3.1.1. Urine and Peripheral Blood Transcriptomics

The CTOT 04 trial has analyzed mRNA transcripts in urinary sediment cells and identified a three-gene signature (CD3ε mRNA, CXCL10 mRNA, and 18S rRNA) predicting TCMR up to 20 days before biopsy-proven diagnosis [51]. A more recent study by the same group analyzed gene expression in urinary cells and renal biopsies during AR and identified unique and shared gene signatures associated with biological pathways involved in TCMR and ABMR. Furthermore, they demonstrated the enrichment of biopsy gene signature in urinary cells and of immune cell types in urine compared with renal tissue. These findings support the hypothesis that urine gene expression patterns can reflect and even amplify ongoing renal tissue immune pathways and may help diagnose rejection and monitor its dynamics [85]. This is consistent with evidence from previous studies suggesting that graft can sort renal tissue infiltrating cells in urine as an in vivo flow cytometer [86].

Several studies have tried to identify a peripheral blood gene expression signature to diagnose subclinical AR at an early stage.

Christakoudi S et al. analyzed expression of 22 literature-based genes in peripheral blood samples of patients from Kidney Allograft Immune Biomarkers of Rejection Episodes (KALIBRE) study and identified a seven-gene TCMR-signature (IFN-γ, IP-10, ITGA4, MARCH8, RORc, SEMA7A, WDR40A) which allowed diagnosis of AR 7 weeks before renal biopsy and correlated with response to therapy [52].

Zhang W et al. focused on patients with subclinical TCMR (protocol biopsy at third month) in KTx recipients from Genomics of Chronic Allograft Rejection (GoCAR) study [87] and identified a 17-gene peripheral blood signature which characterized ongoing subclinical TCMR and predicted an increased risk of clinical TCMR at 24 months and decreased graft survival [53].

A peripheral blood mRNA assay based on eight genes (CXCL-10, FCGR1A, FCGR1B, GBP1, GBP4, IL15, KLRC1, TIMP1) was developed in a multicenter, prospective study and correlated with histological features of acute and chronic ABMR (microvascular inflammation, transplant glomerulopathy) but not of TCMR. Diagnostic accuracy was high (ROC AUC 79.9% *p* < 0.0001), even in the setting of stable graft function [54].

A blood molecular biomarker based on multiple gene expression signatures was designed to distinguish "immunological quiescence" from subclinical AR in a multicenter study (CTOT-08). This correlated with clinical (AR, renal function) and histological outcomes (IFTA) and with de novo DSA. This biomarker was validated with surveillance biopsies data and proved to be especially useful in ruling out subclinical rejection (NPP: 78–88%) [55].

In a multi-center "Assessment of Acute Rejection in renal Transplant (AART)" study, peripheral blood transcriptome analysis identified a 17-gene signature called "Kidney Solid Organ Response test" (kSORT), which predicted both TCMR and ABMR up to 3 months before histological diagnosis in an independent prospective cohort. This tool is characterized by a high accuracy in predicting AR, especially when compared with performance of other biomarkers in the same setting (AUC = 0.94; sensitivity: 83%, specificity: 90.6%; PPV: 93,2%) [57,88].

kSORT has also been used in association with an IFNγ Elispot in the ESCAPE study, resulting in a higher PPP (AUC > 0.85) for subclinical TCMR and ABMR [74].

Peripheral blood transcriptomic analysis allowed to build a classification model capable of discriminating ABMR from accommodation in ABO-incompatible kidney transplants [89].

A blood test (TruGraf v1) has been developed to study a set of microarray-based gene expression in order to discriminate patients with a stable graft and immunological quiescence ("Transplant Excellence") from those with renal dysfunction or AR. This tool was proposed as a tool to avoid unnecessary surveillance biopsies on the basis of high accuracy in detecting AR (74%) and high NPP (90%) [90]. The value of serial TruGraf testing to confirm immunoquiescence and avoid surveillance biopsies has been confirmed in a recent study [91].
