If your doctor suspects that you have an autoimmune disorder such as systemic lupus erythematosus (SLE), rheumatoid arthritis or scleroderma, he may arrange to have you take an antinuclear antibody (ANA) test. In these kinds of connective-tissue diseases, the immune-system antibodies that normally protect you from disease turn against your body's own cells. These self-produced 'autoantibodies' are churned out in large amounts, and cause inflammation throughout the body.
The ANA test is straightforward: a blood sample is taken, then tested for the presence of autoantibodies. Autoantibodies may attack different parts of a cell, but the ANA test is particularly good at detecting those that attack the cell nucleus.
Is the test accurate?
Not especially. A positive result is only an indication of the likelihood of your having the disease - it is not a definite confirmation of its presence.
The ANA test result is reported as a titre - a measure of how much the blood sample can be diluted and yet still show the presence of antibodies. The higher the titre, the more likely the presence of a connective tissue disorder such as SLE (Postgrad Med, 1993; 94: 55-66).
But what constitutes a positive result? In some studies, a titre of less than 1 in 40 was regarded as a positive result. However, a higher titre is usually considered more conclusive. For example, a titre of 1 in 80 suggests that an autoimmune disease is highly likely, and at least one commercial lab (Arup Laboratories, owned by the University of Utah) states in its guidelines that only titres of 1 in 160 or higher are to be considered significant positives.
How specific is the test?
At best, a positive ANA test only tells the doctor that there's a likelihood of connective tissue disease, but it can't tell you which one as a number of such conditions will produce a positive result. Whereas 95 per cent of SLE sufferers will test positive with the ANA (Rheum Dis Clin North Am, 1990; 16: 617-39), only 30-50 per cent of those with rheumatoid arthritis, 40-70 per cent with Sj"ogren's syndrome, 60-80 per cent with scleroderma and 20-50 per cent with chronic juvenile arthritis will (Arch Pathol Lab Med, 1999; 124: 71-81).
Healthy people can also show a positive result: around 2 per cent of the population have mildly elevated antibodies without symptoms (Adv Immunol, 1989; 44: 93-151). One study involving 15 international laboratories found that ANA tests of the general population were positive in 32 per cent of cases at a titre of 1 in 40, and in 5 per cent of people at a dilution of 1 in 160 (Arthritis Rheum, 1982; 25: 1271-7).
Drugs used to treat other disorders, such as procainamide (for heart arrhythmias), hydralazine (a vasodilator) and even the tetanus vaccine can return a positive result (Science, 1994; 266: 810-3), as can a form of lupus that is drug-induced.
Viral or bacterial infections, lung diseases (such as pulmonary hypertension), ulcerative colitis, cancers (of the skin, breast, lung and kidney) and even skin conditions like psoriasis can cause an increase in the number of antibodies produced.
Diagnosis is further complicated by the considerable overlap in symptoms of many connective-tissue diseases as well as the presence of various antibodies. For example, 'mixed connective-tissue disease' displays the symptoms of SLE, scleroderma and myositis, leading some to ask whether this is truly a separate entity at all (Arthritis Rheum, 1998; 41: 768-77).
Moreover, there is a high potential for false negatives (an all-clear result when you have the disease.) As autoimmune disorders often evolve over time, a significant number of patients produce negative ANA tests early on (Arthritis Rheum, 1999; 42: 1785-96), and only repeat testing at a later date can verify the disease status (Arch Pathol Lab Med, 1999; 124: 71-81).
Yet more inaccurate tests
If your ANA test results are positive, another set of tests can be ordered to differentiate between diseases by looking at specific nuclear proteins. But these tests are also low in sensitivity. So, while a positive result for double-stranded DNA and SM antibodies may confirm SLE, a negative test doesn't necessarily rule it out (Arch Pathol Lab Med, 1999; 124: 71-81).
Although these tests are not harmful in themselves, their biggest danger is their huge room for error. An incorrect diagnosis can set you off on a treatment you don't need, often requiring drugs with debilitating side-effects of their own.