Total knee arthroplasty (TKA) is an effective, safe and increasingly popular procedure for the management of severe degenerative conditions of the knee. Technological and therapeutic advances have improved the durability of implants enabling manufacturers to focus on improving performance as, despite good pain relief and satisfaction rates, the majority of TKA recipients function at levels lower than age and sex matched normal populations. Traditionally, one of the key indicators of success of TKA is knee joint range of motion (ROM) as it is essential for many routine daily activities. Knee flexion is required for tasks such as stair climbing and rising from chairs, and close to full knee extension is necessary during gait. Post-operatively however, knee flexion seldom exceeds 110 degrees, and up to five degrees limitation in knee extension is common and considered normal. This is clearly inferior to the ROM of the native knee joint and is commonly assumed to be a major contributor to the suboptimal functional outcomes observed in TKA recipients.
Many studies have therefore focussed on improving knee joint ROM in order to improve functional outcome. Surgical innovations and prosthetic designs have recently been developed which manufacturers claim can accommodate knee flexion up to 150 degrees. The efficacy of these designs and their effect on function however, has not been established. In contrast, limitation in knee extension or flexion contracture (FC), has received relatively little attention. Like knee flexion, many studies have evaluated aspects of ROM restriction including exploring the aetiology and predictors of post-operative restriction and management techniques. Few studies have investigated the effect of FC on gait or function, the main focus has been on kinetic and kinematic abnormalities. No studies have previously evaluated the effect the increased muscular activation caused by FC has on energy expenditure or function. It is conceivable that the resultant increased exertion with walking could increase fatigue and present a more immediate impact on the elderly TKA recipient. Additionally it may create a cardiovascular risk in a population which is known to have a high prevalence of comorbidity.
The aim of this thesis was to investigate the impact of knee ROM at the extremes of flexion and extension after TKA. In order to do this several studies were undertaken. First, a systematic review of the orthopaedic literature comparing high-flex with conventional TKA designs was undertaken. This sought to determine whether high-flex TKA provided a statistically significant improvement in knee flexion, and if it did, ascertain whether this provided a meaningful functional benefit. Second, a study aiming to verify the validity and reliability of digital photographic goniometry (DPG) which could be used as a simple and accurate tool for measuring knee flexion and extension ROM following TKA was conducted. The third study of the thesis was a randomised controlled trial comparing knee flexion ROM and other clinical outcomes of a high-flex surgical innovation; flexion of the femoral implant in a cruciate-retaining design, with the conventionally positioned implant as the control. Finally, an investigation of the effect of FC on the energy cost of walking (Cw) after TKA was undertaken. To provide information which is clinically meaningful, a sub-aim within this study was to identify the FC angle at which significant changes in Cw, if present, occurred.
The systematic review revealed a lack of consensus in the literature regarding the impact of high-flex TKA designs on physical and functional outcomes. Five of the nine studies reported significant increases in knee flexion and or overall knee ROM. However the majority of high-flex studies had significant methodological flaws including no or poor randomisation, biases in group allocation, inadequate blinding and insufficient detail surrounding measurement techniques. Additionally, the functional outcome measures used were inadequate to thoroughly evaluate the effect of increased flexion on function. Conclusions were therefore unable to be drawn on the efficacy of high-flex TKA designs.
The study evaluating DPG found excellent criterion-related validity and both inter and intra-tester reliability for both knee flexion and extension. It was concluded DPG could be used instead of, or interchangeably with a universal goniometer, and offers several advantages including long term digital storage enabling review and re-measurement. This has potential application to both clinical follow-up and increasing the rigour of studies assessing ROM interventions.
The randomised controlled trial was designed to address many of the limitations identified in the systematic review. ROM measurement was blinded and validated and functional testing was expanded including both patient reported and physical testing components. Health related quality of life surveys and multi-domain satisfaction assessment were also included. A significant difference in knee flexion ROM was found between the flexed and control groups and this difference was particularly apparent at the measurement taken immediately following surgery. Both groups lost knee flexion over the following year reflecting the multi-factorial nature of post-operative knee flexion. There were no differences however, in any functional outcome. Moreover, knee flexion did not appear to be related to functional outcome, as the mean flexion of the control group was substantially reduced post-operatively, yet their functional outcomes were similar to the flexed group.
The gait study found the energy cost of walking increased with greater magnitudes of knee FC, but this effect did not reach significance until the contracture exceeded 20 degrees in the TKA group, a greater restriction than that seen in the control group (15 degrees). This finding may be reassuring to surgeons with regard to the exertion with walking for their patients, as FC’s of this magnitude are uncommon. Despite the TKA group comprising only high functioning participants with no pre-existing FC however, their gait was found to be abnormal. They walked significantly slower and with significantly less knee extension than normal control participants. Furthermore, in contrast to the normal group, the TKA recipients did not perceive the imposition of walking with a FC as more difficult than without a FC, suggesting they may be accustomed to walking with this uneconomical flexed pattern. Walking with a flexed pattern has been shown to increase the risk of anterior knee pain and FC’s of 15 degrees have been shown to cause abnormal stresses on implants.
The studies from this thesis have investigated knee joint ROM following TKA and have focussed on its effect on function. A great deal of research and development on high-flex designs and techniques has occurred in the last decade but there appears to be no discernible benefit from these innovations when compared to conventional designs and methods. Our study evaluating the surgical technique of a flexed femoral implant demonstrated a significant difference in knee flexion in the flexed implant group in comparison to the controls but this did not provide a meaningful clinical or functional benefit. These studies suggest the theoretical relationship between knee flexion and functional outcome is not as strong as is generally believed as it is heavily influenced by many other confounding factors. Further exploration of the role of weight bearing knee flexion and quadriceps muscle strength is recommended. Likewise, limitations in knee extension did not adversely affect the function of TKA recipients, specifically walking in the study in this thesis. Flexion contracture did not increase Cw until severe, clinically uncommon FC’s were simulated. This should not diminish concerns regarding the effects of FC on other important biomechanical factors which could affect function and implant durability. This thesis demonstrated that whilst knee ROM remains a key outcome following TKA achieving ROM approaching that of a native knee may be less important than was previously assumed.