*Purpose: Surveillance renal biopsy data indicate that up to 36% of clinically stable kidney allograft recipients have subclinical rejection, emphasizing the need to implement automated and sensitive technology to bring biomarkers into clinical practice. In this study, we used a novel platform to measure established urine biomarkers (CXCL9, CXCL10, CCL2 and VEGFA) simultaneously with a high level of precision, and test the hypothesis that patterns of expression serve as a new gold standard to identify stable patients post transplant.

*Methods: We collected urine from a total of 515 adult and pediatric renal allograft recipients who underwent surveillance or indication biopsy at four sites. Biomarkers were measured on the SimplePlex platform (ProteinSimple Inc.), and levels were compared in patients with normal, borderline or acute rejection by histology. Principal component (PC) analysis, non-parametric rank analysis and ROC curve analyses were used to analyze biomarker performance.

*Results: Of the 515 test samples, biopsy histology was reported as normal (n=320), acute rejection (n=102) and borderline rejection (n=92). Unsupervised PC analysis of all 4 biomarkers revealed heterogeneity among each diagnostic group (PC1=75%), which was attributed to changes in CCL2 and CXCL10. Moreover, there was clustering of normal and borderline diagnoses while the greatest degree of heterogeneity was within patients with acute rejection. Using the rank sum test, we found that each biomarker independently discriminated normal vs. acute rejection (P<0.00001), whereas only CCL2 trended higher in patients with borderline rejection (P=0.03). Of all biomarkers, CXCL9 and CXCL10 provide the best diagnostic performance for the diagnosis of acute rejection (AUC 0.79 and 0.84, respectively). Based on a Bayes analysis and assuming a 20% incidence of acute rejection, the negative predictive value of CXCL10 was 88%. To evaluate the significance of our findings, we subsequently tested 79 samples in an additional cohort of clinically stable patients and identified 67% as stable vs. 33% with concern for subclinical disease using biomarker assays alone.

*Conclusions: This study represents one of the largest cohort analyses of multianalyte urinary biomarker profiling following renal transplantation. It suggests that automated monitoring of urinary biomarkers accurately classifies stable patients vs. those with active intragraft inflammation. Sensitive multianalyte assays have potential to outperform surveillance biopsy for the monitoring of at-risk patients and may serve as a precision tool to identify eligibility for minimization or pro-tolerogenic regimens.

« Back to 2020 American Transplant Congress