged in order to exclude patients with significant bone-on-bone arthritis of the lateral patellar facet. Otherwise, moderate lateral facet disease or advanced diseased of the medial patellofemoral compartment should not preclude the use of UKA.
These criteria were elucidated in a radiographic decision aid, which was developed by a collaboration of joint arthroplasty surgeons after review of current literature . In a retrospective review of over 500 patients, those meeting the radiographic standards irrespective of patient factors such as age and weight displayed a 5-year implant survival rate of 99% compared to 93% in those patients failing to meet these standards. Furthermore, functional outcomes measured by knee flexion, Knee Society score function component, and University of California Los Angeles activity score were significantly higher in those patients meeting the radiological criteria.
Spontaneous osteonecrosis of the knee (SPONK) that is focal and localized to the medial femoral condyle or the medial tibial plateau is also an indication for mobile-bearing UKA . In the early stages of disease, SPONK may only be detected on MRI prior to subchondral collapse, while also ruling out secondary osteonecrosis, which frequently involves both condyles . As the disease progresses, some patients may demonstrate subchondral collapse in conjunction with joint space narrowing as the osteonecrosis is accompanied by a degenerative process. In all forms of the disease, the pathoanatomy resembles anteromedial osteoarthritis in that it is limited to the medial compartment and both the ACL and MCL are functionally intact. This should not be confused with secondary osteonecrosis, which occurs frequently in younger patients after corticosteroid, renal, or systemic disease . This entity is often bilateral and involves both compartments, thus making unicompartmental arthroplasty futile. More recently, osteonecrosis in the postoperative knee (ONPK) has been described following arthroscopic surgery and is similarly focal in extent and localized to the medial femoral condyle in most cases [44, 45]. Outcomes and survival of UKA for SPONK or ONPK localized to the medial compartment have been encouraging [46, 47]. Furthermore, the success appears to be independent of the size of the osteonecrosis lesion as we have found a survival rate of 94.6% at 5 years in a cohort of 64 patients with mean lesion width amassing 64% of the medial femoral condyle width. Only one patient suffered from aseptic loosening of the femoral component in this cohort .