MOLECULAR MECHANISMS OF CHRONIC KIDNEY GRAFT INJURY

J.C. Papadimitriou¹, C. Drachenberg², A. Haririan³, R. Munivenkatappa^4
1Pathology, University of Maryland School of Medicine, Baltimore/UNITED STATES OF AMERICA, ²Pathology, University of Maryland School of Medicine, Baltimore/MD/UNITED STATES OF AMERICA, ³Medicine, University of Maryland School of Medicine, Baltimore/MD/UNITED STATES OF AMERICA, ⁴, University of Maryland School of Medicine, Baltimore/MD/UNITED STATES OF AMERICA

Body: Introduction: TG is the most characteristic and important histological feature of chronic antibody mediated rejection (AMR). Its pathogenesis and early morphological recognition is a subject of significant academic as well as clinical/diagnostic relevance. Methods: Biopsies (Bx) from 240 patients with grafts functioning for ≥1 year (1-6 Bx/ patient,n=649) were examined with routine stains, C4d stain and electron microscopy (EM). In addition, CD3, CD20 and CD68 stains were done in all Bx in order to identify T and B lymphocytes and macrophages respectively. For each patient the Bx with earliest G pathology or the last Bx if no G pathology was present, was selected and inflammatory cells were quantitated in the least and most inflamed G. Results: Based on the G pathology Bx were classified in eight groups: normal 20.8% (n=50), ischemic 30.8% (n=74), diabetic glomerulosclerosis (DG) 8.4%, various types of glomerulonephritis (GN) 7%, FSGS 8.4%, transplant glomerulitis (TGitis) 6.6%, TG 16.3% and thrombotic microangiopathy 1.7%. TG had a significantly higher number of C68+ cells compared to all other categories (p=.015 vs TGitis, p=.001 vs. all others). Similarly, CD3+ were higher in TG than all other goups except TGitis, but much less so than CD68+. The number of neutrophils correlated with CD68+ (r=.61) but by themselves neutrophils did not discriminate TG from other pathologies. Using ROC analysis a cutoff of CD68 of >12 in the most inflamed glomerulus was the most reliable predictor of TG (84.6% sensitive and 86.7% specific), even more than DSA. This value correlated strongly with DSA (p=.001), C4d staining in peritubular capillaries (PTC) (p=.001), PTC multilamellation on EM (p<.001), diffuse C4d staining in G loops (p=.005), Banff “g” scores (p<.001) and Banff “cg” scores (between 0 vs. the other scores (p<.001), indicating that macrophages correlate linearly more with activity than chronicity). CD68>12 was also associated with worse graft survival (mean follow-up 13.8 mo), even after adjusting for potential confounders including degree of interstitial fibrosis and inflammation, PTC C4d staining, TG and DSA (HR=2.16, P=0.016, 95%CI:1.2-4.0). Interestingly, after adjustment for glomerular CD68>12, DSA was no longer independently predictive of graft loss (HR:1.3, P=0.33; 95%CI:0.7-2.4). In contrast to the number of CD68+ cells, Banff g score was not significantly associated with graft failure.

Conclusions: CD68_Max >12 appears to be the most sensitive predictor of TG, even more than DSA, and independently associated with graft survival. Although multilammelation of PTC on EM of >5 layers was more specific, a combination of CD68_Max>12 and C4d in PTC had comparable specificity to EM. Thus, the number of G CD68 + cells appears to be of both diagnostic as well as pathogenetic/mechanistic importance for TG and overall graft survival.

Disclosure: All authors have declared no conflicts of interest.