Donor Specific Antibodies are Associated with more Inflammation, More Fibrosis and Higher Expression of Rejection Associated Transcripts in Subclinically Rejected Liver Allografts

Abstract

Background and Aims Subclinical rejection (SCR) is a common event in protocol biopsies after liver transplantation and has a good medium-term prognosis even if left untreated. We recently found hints for an intrahepatic regulation of cytotoxic T cells by regulatory T cells (Treg) in SCR. However, the role of donor-specific anti-HLA antibodies (DSA) has not been investigated in this setting. Material and Methods We included all liver biopsies with SCR (rejection activity index (RAI)≥1+1+1; liver enzymes (ALT, AST, AP)≤2x upper limit of normal (ULN)) from virus negative patients in our program. Comparator cohorts consisted of biopsies with acute cellular rejection (ACR: RAI≥3; liver enzymes >2x ULN), and no histological rejection (NHR: RAI<2, normal liver enzymes). RT-PCR was performed from 77 liver biopsies with gene panels using 90 genes for rejection, endothelial cell activation, operational tolerance, T cell exhaustion and immune regulation. DSA were detected with bead assays. DSA positivity was analyzed using an MFI threshold of 1000 or more. Results and Discussion Graft gene expression in SCR biopsies broadly overlapped with the distinct expression pattern from ACR and NHR in a cluster analysis of RT-PCR results (p<0.05 and q<0.08). Thereby, SCR was characterized rather by an intermediate expression level of those genes and molecular pathways that were upregulated in ACR than by a unique set of genes or pathways. However, expression levels of GARP, a marker for activated Treg, but not of FOXP3, the Treg lineage marker, were higher in grafts with SCR compared ACR. Additionally, granzyme B, an effector molecule of cytotoxic lymphocytes, was downregulated in SCR compared to ACR, while CD8 was not significantly different in SCR and ACR. Next, the influence of DSA on the graft gene expression was evaluated, because the graft gene expression in SCR was inhomogeneous. DSAs were found in paired blood samples in 30% of SCR, 8% of NHR and 47% of ACR biopsies, while the overall DSAs frequency was 27% in our biopsy program. DSA+ SCR biopsies had significantly higher scores for lobular and portal inflammation, central perivenulitis and perisinusoidal fibrosis compared to DSA- SCR biopsies. Thereby, rejection associated transcripts were significantly overexpressed in DSA+ compared to DSA- SCR biopsies. No DSA+ liver biopsy showed a relevant C4d staining with our protocols. Nonetheless, eight SCR biopsies (10% of all SCR biopsies, 43% of DSA+ SCR biopsies) additionally fulfilled the 2016 Banff criteria for possible chronic antibody mediated rejection (cAMR). However, the biopsies numbers are still too low for prognostic evaluations. Conclusions SCR overlaps transcriptionally with NHR and ACR. Thereby, DSA seems to mark SCR biopsies with more inflammation and fibrosis. The combination of SCR in biopsies with DSA positivity might therefore identify patients with more pronounced graft inflammation, which might require a change in immunosuppression.