Prevalence of antineutrophil cytoplasmic antibodies in a large inception cohort of patients with connective tissue disease

Ann Intern Med. 1997 Jun 1;126(11):866-73. doi: 10.7326/0003-4819-126-11-199706010-00003.


Background: Two types of antineutrophil cytoplasmic antibodies (ANCA), antiproteinase 3 antibodies (anti-PR3) and antimyeloperoxidase antibodies (anti-MPO), are useful in the diagnosis of such types of vasculitis as Wegener granulomatosis and microscopic polyangiitis. Connective tissue diseases frequently appear in the differential diagnosis of this spectrum of vasculitis.

Objective: To determine the prevalence of ANCA in patients with connective tissue disease.

Design: Blinded, controlled study of a 5-year inception cohort.

Setting: Tertiary-care university teaching hospitals.

Patients: 70 patients with rheumatoid arthritis, 70 patients with systemic lupus erythematosus, 45 patients with scleroderma, 36 patients with inflammatory myositis, 44 patients with the sjögren syndrome, 33 patients with the antiphospholipid syndrome, and 165 patients with early undifferentiated connective tissue disease (EUCTD). Serum was taken from 200 blood donors and 52 patients who had known vasculitis and positive results on tests for anti-PR3 or anti-MPO; these patients served as controls.

Measurements: The presence of anti-PR3 and anti-MPO was determined by combining the results of indirect immunofluorescence tests for cytoplasmic (C-ANCA) and perinuclear (P-ANCA) patterns with the results of enzymelinked immunosorbent assays (ELISAs) directed to measure antigen.

Results: Cytoplasmic ANCA was not detected in any study or control patient. Perinuclear ANCA was commonly detected among patients with lupus (31%) but was uncommon among patients in other groups (0% to 5%). In all cases, P-ANCA was associated with the presence of antinuclear antibodies. Atypical ANCA immunofluorescence patterns were fairly common in all groups (11% to 39%). Antiproteinase 3 was detected by ELISA in study patients (1 patient with rheumatoid arthritis, 1 with lupus, 1 with polymyositis, and 6 with EUCTD). Antimyeloperoxidase was detected by ELISA in 2 study patients (1 with rheumatoid arthritis and 1 with lupus). None of the patients with positive ELISA results had evidence of renal vasculitis during follow-up. When an ANCA scoring system that combines immunofluorescence and ELISA was used, the test specificity for vasculitis was 99.5% among patients with connective tissue disease.

Conclusions: Patients with connective tissue disease are known to develop multiple autoantibodies; positivity for anti-PR3 and anti-MPO ANCA in such patients is highly specific for anti-PR3. However, P-ANCA immunofluorescence, which may have positive results because of the presence of antinuclear antibodies, is not a specific marker of anti-MPO. A rigorous ANCA testing system that combines the results of immunofluorescence with those of ELISA is highly specific for Wegener granulomatosis and related vasculitides even in patients with connective tissue disease.

Publication types

  • Clinical Trial
  • Controlled Clinical Trial
  • Multicenter Study

MeSH terms

  • Antibodies, Antineutrophil Cytoplasmic / blood*
  • Autoantigens / immunology*
  • Connective Tissue Diseases / diagnosis
  • Connective Tissue Diseases / immunology*
  • Enzyme-Linked Immunosorbent Assay
  • Fluorescent Antibody Technique, Indirect
  • Follow-Up Studies
  • Humans
  • Myeloblastin
  • Peroxidase / immunology*
  • Prospective Studies
  • Serine Endopeptidases / immunology*


  • Antibodies, Antineutrophil Cytoplasmic
  • Autoantigens
  • Peroxidase
  • Serine Endopeptidases
  • Myeloblastin