Essential facts

1. Investigations are useful to support a diagnosis, but should not be performed in the absence of a clinical indication; remember that false-positive tests for autoantibodies are common, especially for rheumatoid factor and anti-nuclear antibody (ANA).

2. A number of drugs can induce the development of autoantibodies, and only occasionally does this lead to clinical disease.

3. Anti-cyclic citrullinated peptide antibody (ACPA) testing for rheumatoid arthritis (RA) is very specific in early disease, but lacks sensitivity.

4. Many rheumatic diseases and their treatments can result in systemic abnormalities including anaemia, low white cell count, thrombocytopenia and abnormal liver function.

5. The ‘ALARA’ principle (‘as low as reasonably achievable’) underpins the use of ionizing radiation.

6. Treat the patient, not the image.

After a thorough history and clinical examination, further investigations can help to confirm or exclude a diagnosis. Important factors to consider are age, sex, family history of rheumatic disease, medication, pregnancy, infection, malignancy, and liver, pulmonary, haematological, endocrine or skin diseases. Laboratory or imaging investigations used in isolation will not usually provide a diagnosis and the result will depend on the pre-test probability of the diagnosis being present, which in turn is influenced significantly by the clinical findings. This principle is a very important one to follow. It can prevent unnecessary and sometimes costly tests being performed and potentially causing distress to patients. Laboratory tests are important in monitoring drug toxicity, especially for patients being treated with disease-modifying anti-rheumatic drug therapy, and can be used to monitor disease progress or to identify complications of disease. Further specialist investigations may be required in some circumstances, especially in patients with multi-system disease (such as echocardiography, lung function testing, high-resolution computed tomography (CT), positron emission tomography (PET)), but these are beyond the remit of this book. Their use is indicated in the relevant disease area.

Haematology

Most systemic rheumatic diseases are associated with normocytic normochromic anaemia; occasionally the anaemia may be the first clue to a diagnosis. Anaemia may be a result of drug therapy. Iron-deficient anaemia may result from non-steroidal anti-inflammatory drug (NSAID)-induced peptic ulceration and bleeding. Macrocytic anaemia can be due to methotrexate-induced folate deficiency, which can be avoided by providing routine folate supplements. Haemolysis may be a manifestation of systemic lupus erythematosus (SLE). Platelet counts are usually raised in response to inflammation, but may be low as a direct result of disease such as SLE or, rarely, Felty’s syndrome. More commonly, platelet counts are reduced as a result of drug toxicity. Neutropenia occurs in SLE, but can also be a sign of marrow toxicity from drug therapy. Raised neutrophil counts may warn of a possible infectious complication; however, it is important to remember that steroids reduce the ability of neutrophils to marginate in blood vessels, which results in an apparent rise in neutrophil counts, typically by one-third. Lymphopenia is common in SLE and may be induced by drugs. One particular treatment for RA (monoclonal B-cell therapy) actually depletes B lymphocytes by targeting the CD20 molecule on their cell surface. Very low lymphocyte counts predispose to fungal and viral infections. Abnormal clotting is typical of primary and secondary antiphospholipid syndrome, with delayed activated partial thromboplastin time (APTT) and prothrombin time (PT), together with the presence of a circulating lupus anticoagulant and, typically, anticardiolipin antibodies.

Most of the disease-modifying anti-rheumatic drugs and some biological agents have the potential to cause marrow toxicity. Regular monitoring of platelet counts and white cell counts, especially neutrophils, can anticipate and prevent serious illness, by withdrawing or reducing the dose of the drug before clinical problems arise.

ESR and CRP

The erythrocyte sedimentation rate (ESR) is a very non-specific measure, which is usually increased in inflammatory states (including systemic infections), but may also be increased in pregnant women, older people, or in individuals with high circulating levels of immunoglobulins or lipids. The C-reactive protein (CRP, an acute-phase protein) is a more direct measure of inflammation, as it is produced by the liver in response to systemic effects of pro-inflammatory cytokine production, such as interleukin (IL)-1, IL-6, IL-17 and tumour necrosis factor (TNF) α. CRP concentration is also increased in patients with infection. It is a useful investigation in bone and joint infection to monitor response to therapy.

Microbiology

In all suspected cases of septic arthritis or osteomyelitis, attempts should be made to obtain a specimen of infected fluid or tissue for microscopy and culture prior to starting antibiotics. Blood and bone marrow culture may be helpful. Some infections can be inferred by using serial serological tests to demonstrate a change in the immune response to a specific antigen. For example, a transient rise in immunoglobulin (Ig) M antibodies directed against parvovirus B19, followed by a subsequent rise in IgG levels against the virus, is a good indicator of recent exposure to the virus, which is known to cause a self-limiting arthritis in adults.

Biochemistry

All patients receiving methotrexate, leflunomide or sulfasalazine require regular monitoring of liver function. In patients with active inflammatory joint disease, it is common to observe a mild increase in liver enzyme levels as part of the systemic inflammation, but this may cause confusion when starting disease-modifying anti-rheumatic drugs (DMARDs). In practice, it may be necessary to induce brief improvement in such patients with steroids, settling the inflammation and normalizing liver function before starting a DMARD. Renal function should be tested regularly in patients with SLE and vasculitis. Patients receiving ciclosporin (not commonly used in rheumatology practice) need to have regular creatinine monitoring for possible nephrotoxicity.

Clues to a diagnosis of gout might include a good history, a normal physical examination, except during an acute attack, but supported by a raised serum urate concentration. Urate levels may be transiently normal following an acute attack, as a result of shedding of crystals into the joint, with a subsequent rise to above normal when the patient recovers.

Muscle enzyme levels are usually increased in inflammatory muscle disease and metabolic storage diseases, compartment syndrome and injuries associated with muscle necrosis. Estimation of creatine kinase (CK) levels may be helpful in patients who are suspected of having muscle necrosis, but who are unable to provide a reliable history. Apart from CK, aldolase, lactate dehydrogenase, alanine aminotransferase and aspartate aminotransferase levels can also be increased. Often the levels of CK are highest in metabolic muscle disease (typically in the several thousands) compared with levels in inflammatory muscle disease (typically a few thousand or less).

Bone biochemistry includes serum calcium and phosphate levels, and the level of bone-derived alkaline phosphatase. It is now possible to measure parathyroid hormone levels in blood, as well as levels of vitamin D, so that a more precise diagnosis can be made. Markers of bone turnover, such as P1NP (pro-collagen type 1 nitrogenous pro-peptide), can help in determining response to therapy in osteoporosis. Osteomalacia and rickets are typically associated with low calcium, high alkaline phosphatase and low vitamin D levels. High levels of alkaline phosphatase are typically found during active Paget’s disease, but also occur in the presence of bony metastases or in patients with liver disease. Alkaline phosphatase levels are higher than normal in pregnancy and during childhood and adolescence.

Immunology testing

Immunoglobulin levels may be increased in many autoimmune diseases, especially if significant amounts of autoantibody are being produced. This will result in a polyclonal increase in immunoglobulins, typically of the IgG class. Monoclonal increase of immunoglobulin suggests the presence of a myeloproliferative disease, such as myeloma, which may occasionally present with aches and pains and a raised ESR, but typically a normal CRP level. By contrast, repeated treatment with rituximab or other anti-B-cell agents has been found to reduce IgG and IgM levels, raising concern over potential long-term immunocompromise.

Many healthy individuals make rheumatoid factor and anti-nuclear antibodies. Approximately 25% of healthy elderly people make rheumatoid factor (compared with around 4% of 20-year-olds); one-third of healthy individuals produce low-titre anti-nuclear antibody (ANA). At all ages, for unknown reasons, healthy females are more likely than healthy males to have circulating ANAs. Although up to one-third of first-degree relatives of patients with SLE have positive ANA tests, the majority never develop SLE.

In many pathological conditions involving tissue damage, immune activation may be directed against normal healthy tissue. The initial damaging events usually involve T cells, but released antigens may become targets for an antibody response. The detection and quantification of autoantibodies is an important aspect of diagnosis of autoimmune diseases, such as rheumatoid arthritis (RA), SLE, systemic sclerosis, and the systemic vasculitides. Many of these diseases are associated with a specific autoantibody or group of autoantibodies. They are usually detected by their reaction against tissue components using subjective methods such as indirect immunofluorescence. Positive samples should be further analysed using specific, quantitative methods. Autoantibodies are not ‘gold standard’ tests; they are markers of the disease, but with significant limitations. They should be used as part of a diagnostic workup, including the clinical history and examination findings, rather than as a marker indicating one particular disease. However, autoantibodies may be of prognostic value; for example, patients with RA who are rheumatoid factor (RF) or anti-cyclic citrullinated peptide antibody (ACPA) positive have a worse prognosis compared with those with seronegative RA.

Techniques are gradually improving, giving numerical results rather than titres, but a lack of standardization makes the results variable. Many of the antibodies show no correlation with disease activity, and should be regarded as indicators of the potential presence of disease. Their use should be restricted to the initial investigation in most instances and not repeated every time the patient is followed up (e.g. RF and ANA). Exceptions to this include serial measurement of complement components levels C3 and C4 (which typically fall below normal in active disease because complement is being consumed), and double-stranded DNA (dsDNA) in SLE, anti-neutrophil cytoplasm antibody (ANCA) titres (but interpret with caution) in small-vessel vasculitis and C-reactive protein in RA, all of which may provide useful information in the appropriate clinical context.

RF and ACPA

Rheumatoid factors are autoantibodies, usually of the IgM class, directed against the Fc portion of normal IgG molecules, and are the commonest autoantibodies in humans. They are present in polyclonal form in a large number of autoimmune disorders as well as in some inflammatory states (Table 4.1). In monoclonal form, they are produced by proliferating B cells (for example in Waldenström’s macroglobulinaemia and chronic lymphoid leukaemia). In addition, RFs are part of the normal antibody repertoire, especially in the fetal and neonatal periods, suggesting that they may play a role in the development of the immune system. Healthy individuals are capable of making RF, but not usually in large amounts. Some 70% of patients with RA have increased levels of RF. The presence of RF is associated with a worse prognosis and a higher incidence of systemic manifestations of the disease. Early in the course of RA, however, RF may not be detectable, and it is worth repeating the test some months later in patients who have clinical features of RA. RF is also produced in other autoimmune rheumatic diseases, chronic infection and vasculitis, resulting in low specificity for RA. Anti-filaggrin antibodies, anti-keratin antibody (AKA) and anti-perinuclear factor (APF) are more specific than RF and may be detected in early RA. However, the assays are not standardized. Cyclic citrullinated peptides (CCPs) of filaggrin are the target of those antibodies and a more specific enzyme-linked immunosorbent assay (ELISA) has been developed. In established disease, ACPAs have a specificity greater than 90% and a sensitivity of 70%, and may replace RF testing in future. However, in patients with clinical evidence of early RA, although the specificity of RF and ACPA testing is very high (above 95%), around 20–30% of patients are negative for RF and/or ACPA.

Table 4.1 Conditions in which a positive rheumatoid factor test may be found

Normal health	Normal population, especially the elderly
Autoimmune rheumatic disease	High prevalence in rheumatoid arthritis, but also seen in Sjögren syndrome, systemic lupus erythematosus and hepatitis C-associated mixed cryoglobulinaemia; chronic inflammatory lung disease
Juvenile arthritis	Found in a small number of children with polyarticular juvenile idiopathic arthritis
Infectious diseases	Chronic bacterial infections, including sub-acute bacterial endocarditis, infective exacerbations of cystic fibrosis, tuberculosis, leprosy, trypanosomiasis, visceral larva migrans, infectious mononucleosis, influenza A, hepatitis A and cytomegalovirus infections
Endocrine diseases	Graves’ disease and hypothyroidism
Liver disease	Chronic liver disease, autoimmune hepatitis
Malignancy	Some malignancies including B-cell proliferative diseases, adenocarcinoma of the bladder

ANA, double-stranded DNA antibodies and ENA

Over 85% of patients with SLE will be positive for ANA, and in the correct clinical setting this is a very useful test in diagnosis. The ANA test is sensitive, but not specific (Table 4.2). Healthy people often have positive results. The test should not be used to evaluate vague symptoms, but to support a diagnosis of a autoimmune rheumatic disease if a patient has appropriate symptoms or signs. The final diagnosis is based on clinical findings, although test results may support the diagnosis, or prompt further specific investigation.

Table 4.2 Conditions in which positive anti-nuclear antibody tests may be found

Normal health	Normal population, especially the elderly, transient during pregnancy
Autoimmune rheumatic disease	High prevalence in systemic lupus erythematosus, but also seen in scleroderma, dermatomyositis, Sjögren syndrome, mixed connective tissue disease and long-standing rheumatoid arthritis
Juvenile arthritis	Highest prevalence in girls with oligoarticular juvenile arthritis, but also seen in polyarticular and systemic disease
Infectious disease	Malaria, bacterial infections, transient rise during many viral infections
Endocrine disease	Type 1 diabetes, autoimmune thyroiditis
Liver disease	Chronic liver disease, autoimmune hepatitis
Malignancy	Breast cancer, squamous cell carcinoma of the lung and hepatocellular carcinoma
Haematological disease	Autoimmune thrombocytopenic purpura and haemolytic anaemia
Drugs	Many reported including minocycline, penicillamine, hydralazine, procainamide, isoniazid, methyldopa and quinidine