10.7 Osteoarthritis

Osteoarthritis (OA) is the commonest joint disorder in the world and represents a heterogeneous disease process with the final outcome being joint failure. It can be defined pathologically, radiographically, or clinically. While in the past radiographic classification grades have been used to define OA, more recently OA is seen as a failed repair of damage that has been caused by excessive mechanical stress (defined as force/unit area) on joint tissues and combines both structural and symptomatic features to describe the failed joint. There are many causes of OA and the disease process affects not only the articular cartilage but also the subchondral bone, ligaments, joint capsule, synovial membrane, and periarticular muscles. Ultimately there is articular cartilage degeneration with fibrillation, fissures, ulceration, and full thickness loss at the joint surface with sclerosis and eburnation of the subchondral bone, osteophytes, and subchondral cysts. Clinical features are characterized by joint pain, tenderness, limitation of movement, crepitus, and occasional effusion without systemic effects. The failure of the joint to repair leads to pain and dysfunction. However, many people with severe radiographic changes are asymptomatic and progression, when it occurs, is usually slow and may halt intermittently.

OA is the most common joint disorder in the world. It is more common in women and the prevalence increases rapidly with age, with 10% of the world’s population aged over 60 years having symptoms attributed to OA. In 1990, OA was the tenth leading cause of non-fatal health burden in the world. In 2002, OA was the fourth leading cause of years lived with disability in the world. The global prevalence rates are shown in Figure 10.7.1. Using radiographic definitions, knee OA is present in 37% in those aged over 60 years and hip OA 27%. Symptomatic OA is much less common with rates of 4.9%. The lifetime risk of developing symptomatic knee OA is 45% (Figure 10.7.2).

OA is a dynamic process with a number of antecedents and represents a final common pathway for all insults to the joint. The typical initiating factor is a mechanical insult and the response of OA is a manifestation of the joint to repair and minimize the abnormal biomechanics. From community studies, at the knee OA may affect medial/lateral tibio-femoral joint (40%) and/or patellofemoral joint (60%).

The function of the chondrocytes in healthy cartilage is to maintain the matrix in terms of collagen, aggrecan, matrix metalloproteinases (MMPs), and cytokine constituents in response to loading. Risk factors for OA development divide into abnormal loading on normal cartilage or normal loading on abnormal cartilage. In OA, the abnormal loading experienced by the chondrocyte leads to promotion of matrix degradation and downregulating cartilage repair processes (Figure 10.7.3). The constituents of the articular cartilage change in OA with increased type II, IX, and XI collagen and aggrecan representing a matrix constitution similar to that seen in fetal development. The type of loading is critical, while repetitive loading promotes homeostasis of the articular cartilage, a single pathological compression of cartilage leads to an increase in catabolic MMP and aggrecanases in an attempt to resorb the damaged cartilage. Histologically, damage to the articular cartilage is evidenced by fibrillation, horizontal splits, partial and full thickness loss. Vascular invasion of the calcified cartilage leading to chondrocyte hypertrophy and replacement by bone further thins the articular cartilage.

There is a growing body of evidence demonstrating the role of bone as a key tissue involved in the pathogenesis of both structural and symptomatic features of OA. The principal role of the subchondral plate is mechanical with the cortical and trabecular bone compartments acting as shock absorbers, continually responding to loads applied to them by viscoelastic dampening. In the longer term, stress on the subchondral bone lead to remodelling, via the osteocyte network and Wnt signalling and alteration of the material properties. In OA, these changes in material properties reduce its capacity to absorb and dissipate the energy from loads through the joint. Perturbations in the trabecular structure result in increased load associated in interosseous pressure in the medullary area and this may be represented by bone marrow lesions as seen on magnetic resonance imaging (MRI), which have also been associated with arthritic pain and change in bone marrow lesions may correlate with change in symptoms.

In addition to structural changes, the subchondral plate of bone and underlying trabecular/cortical structures are innervated with nociceptive sensory nerves and there is now increasing evidence identifying bone as a source of pain in OA. From bone scintigraphy, extended bone uptake was strongly associated with knee pain in patients with knee OA (odds ratio 3.71 (2.26–6.12)). Of the neuropeptides, both calcitonin-related peptide and substance P are expressed in bone with the potential of mediating bone pain. These neuropeptides can function as both afferent nociception via substance P as well as efferent influences on bone remodelling via CGRP (calcitonin gene-related peptide), an inhibitor of osteoclast motility and stimulator of osteoblast proliferation.

Osteophytes are fibrocartilaginous outgrowths localized to the joint margins that form through a process of endochondral ossification and are the radiographic hallmark of OA (Figure 10.7.4). Recent work has highlighted the role of the Wnt signalling system as critical in their formation. The significance of osteophytes as a marker of skeletal adaption versus a pathological feature of maladaptation remains unresolved.

While OA has been traditionally been regarded as a degenerative disease, it is now clear that local synovitis is very common in OA—these inflammatory changes are likely to be the consequence of cartilage and bone catabolism and are associated with severity of knee pain. The ability to visualize the joint non-invasively using both contrast and non-contrast imaging has demonstrated the almost universal presence of synovitis in OA. From histological specimens, the synovial membrane in OA is hyperplastic and may contain lymphocytic infiltrates. The synovium not only is a source of catabolic inflammatory enzymes but also is richly innervated and represents another source of pain.

In the knee, the semicircular fibrocartilaginous menisci cover two-thirds of the articular surface and are responsible for load distribution and shock absorbing in addition to roles in proprioception, joint stability, and joint lubrication. Damage to the menisci can be the result of trauma or degeneration. Traumatic injury to the menisci is manifest as vertical or radial tears. Degeneration is associated with flaps, complex tears, or maceration. Damage to the menisci also commonly involves extrusion of the mensci. Meniscal damage is extremely common in OA as well as in those without joint pain and the contribution of meniscal damage as a cause versus consequence of OA is not yet clear.

Muscles not only provide movement but also provide stability for the joint and absorb reaction forces to loading. During normal walking a load of three to four times body weight is generated and at the knee, periarticular muscles are activated in specific patterns to actively absorb this energy and protect the other tissues of the joint. There is growing evidence for the role of the quadriceps weakness not only as a consequence of symptomatic knee OA but also as a determinant, through the potential role of muscle to absorb energy, facilitating patella tracking and acting as a sensory input for joint proprioception, together with mechanoreceptors in the ligaments, capsule, menisci, and skin. Damage to the cruciate ligaments is recognized as a key antecedent for knee OA, although the onset for symptomatic disease may be delayed for many years.

OA is a common complication of obesity. Obesity through increased mechanical loading is strongly associated with incident and progression of OA especially at the knee and is a risk factor for total joint replacement. In addition, obesity is associated with a loss of muscle strength relative to weight and poor balance. To compensate for this, obese people have an altered gait exerting greater forces during walking. In addition to mechanical factors, obesity is associated with low-level inflammation which may further contribute to joint damage.

Despite the mechanical aetiology to OA, the significant heritability of OA has been confirmed using familial studies, twin studies, linkage analysis, candidate gene, and genome-wide association studies. Genetic influences on bone density and geometry, obesity, synovitis, and chondrocyte function all potentially influence the natural history of OA. Polymorphisms in the Wnt signalling pathway as well as metalloproteinases and cartilage constituents, such as cartilage intermediate layer protein and oligemeric matrix protein, have been associated with knee OA. While the individual genetic traits carry a modest increase in OA risk, identifying the combination of traits in the form of gene–gene and gene–environment interactions have the potential to both inform our understanding of the pathogenesis of OA as well as identify high-risk groups and potential therapeutic targets.

In addition to local nociceptive stimulus at the joint leading to pain in OA, there is a growing body of evidence that a mechanism involving the peripheral nerves through increased innervation density, spinal cord via gating and hyperexcitability, and altered cortical processing such as locus of control, anxiety, and depression leading to altered inhibitory descending tracts have major affects on symptoms and response to therapy. The characteristic feature of central sensitization leading to chronic pain is that previously non-noxious stimuli such as walking are perceived as painful.

The purpose of clinical risk factors is to assess risk of disease, aid diagnosis and identify possible therapeutic methods. They can be divided into modifiable and non-modifiable risk factors (Table 10.7.1).

Of the risk factors, some are associated with incident disease (high bone mineral density), others with progression (malalignment) and some with both (obesity).

OA presents as a gradual onset, episodic joint pain worsened by activity and relieved by rest. As pain advances, range of movement and function declines and there is a reduction in joint stability with buckling/giving way symptoms (Box 10.7.1). Finally, OA leads to pain at rest and night which through loss of sleep further exacerbates the pain. Physical examination should include assessment for obesity, joint range of motion, crepitus, bony enlargement, focal tenderness, signs of local inflammation, muscle strength, and joint alignment. It is important to exclude referred pain, e.g. hip OA presenting as knee pain.

Imaging is used to confirm clinical suspicion and rule out other diagnosis. While radiographs may have features of OA, the absence of radiographic features does not exclude a diagnosis of OA; in addition the presence of features does not exclude other causes for the patient’s symptoms. MRI is used to exclude other diagnosis where appropriate. Blood tests can not confirm OA and a full blood count, and renal and liver function tests should be performed if there is a plan to start non-steroidal anti-inflammatory drugs (NSAIDs), especially in the elderly. While biomarkers have been developed for OA, their use is limited to research studies at present.

The aim of imaging is to diagnose, follow-up, and elucidate underlying mechanisms of disease in OA. Standard imaging assessment of a joint affected with OA is the plain projectional radiograph which gives a two-dimensional representation of the bone (osteophytes, subchondral sclerosis, subchondral cysts, and attrition) and inferences regarding soft tissue such as joint space narrowing, which represent both cartilage thickness and meniscal extrusion. The Kellgren and Lawrence score is maintained as one of the key assessment of disease severity (Table 10.7.2). More recent atlases (OARSI) classify each compartment separately. The standard views are semiflexed anteroposterior (AP) to assess joint space width and tibiofemoral bone changes, and the skyline or lateral to examine the patellofemoral joint. In contrast, MRI offers a three-dimensional view in any given plane of all the anatomical structures within the joint including cartilage (as measured by defects or volume), osteophytes, ligaments, tendons, synovium, capsule, bone contours, and marrow signal. MRI can detect lesions at an earlier stage in more subregions of the knee. In addition to structural features, the use of specific MRI sequences such as T₂ mapping and T₁rho, allow indirect assessment of the cartilage composition

Ultrasound can easily detect cartilage thickness, meniscal integrity, synovitis, and effusion. Scintigraphy use is limited due to the radiation exposure and poor specificity. Newer computed tomography (CT) scanners are able to offer detailed imaging of the bone structures but have lower soft tissue contrast than that of MRI, for example. CT arthrography provides an indirect method to visualize the cartilage and structures within the knee.

The treatment aims of knee and hip are to reduce clinical symptoms of pain and stiffness, improve or maintain function, reduce disability, improve quality of life and limit disease progression, both symptomatic and functional. A key aim is patient education not only in terms of OA causes and consequences but also in terms of expectations regarding treatment, a major predictor for willingness to accept arthroplasty. The recent OARSI guidance, by consensus produced a list of recommendations for the treatment of hip and knee OA (Table 10.7.3). From a mechanical perspective, therapies that reduce the abnormal joint loading are likely to result in better patient outcomes rather than tissue directed therapies.

Patient education should address the contributions of biological, psychological, and social factors in OA pain. Hence the treatment should be individualized and education should be tailored to the patient’s needs and capabilities. Further, in terms of willingness to accept more invasive interventions such as arthroplasty, information regarding postoperative pain and its management, expected time to recovery, and duration of benefit are all likely to influence patient expectations and reported satisfaction after surgery. While difficult to achieve, weight loss is a key management strategy for obese patients presenting with OA. A modest weight reduction of 10% is associated with clinically significant benefits. Strategies that combine incentives, raising outcome expectations and building on self-efficacy, are more likely to succeed. Further, combining weight loss with exercise leads to a preservation of muscle strength.

Exercise is an integral part of the management of the patient with OA, especially of the knee. However, failure of adherence to exercise regimens and underuse by practitioners to recommend the correct exercise or refer to physiotherapy remain key hurdles. At the knee, a key aim is to improve muscle strength. There is observational evidence that greater quadriceps strength increases disease progression in the presence of malalignment but not in those with neutral knees (±5 degrees); there is no robust interventional evidence for this. Through appropriately targeted exercise, proprioception can also be improved. An exercise prescription is now recommended for muscle rehabilitation for patients with OA (Box 10.7.2).

The principal aim of a walking device is to reduce the compressive load across the affected joint surface. This is maximally achieved by using the walking device in the contralateral side which not only shifts body weight from the symptomatic limb but also lessens the adduction moment. The optimal height of the cane is to the superior tip of the greater trochanter, with the arm at 20–30 degrees flexion.

A properly fitted unloader brace of the knee attempts to correct the relative alignment of tibia to femur and improve mediolateral instability. Care needs to be taken to ensure adequate adherence to using the brace and these braces are not appropriate in those with bicompartment disease. In those with symptomatic patellafemoral OA, medial taping of the patella reduces symptoms by increasing the patellafemoral contact area during weight-bearing exercise that involves knee flexion rather than maintaining the alignment of the patella.

At the initiation of the gait, the heel strike transient represents the ground reaction force at the point of initial ground contact and is attenuated by many factors including the heel fat pad. The use of viscoelastic insoles lead to a more gentle heel strike moment. The use of lateral or medial wedged insoles to counter degrees of malalignment has not been shown to produce meaningful symptomatic benefit.

The role of analgesics has been questioned by the observation that patients treated with a NSAID had reduced joint pain but an increase in the medial joint loading, with the potential of accelerating progression. NSAIDs and COXII inhibitors are also associated with significant side effects and careful use is advocated by national guidance. The use of topical agents represents an alternative route of administration for NSAIDs and they are better tolerated with fewer side effects and good short-term pain relief. Other topical agents such as capsacin work by initiating an intense local irritation that then depletes substance P. These agents need to be applied three to four times a day to maintain efficacy.

A number of intra-articular agents have been used in OA. Corticosteroids have a good short- to medium-term effect in reducing pain and improving function. The current guidance is to limit injections to four a year. Hyaluronic acid derivatives have been used to inject large joints and while they improve pain and function do not appear to have a disease-modifying role.

Ongoing trials are evaluating the benefit of vitamin D supplementation. Vitamin K use has not been shown to be beneficial in the general population. Other micronutrients such as selenium, carotene, and vitamin E and C have not been clearly proven to improve clinical outcomes.

A large number of complementary therapies have been used in the treatment of OA pain including glucosamine, chondroitin, avocado-soybean; however none have been robustly shown to improve clinical outcomes.

Although OA is the commonest arthritis globally, we still have little understanding of why patients develop the disease, why symptoms occur in some not others, and how to predict those that will develop severe symptomatic disease. This limitation of understanding has restricted the development of clinically effective interventions to date. In contrast, there is an historical low level of utilization of non-pharmacological interventions which have been shown to improve symptoms. There is an urgent need to combine universal implementation of recent guidance together with translation research to improve patient outcomes in OA.