In renal transplantation, non-invasive biomarkers that are predictive and allow for early intervention are limited. Here we report findings from a prospective longitudinal study of B cell cytokines in an unselected cohort of renal transplant recipients to predict rejection and outcome.

164/374 patients transplanted between dates 1/13-12/14 had serial biopsies (for-cause, and protocol at 3 and 12 mos) and blood drawn at 0, 1, 3, 6 and 12 mos. 315 biopsies and 534 PBMC samples were examined. 33% of patients had acute rejection (AR; all ACR and 3 ACR&ABMR; 19% subclinical, 14% clinical) within the first yr. Of patients with AR: 37% had early AR (≤3mos) that responded to therapy, 20% had early AR with persistence/recurrence at 12mos despite treatment, and 43% had late AR (6-12mos) despite a normal early biopsy. In this regard, the utility of 3 mos histology was limited in that it could not predict the response to treatment nor late AR.

Based on prior work, we assessed the ratio of B cell IL-10:TNFa expression (CD40L & CpG stimulation X24h) to serve as a biomarker. Patients with AR (clinical or subclinical) had a significantly lower B cell IL-10:TNFα ratio, especially within the T1 transitional subset (T1B). Using ROC analyses, an optimal cut-off value of 1.26 for T1B IL-10:TNFα ratio at 3 mos independently predicted AR at any time, with an 82% sensitivity and 86% specificity (OR 0.31, per increase in ratio by 1; p=0.0001). Moreover, T1B IL-10:TNFα ratio at 3 mos strongly predicted rejection at that time (PPV 76%, NPV 85%). Even in patients with normal 3 mos biopsy, a low T1B IL-10:TNFα ratio predicted subsequent rejection with a or PPV 82% and NPV 85% (AUC 0.86, p<0.0001). Moroever, patients treated for early AR who subsequently has persistent/recurrent AR, had a significantly lower T1B cytokine ratio at 3mos (0.58 vs. 1.52, p=0.04). Importantly, this biomarker is not confounded by “inflammation” 2[deg] infectious causes. Patients with BK or CMV viremia had significantly higher T1B IL-10:TNFa ratio compared to those with AR (low ratio). Finally, patients with a low T1B IL-10:TNFα (<1.26) ratio had significantly worse creatinine at 18 and 24 mos post-transplant and this was associated with interstitial fibrosis with inflammation (IF+i) on 12 mos biopsy (OR 0.31, p=0.003). To conclude, T1B IL-10:TNFα ratio at 3 mos appears to be a strong and a specific biomarker for the prediction and course of renal allograft rejection, and possibly long–term outcome. If confirmed, it may help inform individualized therapeutic intervention.

CITATION INFORMATION: Cherukuri A, Mehta R, Ogunde S, Hariharan S, Rothstein D. Human B Cell Cytokines Predict Renal Allograft Rejection and Early Transplant Outcomes. Am J Transplant. 2016;16 (suppl 3).

« Back to 2016 American Transplant Congress