MO444MACROPHAGE DENSITIES CORRELATE WITH LONG-TERM FUNCTION IN PAUCI-IMMUNE AND MEMBRANOUS GLOMERULONEPHRITIS AS WELL AS IN HYPERTENSIVE NEPHROPATHY

Abstract

Abstract Background and Aims Macrophages and monocytes are main players in innate immunity. In renal diseases, their role is poorly understood. Our multicentric cross-sectional study aimed to study the prevalence of macrophages and monocytes in various human native kidney diseases. For this, we used precise pixel-based digital quantification of their densities in renal biopsies and correlated our findings with clinical data. Method We included 324 patients, who underwent a diagnostic renal biopsy. Additional normal kidney samples from 16 tumour nephrectomies were used as controls. According to the diagnosed diseases, we established 17 patient groups. Biopsies were stained for CD68+-macrophages using automated immunohistochemistry (Ventana Ultra) and selected groups were further subtyped for CD14+-monocytes and CD163+-M2-macrophages (67 cases, pauci-immune glomerulonephritis (PIGN), IgA-nephropathy (IgAN) and control samples). Digitized sections (Leica) were analysed using the open-source software QuPath to quantify cell densities (positively stained areas displayed as percentages of ROI) in renal cortex, medulla and extrarenal tissue, respectively. Detailed clinical and laboratory data at timepoint of biopsy were available for all patients. Additional data for follow-up were achievable in 158 cases. Results Renal disease samples presented higher mean macrophage densities compared to control cases (CD68: cortex 1.2 vs. 0.2%, p<0.001, medulla 0.8 vs. 0.04%, p<0.001; CD163: cortex 3.2 vs. 0.5%, p<0.001, medulla 2.3 vs. 0.6%, p<0.05), but CD14+-density did not differ between patients and control samples. The highest cortical CD68+-density occurred in PIGN (1.98%) and in medulla in ascending infections (1.86%). The lowest cortical CD68+-densities were measured in thin basal membrane syndrome / Alport-syndrome (0.56%) and in medulla in immunotactoid and fibrillary glomerulopathy (0.26%). Chronic kidney disease displayed lower percentages of CD68+-densities (cortex: 1.15%; medulla: 0.49%) compared to acute kidney injury (cortex: 1.84%, p<0.001; medulla: 1.08%, p<0.001) and acute on chronic kidney injury (cortex: 1.81%, p<0.001; medulla: 1.43%, p<0.001). We detected a correlation of CD68+- and CD163+-infiltration with kidney function (eGFR) in cortex and medulla at the time of biopsy (CD68: r=-0.51 for cortex, r=-0.60 for medulla; CD163: r=-0.71 for cortex, r=-0.73 for medulla; p<0.001) and follow up (CD68: r=-0.41 for cortex, r=-0.34 for medulla, p<0.001; CD163: r=-0.46 for cortex, r=-0.50 for medulla, p<0.05). Older patients (>64 years) showed a higher medullary M2-infiltration (1.81% vs. 4.34%, p<0.005). The eGFR at the time of biopsy inversely correlated (p<0.05) with cortical CD68+-density in IgAN (r=-0.39), PIGN (r=-0.53), membranous glomerulonephritis (MGN; r=-0.70), focal segmental glomerulonephritis (r=-0.63), and hypertensive nephropathy (HNP; r=-0.44). At follow-up, this correlation (p<0.05) was still present in PIGN (r=-0.43), MGN (r=-0.58), and HNP (r=-0.77). In PIGN, cortical CD163+-density and eGFR were associated (p<0.001) at timepoint of biopsy (r=-0.51) and follow-up (r=-0.51). Particularly, cANCA-vasculitis showed a strong correlation between eGFR and cortical CD68+- as well as CD163+-densities at time of biopsy (CD68: r=-0.78; CD163: r=-0.75, p<0.001) and also for follow-up (CD68: r=-0.48; CD163: r=-0.68, p<0.05). Conclusion Macrophages may promote progression of human renal diseases, whereas monocytes do not correlate with eGFR-decline. Especially, in cANCA- vasculitis CD163+- infiltration is associated with renal outcome. Additional studies are needed to investigate, whether macrophages can serve as predictive markers or therapeutical targets in native renal diseases.