Background: Kidney transplantation is the best treatment option for end stage kidney failure. However, long term management for a prolonged life of these transplanted organs is still a challenge because of lack of efficient diagnostic biomarkers. In this study, we analyzed sera collected from transplant patients to identify blood protein markers that are specific to transplant associated injury.

Method: Serum samples (n=121) from renal transplant patients enrolled at Lucile Packard Children's Hospital at Stanford University which included 111 transplant (AR; n=27), calcineurin inhibitor toxicity (CNIT; n=20), chronic allograft nephropathy (CAN; n=25), BK nephropathy (BK; 14) and normal graft function without significant pathology (STA; n=25) and 10 healthy normal controls (HC). 20 most abundant serum proteins were depleted from serum samples. The resulting proteins from each sample was subsequently subjected to trypsin digestion and were analyzed by a LTQ Orbitrap Velos mass spectrometer. The IPI human database was searched, using a 50 ppm mass window on the precursor ion. We identified 137 common serum proteins in the samples analyzed with cumulative spectral counts for each protein. In order to identify transplant injury protein panels, we fitted 100 Elastic Net logistic regression models to all the proteins using bootstrap samples to classify injury phenotypes. For each bootstrap a nested cross-validation loop estimated the best value for penalized parameter. Multivariate analysis was performed for different clinical confounders.

Result: This study identified a panel of 10 proteins that could segregate samples with biopsy confirmed graft injury (AR, BKVN, CAN, CNIT) from samples without injury (STA, HC) with a classification accuracy of 92%. Five of these 10 proteins could further segregate samples with acute graft dysfunction (delta CrCL>20% over baseline) from the other injury phenotypes that included chronic graft injury with a classification accuracy of 98% using penalized logistic regression modeling. Importantly, a panel of 10 proteins was able to differentiate AR from BKVN with a PPV 85% and an NPV 94%. Multivariate analysis for different clinical confounders confirmed the high specificity and sensitivity of this proteomic panel differentiating the most common etiologies of biopsy-confirmed graft injury, specifically AR, CAN, CNIT and BKVN.

Conclusion: By utilizing mass spectrometry based proteomics we have identified novel diagnostic serum protein panels for transplant injury monitoring.

« Back to 2013 American Transplant Congress