5.4 Rheumatoid arthritis of the elbow

Rheumatoid arthritis (RA) is the most common form of inflammatory arthritis and affects 3% of women and 1% of men. The classification criteria developed by the American Rheumatism Associated require four of the following:

The elbow is involved in 20–50% of patients with rheumatoid disease and, in the majority of cases, it is bilateral. Women are affected three times more commonly than men. The elbow is the first joint to be affected by RA in only 2.1–3% of cases. In the early stages, the synovitis causes pain and tenderness, especially over the radiohumeral joint line, with associated loss of elbow extension. Later, the whole elbow may become swollen and stiff. Finally, when bone destruction is severe, instability and capsular rupture result in a flail elbow (see Figure 5.4.1). Ulnar collateral ligament incompetence may cause valgus ulnar humeral instability and ulnar nerve dysfunction. Annular ligament incompetence can lead to radial head subluxation.

The aetiology in RA is unknown. The genetic predisposition, the involvement of activated immune cells, and the response to immunosuppressing therapy, all suggest the disease is immune mediated. The cell-mediated immune response (T cells that incite an inflammatory response) is initially against soft tissue (synovitis) and later, against cartilage (chondrolysis), and then bone (periarticular bone resorption). The pathological spectrum of RA spans across early disease when joints exhibit active synovitis without structural damage, to late disease when the joints may be mechanically damaged, mal-aligned, and unstable without persistent active synovitis.

RA is a rare disease in men under the age of 30 years. The incidence rises to peak at 60–70 years. In women, the prevalence of disease increases from the mid twenties to a fairly constant level at 45–75 years with a broad peak at 65–75 years. The genetic contribution in RA is around 15–20% and this can exert a significant effect on disease expression. The risk of RA in first-degree relatives is almost double that amongst the general population.

The standard radiographic assessment of RA was proposed by Larsen and colleagues in 1977. He grades the disease into six stages (Figure 5.4.2).

An understanding of the function and biomechanics of the elbow is important when considering the design and features of total elbow replacement. The main role of the elbow joint is to position the hand in space but a full range of movement of the elbow is not required for many activities of daily living. It has been shown in normal volunteers that most tests can be carried out with a 100-degree arc of flexion between 30–130 degrees, and a 100-degree arc of forearm rotation equally divided between pronation and supination. In patients with elbow RA the flexion arc is often well preserved until a late stage in the disease.

The other main role of the elbow joint is to act as a fulcrum for the forearm lever, and this stability is essential for normal strength and function. Increasingly, severe erosion of the subchondral bone of the elbow joint, due to RA, gives rise to varus/valgus and anteroposterior instability and thus weakness, because of the loss of the elbow joint as a stable fulcrum.

Many RA patients are unable to carry out the more demanding activities of daily living because of pain and weakness caused by reflex inhibition of adjacent muscles. One group found the average flexion strength of the elbow in rheumatoid patients tested at 90 degrees, was approximately 50% of the normal young adult. The activities of daily living, such as eating and dressing, can apply compressive loads of up to half of the body weight to the elbow and may be as high as three times body weight with certain actions, such as pushing up out of a chair or lifting heavy weights.

When assessing the elbow in a patient with RA, it is important to perform a thorough evaluation of the cervical spine, shoulder, wrist, and hand in both upper limbs, to determine if the patient’s complaints are limited to the elbow or are more diffuse in nature. The age, occupation, hand dominance, and age of onset of RA should be assessed.

Isolated elbow involvement in RA is rare. It should be noted that associated wrist (distal radioulnar joint) pathology can contribute to loss of forearm rotation and limited shoulder abduction can prevent the ability to compensate for loss of pronation. Specific elbow problems that require assessment are pain, stiffness, swelling, instability, and ulnar nerve symptoms.

Pain arising in the joint is often diffuse in nature. It is commonly felt on the lateral side initially. To gauge the level of pain in the elbow an assessment of night and rest pain should be made together with the number of anti-inflammatory painkillers taken per day. In addition, an assessment of whether there is pain on certain specific activities of daily living is made, including washing, dressing, and lifting simple household objects, for example, the ability to lift a mug, kettle, or a pan.

Mild stiffness may hardly be noticed. If it is severe, it can be very disabling. The patient may be unable to reach to the mouth (loss of flexion), or to the perineum (loss of extension). Limited supination makes it difficult to carry large objects. It is very important to consider the function of the whole arm as a unit as well as the patient’s age, occupation, and overall condition prior to forming a treatment plan.

An in-depth medical history is taken of age at onset of the RA, current drug treatment relating to anti-inflammatory drugs, disease modifying drug therapies, and specifically whether the patient is taking prednisolone or anti-tumour necrosis factor (TNF) therapies. An assessment of the extra-articular complications of RA should be made, specifically relating to pulmonary and cardiac symptoms and the neurological effects of peripheral entrapment of the ulnar nerve. It is important to assess the effects of the RA on other joints, particularly in the lower limb, and whether the patient has had or is awaiting total hip, total knee, or forefoot arthroplasty surgery. If surgery is being contemplated a general medical history must be taken to assess fitness for anaesthesia (diabetes, ischaemic heart disease, use of warfarin, lung pathology).

Finally, a social history is useful in noting where the patient lives (flat, a bungalow, or two storey house), who the patient lives with, the need for social service support and an assessment of mobility, particularly the use of walking aids.

General considerations include a rapid assessment of the patient’s habitus and whether walking aids or a wheelchair are being used. The patient should be asked to undress to reveal the whole of the upper limb. Screening examination of the neck and shoulders should be made to assess range of movement and pain. Rheumatoid involvement of the neck or shoulder may make examination of the elbow difficult because of pain and stiffness.

With both upper limbs exposed, the patient holds their arms by the side of the body with palms forward; this enables varus or valgus deformity at the elbow to be assessed. The patient then holds the arm out sideways at right angles to the body with palms upwards and elbows straight. An assessment of muscle wasting, swelling, lumps (Figure 5.4.3), surgical scars, and skin colour can be made.

The back of the elbow joint is palpated for warmth, subcutaneous nodules, synovial thickening, and fluid. Laterally, the joint line can be palpated for swelling and crepitus. The radial head should be assessed for tenderness, and to make sure it is properly aligned and not subluxed or dislocated. The ulnar nerve can be palpated behind the medial epicondyle and assessed for sensitivity and Tinel’s test for cubital tunnel syndrome.

Flexion and extension are compared on the two sides. Pronation and supination can be measured and compared with the elbows tucked into the sides and flexed to a right angle.

Rheumatoid nodules (Figure 5.4.3) occur in 30% of patients with progressive seropositive disease. These are most commonly subcutaneous and occur on the extensor surface of the elbow and especially around the olecranon. They are usually mobile but can be fixed to deeper tissues. In addition, they can be found in the spine, occiput, and other areas exposed to mechanical pressure. Nodules have a characteristic histology in that they are made up of a central area of fibrinoid necrosis surrounded by an area of palisading epithelial cells and fibrocytes.

There is no single test to diagnose RA. Instead, the diagnosis is based upon many factors, including the characteristic signs and symptoms, the results of blood tests, and the results of radiographs. Laboratory studies are performed to help with the diagnosis, to monitor the progress and complications of the disease, and to assess the side effects of drugs. Blood tests include a full blood count to measure haemoglobin (anaemia is a common presenting feature of RA), white blood cell count and platelet count, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), rheumatoid factor, antinuclear antibodies, liver enzymes, and renal function.

Rheumatoid factor (immunoglobulin M) is detected by the Rose–Waaler assay and is positive in approximately 80% of RA patients. Rheumatoid factor may also be positive in a number of other conditions.

The ESR and CRP are non-specific markers of inflammation. A high ESR and CRP suggest the presence of inflammation, but they do not indicate the cause of this inflammation. The anticitrullinated peptide antibody test is more specific than rheumatoid factor for diagnosing RA. It may be positive very early in the course of the disease. The test is positive in 50–90% of patients with RA. Between 30–40% of people with RA have autoantibodies called antinuclear antibodies (ANAs). However, many healthy people also have a positive ANA test.

Plain anteroposterior and lateral radiographs are an important measure of disease progression (see Fig 5.4.1C). In established RA, radiographs are used to classify and monitor disease progression from osteoporosis, soft tissue swelling, periarticular cysts, joint space narrowing, alteration of joint architecture, and finally, gross joint destruction. About 15–30% of patients with RA will have radiographic changes in the first year. However, after the first 2 years of RA, more than 90% of people have changes on radiographs.

Magnetic resonance imaging (MRI) scans are more sensitive than radiographs for detecting the cartilage damage caused by RA. Therefore, MRI scans may be more effective than radiographs for detecting the early changes of RA.

In patients where the elbow joint is the first joint to present, aspiration of the joint for microscopy, Gram staining, crystals, and culture, may aid in the diagnosis by excluding other conditions that may cause a swollen elbow. These include infection, gout, pseudogout, and other connective tissue diseases.

The therapeutic goals are the control of synovitis and pain, maintenance of joint function, and the prevention of deformities. The management of RA requires a multidisciplinary team approach, usually headed by a rheumatologist and may involve the use of drugs, physical therapy, and sometimes surgery (Table 5.4.1).

Education by a clinical nurse specialist plays a vital role in helping patients understand their disease, and associated drug therapy. The management of RA has changed substantially over recent years with the emphasis now being on early diagnosis and aggressive early intervention with disease-modifying antirheumatic drug (DMARD) therapy. This approach has been shown to make a long-term difference to prognosis. Anti-TNF drugs are indicated in patients not responding to conventional DMARDs. The aim of initial drug therapy is to prevent disease deterioration as well as to alleviate pain.

A multidisciplinary team should be involved early and throughout the course of the disease. This involves specialist nurses who can take patients through the details of drug regimens, side effects, and monitoring requirements, and physiotherapists who can help with non-pharmacological pain relief, introduce appropriate exercise, and discuss the balance between activity and rest. Occupational therapists provide resting splints for painful joints, and aids which maintain function, independence, and employment.

The key role of all professionals in early RA is education. Evidence exists that educating patients with early RA is an independent predictor of good disease control.

Non-steroidal anti-inflammatory drugs (NSAIDs) represent first-line treatment in all types of arthritis, acting to decrease the synovial reactivity and alleviate pain and swelling. Disease-modifying drugs such as sulphasalazine, methotrexate, and azathioprine are used in mild, moderate, and severe disease. Some patients with mild disease may respond with symptom control to a single disease-modifying drug while those with poor prognosis disease may need early aggressive combination drug strategies (Box 5.4.4).

Methotrexate is now the most commonly used disease-modifying drug with a relatively high rate of response. Its risk of toxicity is reduced by the co-prescription of folic acid.

Unlike NSAIDs, the use of steroids slows the course of the disease. Over the short term, they exert a profound effect on the symptoms of RA but their long-term use is contraindicated due to their significant side effects.

TNF is a key cytokine in RA inflammation and damage. Over the past 5 years, drugs that block this cytokine (adalimumab, etanercept, and infliximab) have been introduced into clinical practice. Two of the drugs, adalimumab and etanercept are subcutaneous injections that can be self-administered; infliximab is given intravenously every 8 weeks after initial induction. These drugs have made a huge difference to the management of many patients with active disease who have failed to respond to conventional treatments. It is estimated that up to 5% of all RA patients may require such therapy. In the United Kingdom, anti-TNF therapies are restricted to patients with ongoing active disease who have failed to respond to at least two conventional disease-modifying drugs, where one is methotrexate.

It is important in early RA to monitor treatment response and modify therapy accordingly (Table 5.4.2). This requires a combination of subjective and objective assessment using a disease activity score, which is an amalgamation of the number of tender and swollen joints, the ESR, and a visual analogue score of the patient’s overall health.

Surgery is indicated when appropriate non-surgical management has failed, giving rise to functional limitations due to pain or loss of motion (Table 5.4.3). It is important when surgery is being contemplated to preoperatively assess the cervical spine with radiographs and to be certain that the patient will be compliant with postoperative rehabilitation.

The primary aim of surgery on the elbow is to relieve pain and/or restore joint function.

Synovectomy with radial head excision is an accepted procedure for the early stages of RA (Larsen grades 1 and 2), with chronic synovitis and pain. Following synovectomy, the synovium initially regenerates normally, but with time it will generate back into rheumatoid synovial tissue.

Excision of the radial head was recommended by most authorities in the past but today the radial head is preserved if it is not severely involved in the disease process or causing pain with pronation or supination. The negative effect of radial head resection is to increase the loading of the ulnar compartment of the joint. In addition, resection of the radial head results in deterioration of the postoperative outcome from 70% to 45% during a 6-year period.

Previous synovectomy does not affect the results of total elbow replacement and can therefore be considered in the early, painful stages of rheumatoid destruction of the elbow joint. A lateral approach to the elbow is used and if the radial head is excised there is excellent exposure of the anterior compartment for synovectomy.

Lee and Morrey have reported arthroscopic elbow synovectomy results similar to those achieved by open synovectomy.

Review of the literature suggests up to 90% excellent pain relief for the first 2–5 years after surgery, with late results demonstrating deterioration over time with the success rates falling to 60% at 5–10 years. The largest study reviewed 171 rheumatoid elbows with synovectomy and radial head excision with failure defined as the need for revision surgery or significant pain. The study reported a 1-year survival rate of 85% and a 6.5-year survival rate of 45%.

Ulnar nerve compression at the elbow may at times also be an indication for synovectomy of the elbow, with or without neurolysis and nerve transposition.

Although at one time popular in Europe, this procedure is now rarely carried out in patients with RA. It is contraindicated in patients with significant joint destruction and gross loss of bony architecture. The primary indication is a young, active patient without an inflammatory arthritis, who has a severe post-traumatic arthritis associated with severe pain and limited motion, and is too young or active for total joint arthroplasty. The disadvantage of resection interposition arthroplasty in patients with RA is the risk of progressive bone loss with accelerated instability. This loss of bone may, in the long run, be such that reoperation by prosthesis is rendered impossible. The method has slowly lost ground in the light of the improved results with prosthetic surgery.

Total elbow replacement is indicated for elbows with severe pain, limited range of motion, and pronounced loss of articular cartilage. The patient must agree to postoperative limitations of lifting no more than 2.5–5kg (5.5–11lb). The contraindications to total elbow arthroplasty are sepsis, the need for soft tissue coverage, severe muscle weakness or paralysis, and a non-compliant patient. There are two types that are commonly used: unlinked (Figure 5.4.4) and linked (Figure 5.4.5). In addition, more recently convertible implants have been designed which can be inserted as either unlinked or linked prostheses.

In an unlinked total elbow replacement, there is no linkage between the humeral and ulnar components, with a metal on polyethylene articulation. It relies on the ligaments and muscles for joint stability, with the theoretical advantage of decreased loading at the bone–cement interface. However, it carries a greater risk for instability. Significant pain relief has been reported in 79–94% with an 80%, 12-year survival for the Souter–Strathclyde implant with revision as the end point and a 90%, 16-year survival for the Kudo implant. An improved final arc of flexion has also been reported of 32 degrees to 136 degrees.

In a linked prosthesis, there is a sloppy hinge with a metal on polyethylene articulation that allows limited rotational varus/valgus motion between the humeral and ulnar components, to decrease the bone–cement interface loading. However, there is stress concentration at the hinge, which can lead to polyethylene wear, debris, osteolysis, and component loosening. The outcome of surgery in linked total elbow replacement is significant pain relief in 76–92%, with an improved final arc of flexion/extension of 29 degrees to 131 degrees. The 10–15-year survival of linked total elbow replacements with revision surgery as an end point is 92%.

A recent study compared three groups of consecutive patients who had undergone prosthetic elbow arthroplasty with the Souter–Strathclyde, Kudo, or Coonrad–Morrey implant for the treatment of RA. There were 33 elbows in each group. Clinical function was assessed on the basis of pain relief and the range of flexion. The study showed that this was similar in each group and that the component linkage with the Coonrad–Morrey implant prevented dislocation without an increased risk of loosening. Survivorship was assessed with use of a life-table method, with revision surgery and radiographic signs of loosening as the end points. Survival of the Coonrad–Morrey implant was better than that of the other two implants. The 5-year survival rates, with revision and radiographic signs of loosening as the end points, were 85% and 81% for the Souter–Strathclyde implant, 93% and 82% for the Kudo implant, and 90% and 86% for the Coonrad–Morrey implant. While radiographic evidence of loosening of the Coonrad–Morrey implants was less common, they noted focal osteolysis adjacent to 16% of the ulnar components and half of these cases progressed to frank loosening.

A discussion regarding the risks and benefits of elbow replacement should take place on several occasions prior to surgery. This should include the consequences of continued non-operative management together with discussion of the potential complications.

The complication rate following total elbow replacement is relatively high due to poor soft tissue coverage, the proximity of the ulnar nerve to the elbow joint, and the use of immunosuppressive drug therapy in patients with RA. The predominant complication is ulnar nerve damage, which, as a rule, is usually transient and reversible with permanent ulnar nerve damage uncommon. Deep infection is probably more common after elbow replacement compared to other joints and is reported to be from 2–11%. Wound healing has been a problem in 3–5% of patients, yet 75% of these required no additional surgical procedure.

Instability is a unique complication of the unlinked prosthesis and usually occurs soon after surgery. Inadequate soft tissue balance, either from the collateral ligaments or dynamically from the pull of the biceps, is the basic cause. The incidence of instability is less than 5%.

Intraoperative fracture has been reported in up to 5%, and if this occurs during an unlinked total elbow replacement, may require on-table revision to a linked prosthesis.

Aseptic loosening at 5 and 10 years requiring revision has been reported for the Souter–Strathclyde at 96% and 85% respectively. The longest follow-up of the Kudo implant gives a survivorship of 90% at 16 years. Twelve-year survivorship for the linked Coonrad–Morrey implant is 92%.

Fifty per cent of patients with RA have involvement of the elbow joint, usually bilaterally. Women are affected three times more commonly than men. There have been significant advances in the medical management of RA with the development of the disease-modifying agents. These are now used earlier in the disease, often in combination with each other. Anti-TNF drugs are used when disease-modifying agent combinations have failed to control symptoms. Intra-articular and intramuscular cortisone is an effective way of controlling flare-ups.

Synovectomy with or without excision of the radial head can be used earlier in the disease process when there is failure of medical management to control the symptoms of synovitis, particularly symptomatic radiocapitellar joint problems. The occasion for this procedure is diminishing with the advance of medical therapies.

Total elbow arthroplasty is indicated in severe RA where there is failure of medical management to control symptoms, particularly pain and loss of function. This gives good pain relief and restores a functional range of movement. When there is good bone stock a linked or an unlinked prosthesis can be used. When there is significant bone loss a linked prosthesis is used. However, there is a higher complication rate compared to lower limb arthroplasty, particularly ulnar nerve damage and wound healing problems. There are now published results of total elbow replacement beyond 10 years showing survival rates of between 80% (unlinked) and 92% (linked).