Rao JK, Allen NB, Feussner JR, Weinberger M
A prospective study of antineutrophil cytoplasmic antibody (c- ANCA) and clinical criteria in diagnosing Wegener's granulomatosis

Lancet (Oct) 346:926-931 1995

Rao et al. performed a prospective study to determine how c-ANCA may be useful in patients with suspected vasculutis and in whom Wegener's granulomatosis is in the differential diagnosis. They also looked at the effect of baseline corticosteroid use or disease activity on c-ANCA values in these patients. The patient population included inpatients and outpatients evaluated by a rheumatologist for possible vasculitis at Duke University Medical Center between January and December 1993. They excluded patients with known Wegener's and patients who had received immunosuppressive or disease modifying antirheumatic therapy at the time of evaluation. A second group of patients was obtained by review of records from patients who had an ANCA test in the clinical immunology laboratory. ANCA was performed by indirect immunofluorescence. Diagnosis of Wegener's granulomatosis was based on the 1990 American College of Rheumatology (ARC) criteria for Wegener's granulomatosis.

Of the 346 patients eligible for study, 252 were retained, 174 of whom had not received steroids. 191 patients were entered in the study by direct referral, this group having more symptoms and symptomatic leukocytoclastic vasculitis compared with those obtained by reviewing laboratory records for positive ANCA (n=78). Follow-up information was obtained in 84% (212) patients at a mean of 5.1 months.

By ACR criteria, the diagnosis of Wegener's was established in 25 patients, and the c-ANCA was positive in 7 of these, for a sensitivity of 7/25 or 28%. During the follow up period, six of these 25 patients were diagnosed with Wegener's granulomatosis by typical clinical and histopathologic findings, and five of the six were c-ANCA positive. Nine of the 25 were ultimately diagnosed with other vasculitides, and the remaining 10 were diagnosed with other diseases. None of the patients initially negative by ACR criteria were diagnosed with Wegener's during follow up. Fourteen of the 212 patients overall had a positive c-ANCA. The overall specificity was therefore 96%. The positive predictive value of the test was 0.5. The specificity of the c-ANCA test was not effected by baseline steroid use; however, the sensitivities were poor in all patients, whether on or off steroids. In patients with inactive disease, the sensitivity of c-ANCA was half that of patients with active disease.

The authors conclude that the c-ANCA test is extremely specific independent of steroid use or disease activity. A negative c-ANCA is therefore likely to be a "true negative" in patients with suspected vasculitis. Sensitivity, however, was much less than in previous studies, being only 28%, in patients with ACR-defined Wegener's granulomatosis and only 33% in patients with active disease. It is acknowledged, however, that sensitivity would have been much higher based on the biopsy criteria for a Wegener's diagnosis (5/6 such patients being c-ANCA positive). In any case, nearly 50% of the patients with a positive c-ANCA did not have Wegener's granulomatosis. The authors state that basing therapeutic decisions on a positive c-ANCA test and clinical symptoms of Wegener's granulomatosis alone may be seriously flawed.

The authors also point out that their study differs from others by its prospective nature, using patients with suspected vasculitis and application of a single reference standard, the 1990 ACR criteria. They also pointed out the limitations of their study including the referral bias associated with using university-based rheumatologists and the small number of patients with biopsy proven Wegener's granulomatosis. They also note that of the 16 patients lost to follow-up, some may have eventually developed Wegener's granulomatosis and also, because not every patient had a biopsy, those with occult disease might not have been diagnosed.

Comment: This study helps the clinician put c-ANCA testing in perspective. The c-ANCA must be interpreted within the context of the patient's clinical picture, diagnostic studies, and likelihood of the disease. (Krane)

To go back use the BACK button on your browser.
Otherwise click on the desired link to this article below:
ARF etiology : Acute glomerulonephritis/RPGN
Proteinuria/Hematuria : Vasculitis (Wegener's PAN, etc.)