*Purpose: We analyzed serial gene expression profiles in paired blood and biopsies from kidney transplant recipients phenotyped with a comprehensive clinical algorithm. We compared disease driving pathways in these distinct tissue compartments and tested if gene expression profiles can distinguish kidney injury, clinical and subclinical rejection.

*Methods: Paired biopsies and blood (n=290) run on gene expression microarrays with three diagnoses -subclinical rejection (subAR – normal creatinine), clinical rejection cAR and dysfunction (ADNR) were compared to no rejection (TX). Data were analyzed using Partek Flow software that converts microarray signals into sequencing-like counts using the STAR aligner and normalized using Robust Multichip Average. Reads were quantified and mapped to hg38 gencode_38_v2 genes. Differential expression was done using ANOVA with a false discovery rate <10%.

*Results: In the biopsies, there was a striking overlap between subAR and cAR (95%), and ADNR and cAR genes (95%) but no overlap between subAR and ADNR (12%) All overlapping genes showed higher expression in cAR compared to subAR and ADNR. Pathway analysis revealed similar pathways in both subAR and cAR with overexpression of immune/inflammatory pathways, which was distinct from ADNR, where most perturbations were in cell signaling pathways. In the blood a different pattern was observed, with the overlap in subAR and cAR being lower (40%) thatn in biopsies and more shared genes between subAR and ADNR. Interestingly, although pathway analysis did not show clear immune/inflammatory mechanistic pathways in the blood, the mitochondrial dysfunction and oxidative phosphorylation pathways were dysregulated in subAR and ADNR. Blood and biopsy comparisons, expectedly showed little overlap in the differentially expressed genes (6% for subAR vs TX and 14% in cAR vs TX). Interestingly, biopsies misclassified using a commercially available blood gene expression test show distinct gene expression profiles that were consistent with correctly classified samples.

*Conclusions: We describe the first large scale study comparing paired, serial biopsy and blood gene expression profiles especially in subclinical kidney rejection. Our findings show a distinct molecular profile for subAR, specifically that it is a milder manifestation of acute rejection, and perhaps a reason for evasion of detection until it manifests as cAR with a concomitant rise in creatinine. Our results also show that acute dysfunction is molecularly different from rejection, since ADNR genes shared with cAR were genes associated with injury and not rejection. In contrast, genes shared by subAR and cAR were immune and rejection related. Additionally, misclassified samples with discordant molecular profiles suggests revaluation of clinical phenotyping by incorporating molecular data and not based solely on histology.

« Back to 2022 American Transplant Congress