Testing on formalin-fixed neutrophils is less sensitive and specific for small vessel vasculitis, and less sensitive for MPO-ANCA, than most ELISAs

Abstract

This study evaluated the usefulness of testing sera with perinuclear antineutrophil cytoplasmic antibodies (P-ANCA) on formalin-fixed neutrophils to differentiate between vasculitis and other diseases, and to distinguish between myeloperoxidase (MPO) and other antigen targets. Sera from active (n = 24) or treated vasculitis (n = 23) and non-vasculitic disease (n = 37) were tested in 3 ethanol- and 2 formalin-fixed neutrophil assays, and in 12 MPO-ANCA ELISAs. The sensitivity of demonstrating that P-ANCA became cytoplasmic on formalin-fixed cells in vasculitis, and negative in non-vasculitic disease was 75–79% in different assays compared with 88% for positivity in the majority of MPO-ANCA ELISAs. The sensitivity and specificity of demonstrating that P-ANCA became cytoplasmic where the target was MPO, and disappeared with other antigens were 89% and 74–80% respectively in different assays. P-ANCA specificity for MPO should be confirmed in one of the better-performing MPO-ANCA ELISAs.

Introduction

Antineutrophil cytoplasmic antibodies (ANCA) are autoantibodies directed against the cytoplasmic granules of neutrophils, and are typically found in Wegener's granulomatosis and microscopic polyangiitis (Davies et al., 1982, van der Woude et al., 1985, Savage et al., 1987). The ‘International Consensus Statement on Testing and Reporting ANCA’ recommends that all sera from patients with suspected small vessel vasculitis should be screened by indirect immunofluorescence (IIF) on ethanol-fixed neutrophils, and positive fluorescence confirmed in antigen-specific ELISAs (Savige et al., 1999).

Perinuclear fluorescence (P-ANCA) where the target antigen is myeloperoxidase (MPO) occurs in up to 80% patients with microscopic polyangiitis, but P-ANCA with other specificities is found in many other conditions including inflammatory bowel disease (IBD) and systemic lupus erythematosus (SLE) (Falk and Jennette, 1988, Saxon et al., 1990, Galeazzi et al., 1998). The distinction between these different types of P-ANCA, and the early diagnosis of microscopic polyangiitis is critical because about half of all undiagnosed patients develop renal failure within one year.

Most laboratories confirm the diagnosis of microscopic polyangiitis with an MPO-ANCA-specific ELISA but some persist with IIF on formalin-fixed neutrophils, and others do both (Lock, 1994, Bird, 1999, Chowdhury et al., 1999). The P-ANCA in vasculitis is an artefact that occurs because ethanol fixation permeabilises the neutrophil granule membrane allowing the MPO to diffuse out into the cytoplasm and adhere to the nucleus through a charge effect (Falk and Jennette, 1988).The same sera often produce cytoplasmic fluorescence on formalin-fixed cells because the MPO remains within the granules. In contrast formalin fixation typically destroys the target antigens of other P-ANCA (including those due to antinuclear antibodies) and fluorescence is negative.

Most laboratories screen sera for vasculitis by IIF on ethanol-fixed neutrophils, and confirm P-ANCA-positive sera in MPO-ANCA ELISAs or by IIF on formalin-fixed cells. Previous reports of the sensitivity of formalin-fixed neutrophil testing to distinguish vasculitis from non-vasculitic disease, and to detect or exclude MPO-ANCA have been inconsistent (Chowdhury et al., 1999, Radice et al., 2000). Neither study examined both of these capabilities, determined the sensitivity of this method for the entire range of antibody binding, nor compared IIF testing with the results from more than one MPO-ANCA ELISA. In addition, we understand now that all MPO-ANCA ELISAs perform differently especially where antibody levels are low (Trevisin et al., 2008).

We describe here a study that has systematically determined the sensitivity and specificity of testing P-ANCA-positive sera in formalin-fixed neutrophil IIF assays in both active and treated vasculitis, and in non-vasculitic disease, that is, across the full range of antibody binding. We then compared the performance of the formalin IIF assays with the results from 12 different MP0-ANCA ELISAs. The results demonstrate conclusively that P-ANCA-positive sera should be confirmed in one of the better-performing ELISAs rather than by IIF on formalin-fixed neutrophils.