Molecular rejection signals in chronic-active T cell–mediated rejection: Histological associations and potential clinical implications

Chronic-active T cell-mediated rejection (caTCMR) remains debated, and molecular diagnostics might aid risk-stratification. This retrospective two-center observational study analyzed 26 kidney allograft biopsies with caTCMR, comparing them to 40 borderline T cell-mediated rejection (bTCMR)/acute T cell-mediated rejection (aTCMR) cases (without caTCMR) and 308 subthreshold/negative controls after excluding microvascular inflammation at/above threshold and overlapping pathologies. All biopsies received transcriptomic evaluation through microarray-based gene expression profiling. caTCMR cases with aTCMR (n = 8) showed higher molecular T cell-mediated rejection (TCMR) activity (median probability of molecular TCMR [TCMR_prob] 0.54 [0.30-0.83]) than "pure" caTCMR (n = 11; TCMR_prob 0.03 [0.01-0.28], P = .012) or caTCMR with bTCMR (n = 7; TCMR_prob 0.01 [0.01-0.77], P = .036). TCMR_prob was low in subthreshold/negative controls but elevated in aTCMR, BK-virus nephropathy and pyelonephritis cases (side cohort). Molecular sign-outs classified 4 of 11 (36%) "pure" caTCMR cases as molecular TCMR. Within the TCMR continuum (26 caTCMR plus 40 bTCMR/aTCMR), interstitial inflammation, tubulitis, and total inflammation-lesions were associated in univariable analyses with TCMR_prob >0.2, while interstitial inflammation in areas of interstitial fibrosis and tubular atrophy/tubulitis in areas of interstitial fibrosis and tubular atrophy correlated with molecular chronicity. Seven of 26 (27%) caTCMR and 7 of 40 (18%) bTCMR/aTCMR cases showed mixed molecular rejection activity above thresholds. In conclusion, significant molecular TCMR activity was present in a subset of caTCMR cases and was associated with higher inflammation/tubulitis/total inflammation scores and, at times, molecular antibody-mediated rejection signals, even with microvascular inflammation <2. To confirm these preliminary observations, future studies and external validation are needed.