Antineutrophil cytoplasmic antibodies (ANCA) are associated with a number of vasculitides, including Wegener granulomatosis, Churg-Strauss syndrome, microscopic polyangiitis and idiopathic necrotizing and crescentic glomerulonephritis. A number of connective tissue diseases (CTD's) present with similar findings to the vasculitides. If ANCA is specific to the vasculitic syndromes and remains negative in the CTD's, this would greatly enhance its utility. This study was designed to look at the results of several types of ANCA assays in patients with various connective tissue diseases.
Methods
Patients with connective tissue diseases were part of the Early Undifferentiated Connective Tissue Disease project, funded by the NIH. These patients were enrolled within one year of the onset of signs, symptoms or serology suggesting CTD, and were studied at entry and at years 1, 3 and 5. For the purpose of the current study, serum from 386 patients, obtained at baseline and stored frozen, was obtained and examined for ANCA antibodies.
The CTD diagnoses for these 386 patients, determined at the final 5-year visit, were:
Also studied were subsets of the above patients, 44 with Sjogren's syndrome (in addition to one of the above diagnoses) and 33 with the antiphospholipid syndrome.
Negative controls were serum samples from 200 blood donors. Positive controls were samples from 52 patients with Wegener granulomatosis, microscopic polyangiitis and related vasculitides. Half of these controls had anti-PR3 antibodies, half had anti-MPO antibodies (see below).
Clinically important ANCA is of two types: antiproteinase 3 antibodies (anti-PR3) and antimyeloperoxidase antibodies (anti-MPO). When these antibodies stain neutrophils using indirect immunofluorescence, anti-PR3 produces a cytoplasmic pattern of staining (C-ANCA), whereas anti-MPO produces a perinuclear or nuclear pattern (P-ANCA). Thus, the ANCA antibodies are anti-PR3 and anti-MPO, and the corresponding immunofluorescence patterns are C-ANCA and P-ANCA.
For this study, serum samples were tested for the two ANCA types using both immunofluorescence and ELISA methods.
The C-ANCA pattern of immunofluorescence was appropriately negative in all patients with CTD, in all blood donor controls and in all patients with MPO-positive vasculitis. It was appropriately positive in 23 out of 26 patients with PR-3-positive vasculitis.
The P-ANCA pattern was inappropriately positive in 28 out of the 386 patients with CTD, mainly in patients with lupus (22 out of 70 patients with lupus). It was appropriately negative in all blood donor controls, in all patients with PR-3 positive vasculitis and was appropriately positive in 21 out of the 26 patients with MPO-positive vasculitis.
The atypical ANCA pattern was seen in 89 out of the 386 patients with CTD (of all types). It was also positive in 3 out of 200 blood donor controls and in 5 out of 26 samples from patients with MPO-positive vasculitis.
The direct ELISA assay for anti-PR3 was inappropriately positive in 7 patients with CTD and in 2 patients with MPO-positive vasculitis. It was appropriately negative in all blood donor controls, and appropriately positive in 25 out of the 26 patients with PR3-positive vasculitis.
The sandwich ELISA for anti-PR3 was inappropriately positive in 4 patients with CTD. It was appropriately negative in all blood donor controls and all patients with MPO-positive vasculitis, and positive in 23 out of the 26 patients with PR-3 positive vasculitis.
The direct ELISA assay for anti-MPO was inappropriately positive in 2 patients with CTD. It was appropriately negative in all normal blood donors and in all patients with PR3-positive vasculitis, and positive in all 26 patients with MPO-positive vasculitis.
The final interpretation for anti-PR3 and anti-MPO positivity, which was based on a combination of the immunofluorescence and ELISA tests, performed significantly better than any of these tests individually. Anti-PR3 was negative in all CTD patients, all blood donors and patients with MPO-positive vasculitis. It was positive in 25 out of 26 patients with PR3-positive vasculitis. Anti-MPO was negative in all but 2 patients with CTD, all blood donors and all patients with PR3-positive vasculitis. It was positive in all 26 patients with MPO-positive vasculitis.
The authors note that their study confirms the specificity of ANCA for the diagnosis of Wegener granulomatosis, microscopic polyangiitis and related vasculitides. However, their results also demonstrate the inadequacy of immunofluorescence alone. Much more specific results were obtained with the ELISA techniques, and the best results were obtained using a combination of immunofluorescence and ELISA.
P-ANCA immunofluorescence performed particularly poorly in patients with lupus. This appears to be due to the antinuclear antibodies which can create a P-ANCA and atypical ANCA pattern on immunofluorescence.
Although C-ANCA was highly specific and was negative in all patients with CTD in this study, many of the negative C-ANCA assays were discordant on initial analysis and only negative after a second staining. This underlines the subjectivity of the immunofluorescence method.
This study examining the occurence of ANCA in patients with connective tissue disease is an important one. Its clinical importance lies in when patients have early manifestations of connective tissue disease but may have aggressive organ involvement such as RPGN. One of the criticisms that the authors admit to is that the data was derived from a large study that looked at patients with UCTD. In addition, samples for the patients for anticardiolipin antibody syndrome were obtained from another laboratory. However, the results confirm other studies' findings of the high specificity for cANCA. The low specificity of pANCA is mainly due to the immunofluorescent staining by antinuclear antibodies. To improve specificity, the authors suggest that myeloperoxidase antibodies be measured by ELISA or Sandwich techniques.
Gurjit Kaeley
Rheumatology Fellow. UT, Memphis.
Submit a comment about this article
Return to Journal Club home page