Background. Detection of C4d along peritubular capillaries (PTC) in renal allograft biopsies is an independent prognostic marker of poor long-term graft survival. It is typically associated with circulating donor-specific antibodies. Since only little information is available on the best technique to stain C4d, we compared the two methods most often used for detecting C4d in renal allograft specimens.

Methods. We investigated the expression of C4d along PTC in 64 renal allograft biopsies using a monoclonal antibody (Quidel) and immunofluorescence for frozen (F-IF) and a polyclonal antibody (Biomedica) and immunohistochemistry for formalin-fixed and paraffin-embedded (P-IHC) tissue samples. We compared the staining extent (diffuse, focal, minimal, no staining) in frozen and paraffin sections and evaluated the intra- and inter-observer concordance rates using kappa statistics. In addition, we determined the inter-observer concordance in 240 paraffin-embedded biopsies of a multi-centre study.

Results. The inter- and intra-investigator concordance rate (κ = 0.9) of analysing the C4d expression by F-IF was excellent. In contrast, the detection of C4d by P-IHC demonstrated a substantially lower prevalence and extent of C4d expression with a lower intra- and inter-observer concordance rate (κ = 0.3). Only 69% of diffuse and 13% of focal C4d-expressing cases were in line classified by F-IF and P-IHC. On average, the estimated area of C4d-positive PTC in the diffuse group was 36% lower by P-IHC than by F-IF. The inter-observer concordance rate in paraffin of the 64 renal biopsies and the multi-centre study was good, but not perfect (κ = 0.57 or 0.67).

Conclusions. C4d staining determined on frozen tissue samples using F-IF with a monoclonal antibody appears to be better suited for diagnostic as well as research purposes. Future studies should correlate C4d staining patterns with circulating donor-specific antibodies.