Capillary C4d deposition in kidney allografts: a specific marker of alloantibody-dependent graft injury.

Capillary deposition of the complement split product C4d has been discussed as a marker for antibody-mediated kidney allograft rejection. The relationship between C4d staining and posttransplant alloantibody detection remains to be thoroughly investigated, however. In this study, C4d staining in peritubular capillaries (PTC) and the incidence of alloantibody formation, as detected with sensitive techniques, were evaluated among a cohort of transplant recipients who had undergone biopsies and had not been selected for a specific histologic diagnosis. One hundred thirteen biopsies, obtained from 58 cadaveric kidney transplant recipients, were tested. Serum samples obtained at the time of biopsy were evaluated by flow cytometric crossmatch (FCXM) testing and FlowPRA (One Lambda, Inc., Canoga Park, CA) analysis of anti-HLA panel reactivity. Most biopsies with C4d deposits in PTC (C4d(PTC)(+), n = 21 of 24) were associated with positive posttransplant FCXM results (T and/or B cell FCXM) and/or > or =5% FlowPRA (anti-HLA class I and/or II) reactivity. Approximately 50% of the C4d(PTC)(-) biopsies were observed to be associated with donor-specific alloantibodies. Accordingly, high specificity (93%) but low sensitivity (31%) were calculated for capillary C4d staining (with FCXM testing as the standard method). For clinical evaluation, three patient groups were defined, i.e., a group of recipients with positive C4d staining in at least one allograft biopsy (C4d(PTC)(+), n = 16) and two C4d(PTC)(-) groups, which were discriminated on the basis of posttransplant FCXM results as C4d(PTC)(-)/FCXM(+) (n = 22) and C4d(PTC)(-)/FCXM(-) (n = 20) groups. Univariate analyses revealed significant differences between these groups with respect to serum creatinine levels at 12 mo [median, 2.83 mg/dl (interquartile range, 1.93 to 4.2 mg/dl) versus 1.78 mg/dl (1.47 to 2.24 mg/dl) versus 1.59 mg/dl (1.2 to 1.71 mg/dl), P < 0.001]. Of the five immunologic graft losses, four occurred in the C4d(PTC)(+) group and one occurred in the C4d(PTC)(-)/FCXM(+) group. In a multivariate analysis, C4d positivity was observed to have an independent predictive value for inferior 12-mo graft function (P = 0.02), whereas the observed moderate difference between C4d(PTC)(-)/FCXM(+) and C4d(PTC)(-)/FCXM(-) recipients did not achieve significance. In conclusion, these data demonstrate that positive C4d staining, which is an independent predictor of kidney graft dysfunction, represents a reliable specific marker for antibody-dependent graft injury.