A Pre-transplant Blood-based Lipid Signature for Prediction of Antibody-mediated Rejection in Kidney Transplant Patients

Abstract

There is a lack of biomarkers for pre-kidney transplant immune risk stratification to avoid over- or under-immunosuppression, despite substantial advances in kidney transplant management. Since the circulating lipidome is integrally involved in various inflammatory process and pathophysiology of several immune response, we hypothesized that the lipidome may provide biomarkers that are helpful in the prediction of kidney rejection. Serial plasma samples collected over 1-year post-kidney transplant from a prospective, observational cohort of 45 adult Kidney Transplant [antibody-mediated rejection (AMR)=16; stable controls (SC) =29] patients, were assayed for 210 unique lipid metabolites by quantitative mass spectrometry. A stepwise regularized linear discriminant analysis (RLDA) was used to generate models of predictors of rejection and multivariate statistics was used to identify metabolic group differences. The RLDA models include lipids as well as of calculated panel reactive antibody (cPRA) and presence of significant donor-specific antibody (DSA) at the time of transplant. Analysis of lipids on day of transplant (T1) samples revealed a 7-lipid classifier (lysophosphatidylethanolamine and phosphatidylcholine species) which discriminated between AMR and SC with a misclassification rate of 8.9% [AUC = 0.95 (95% CI = 0.84-0.98), R2 = 0.63]. A clinical model using cPRA and DSA was inferior and produced a misclassification rate of 15.6% [AUC = 0.82 (95% CI = 0.69-0.93), R2 = 0.41]. A stepwise combined model using 4 lipid classifiers and DSA improved the AUC further to 0.98 (95% CI = 0.89-1.0, R2 = 0.83) with a misclassification of only 2.2%. Specific classes of lipids were lower in AMR compared with SC. Serial analysis of SC patients demonstrated metabolic changes between T1 and 6 months (T2) post-transplant, but not between 6 and 12 (T3) months post-transplant. There were no overtime changes in AMR patients. Analysis of SC T1 vs AMR T3 (that at time of AMR) showed sustained decreased levels of lipids in AMR at the time of rejection. These findings suggest that lack of anti-inflammatory polyunsaturated phospholipids differentiate SC from AMR pre-transplant and at the time of rejection, and a composite model using a 4-lipid classifier along with DSA could be used for prediction of antibody-mediated rejection before transplant. HighlightsO_LIDespite significant advancements in kidney transplant treatment and intensive clinical follow-up monitoring, all rejection events are unlikely to be recognized at the beginning. As a result, efforts have been made to identify new biomarkers for kidney rejection detection. C_LIO_LIWhile lipids are known to be potent mediators of inflammation, pro-resolving processes, and other cell signaling cascades, lipidomics can be applied to identify reliable biomarkers to monitor disease severity and may also allow prediction of kidney rejection. C_LIO_LIOur lipidomic study shows lipid profile changes between antibody-mediated rejection group and stable control group as a function of different time point, pre and post-kidney transplantation. Furthermore, our study demonstrates that combining lipid and clinical parameters allow prediction of rejection on the day of the transplant. C_LIO_LIThese findings have the potential to change the present paradigm of pre and post-transplant monitoring and management of these patients by implementing an evidence-based risk stratification technique, resulting in a substantial improvement in kidney transplant success. C_LI