*Purpose: Antibody-mediated rejection (ABMR) is one of the most significant contributors to late graft loss. Hallmark characteristics include peritubular capillaritis (ptc) and glomerulitis (g). Diffuse ptc (>50%) was associated with worse graft survival independent from the ptc score but the current ptc threshold is arbitrarily defined and may not reflect pathophysiological phenotypes accurately enough. The Banff ptc working group has been re-established to study the diagnostic and prognostic relevance of ptc extents in different scenarios. We hypothesize that the assessment of chronic ABMR specimens using Nanostring NCounter based gene expression analysis allows to define novel thresholds of ptc extent, reflecting molecular ABMR phenotypes more accurately.

*Methods: We retrospectively analyzed 44 biopsies with ABMR/chronic ABMR (DSA⁺). Peritubular capillaritis was re-evaluated by an experienced external nephropathologist (M.M.) including the ptc score and ptc extent (focal ptc: 10-50% capillaries, diffuse ptc: >50%). Nanostring nCounter Gene expression analysis was performed with a customized gene set corresponding to the recommendations of molecular ABMR phenotypes in Banff 2017 guidelines. To test the correlation of gene expression levels and histological scores, gene expression levels were divided into quartiles. Gene expressions above the first quartile (ABMR^Q>1) were considered as positive values for ROC analysis.

*Results: Ptc was categorized as no ptc in 13, focal ptc in 23, diffuse ptc in 7 biopsies [ptc extent 25/0-40% (median/IQR)]. ABMR was diagnosed in 27 (67.5%) biopsies, mixed rejection in five (12.5%) and borderline TCMR in three (7.5%). In biopsies with diffuse ptc significant higher gene expressions were found in the ABMR gene set [63/55-83 vs. 32/27-52; (median/IQR); p=0.012], the ABMR exhaust gene set (390/245-609 vs. 245/128-358; p=0.022), the Eculizumab gene set (180/143-339 vs. 65/57-133; p=0.0027) and the TCMR gene set (48/40-75 vs. 25/19-35; p=0.001). Sensitivity analysis revealed improved AUCs for predicting biopsies with ABMR gene expressions over the 1^st quartile with a cut-off of 35% compared to the 50% cut-off [ptc^>35: AUC 0.76/0.61-0.90, p=0.013; ptc^>50: AUC 0.71/0.54-0.88, p=0.039].

*Conclusions: With the Nanostring platform we were able to identify a new threshold of ptc extent. The proposed cut off >35% may reflect molecular phenotypes of ABMR more accurate than the current one and could improve early diagnosis of ABMR.

« Back to 2022 American Transplant Congress