Immune complex deposits in ANCA-associated crescentic glomerulonephritis: a study of 126 cases

BACKGROUND: Necrotizing and crescentic glomerulonephritis related to antineutrophil cytoplasmic autoantibodies (ANCA) is typically referred to as "pauci-immune"; however, it is not unusual for renal biopsies in such cases to exhibit some immune complex deposition within glomeruli on immunofluorescence and/or electron microscopic study. The composition and intraglomerular localization of such deposits in ANCA-glomerulonephritis has not been widely studied, and their potential pathologic and clinical significance is not clear, although a possible synergistic effect between immune complexes and ANCA in producing more severe glomerulonephritis is suggested by some human and animal studies. METHODS: Electron micrographs from 126 renal biopsies showing necrotizing/crescentic glomerulonephritis characterized by positive ANCA serology [C-ANCA, anti-proteinase 3 (anti-PR3), or anti-myeloperoxidase (MPO)] or necrotizing arteritis in the absence of known ANCA results were examined for the presence, quantity, and location of electron-dense deposits. The presence or absence of such deposits was correlated with histologic findings (fraction of glomeruli with crescents and segmental necrotizing lesions, mesangial and endocapillary hypercellularity), immunofluorescence findings, and clinical data, including serum creatinine and 24-hour urine protein levels at the time of biopsy. RESULTS: Sixty-eight (54%) of these biopsies showed glomerular immune complex deposits on electron microscopy; 87% of the latter also showed positive immunofluorescence findings for at least one immunoglobulin or complement component, although staining was relatively mild in most instances (< or =2+ on a 0 to 4+ scale in all but eight cases). Nearly half of biopsies negative for deposits by electron microscopy also showed positive immunofluorescence findings, though even more so than in cases with deposits on electron microscopy the intensity of immunofluorescence staining in these biopsies was typically very weak (trace or trace to 1+ in most cases, none >2+). Hypercellularity within the glomerular tuft was seen in 50% of biopsies with deposits on electron microscopy but only 14% of those without deposits; in each group this was usually mild and mesangial. Notably, the presence of deposits on electron microscopy was associated with a higher median level of proteinuria (3.2 versus 1.3 g/24 hours, P < 0.0001) and a higher median percentage of glomeruli with crescents (62.5% versus 44.0%, P= 0.06). CONCLUSION: Immune complex deposits were found on electron microscopy in just over half of renal biopsies with crescentic glomerulonephritis associated with positive ANCA serology and/or necrotizing arteritis. Clinical correlations suggest that these immune complex deposits may somehow potentiate the effect of ANCA in producing glomerulonephritis.