FC 117: Clinical Validation of Automated Urinary Chemokine Assays for Non-Invasive Detection of Kidney Transplant Rejection: A Large Prospective Cohort Study

Abstract

Abstract BACKGROUND AND AIMS The urinary chemokines CXCL9 and CXCL10 have been extensively studied as biomarkers for kidney transplant rejection. However, final validation in a sufficiently large, real-world prospective cohort is needed to determine the relation to rejection subtypes and clinical confounders, the context of use and the added benefit for clinical practice. METHOD In this single-centre prospective cohort study, we analysed 1559 biopsy-paired urine samples at both indication and protocol time points from 622 kidney transplants performed between April 2013 and July 2019. We quantified urinary CXCL9 and CXCL10 using automated immunoassays and normalized the values to urinary creatinine. In addition to evaluating diagnostic performance for Banff-grade rejection, we also considered rejection as assessed by the molecular microscope. RESULTS The urinary chemokines strongly associated with the inflammation severity in the biopsy, when considering classical semi-quantitative histological scores, but also when considering molecular rejection scores. The diagnostic performance of urinary chemokines alone was moderate for Banff acute rejection, and increased when considering treated rejection or molecular rejection instead (Figure 1). Taking into account confounding factors rather than excluding them, an integrated model of urinary chemokines with eGFR, donor-specific antibodies, time post-transplant and polyoma viremia had a high diagnostic value for acute rejection (N = 150) [ROCAUC 82.6%, 95% confidence interval (95% CI) 78.9–86.2] and added value on top of clinical standard-of-care. The integrated score would help to safely avoid 59 protocol biopsies per 100 patients when the predicted rejection risk is less than 10%. CONCLUSION The integration of urinary chemokines and clinical markers allows for personalized noninvasive monitoring of rejection probability and enables to importantly and safely reduce the number of unnecessary biopsies. We expect this study to pave the way for the final implementation of urinary chemokine measurements as noninvasive markers of rejection in routine clinical follow-up after kidney transplantation.