Rheumatology Advance Access originally published online on November 16, 2004
Rheumatology 2005 44(2):247-250; doi:10.1093/rheumatology/keh467
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Rheumatology Vol. 44 No. 2 © British Society for Rheumatology 2004; all rights reserved
Low seroprevalence and poor specificity of antineutrophil cytoplasmic antibodies in tuberculosis
Department of Internal Medicine, Unité de Recherche Clinique et Thérapeutique (EA 3409), 1 Department of Bacteriology–Virology, 2 Department of Pneumology, 3 Department of Infectious and Tropical Diseases, Hôpital Avicenne, AP–HP, Université Paris-Nord, Bobigny and 4 Department of Immunology, Hôpital Necker, AP-HP, Paris, France.
Correspondence to: A. Mahr, Department of Internal Medicine, Hôpital Cochin, 27, rue du Faubourg Saint-Jacques, 75679 Paris Cedex 14, France. E-mail: alfred.mahr{at}cch.ap-hop-paris.fr
Objectives. Recently published findings suggested that antineutrophil cytoplasmic antibodies (ANCA), particularly those with a cytoplasmic (C-ANCA) labelling pattern and targeting proteinase 3 (anti-PR3), might be markers of tuberculosis (TB). This is a critical issue, because C-ANCA/anti-PR3 were considered to be a highly specific hallmark of Wegener's granulomatosis or microscopic polyangiitis and because TB may clinically mimic Wegener's granulomatosis. We therefore undertook a study with the aim of investigating further the prevalence and specificity of ANCA in TB.
Methods. We evaluated serum samples from 67 patients diagnosed with culture-proven TB and 10 previously untested control samples from patients known to be ANCA positive (four Wegener's granulomatosis and two microscopic polyangiitides) or negative. All 77 sera were screened for ANCA using commercially available indirect immunofluorescence (IIF) and enzyme-linked immunosorbent assay (ELISA) for anti-PR3 and antimyeloperoxidase (MPO). IIF-positive and anti-PR3- and anti-MPO-negative sera were also tested for bactericidal/permeability-increasing protein, lactoferrin, elastase and cathepsin G specificities with commercially available ELISA.
Results. IIF detected ANCA in seven (10%) of the TB sera, including three C-ANCA and four atypical perinuclear-labelling ANCA. Only one IIF-negative specimen was anti-PR3 positive in ELISA. ANCA testing of the control sera yielded IIF and ELISA results concordant with previous findings, except for one borderline ELISA.
Conclusion. Our results indicate that TB is associated with low ANCA seroprevalence and poor specificity, with no test serum showing combined C-ANCA/anti-PR3 activity. In a clinical setting of Wegener's granulomatosis/TB mimicry, such combined reactivity would seem to be more suggestive of Wegener's granulomatosis.
KEY WORDS: Antineutrophil cytoplasmic antibodies, Tuberculosis, Prevalence, Specificity