Background: Peritubular capillaritis (ptc), reported by the ptc score, is a major feature of kidney allograft rejection and microvascular inflammation (MVI). MVI sum scores (ptc + glomerulitis score ≥2) are accepted diagnostic surrogates of human leucocyte antigen (HLA)-antibody interaction. However, low-grade inflammation is common and ptc scores (number of leucocytes/capillary) may not mirror all aspects of ptc morphology. Recently we observed a relationship of the diffuse extent of ptc (inflammation of >50% of the renal cortex) with graft loss and significantly higher donor-specific antibody levels, suggesting potential inclusion of diffuse ptc as an additional surrogate of antibody-antigen interaction.
Methods: We sought to assess how a combination of ptc score and extent in low-grade inflammation (ptc1) affects transplant glomerulopathy (TG) and graft loss risk. Patients (n = 616) were assessed for MVI in first indication biopsies. Cases with a ptc score of 1 but diffuse extent (ptc1diffuse, g-score = 0, n = 26) were considered additional surrogates of HLA-antibody interaction and compared with MVI ≥2 and MVI <2.
Results: The ptc1diffuse and MVI score ≥2 subjects had worse graft survival (42% and 59%) compared with an MVI score <2 (70%) (P = 0.002). The incorporation of ptc1diffuse in the MVI score ≥2 increased the receiver operating characteristics curve for TG [area under the curve (AUC) 0.602; P = 0.008] compared with a Banff MVI score ≥2 (AUC 0.56; P = 0.12); cases with baseline TG were excluded. In multivariate analysis, ptc1diffuse remained independently related to TG (odds ratio 3.89; P = 0.008) and graft loss (hazard ratio 2.64; P = 0.001) even after inclusion of all rejection episodes.
Conclusion: An integrated view of ptc morphology including diffuse ptc in MVI is superior for TG and graft loss risk assessment.