*Purpose: Molecular and animal studies have identified NK cells as important mediators and markers of antibody mediated rejection. Using three color immunohistochemistry for CD16, T-bet, and CD34, we have previously identified intravascular mature NK cells in FFPE renal allograft biopsies. In indication renal allograft biopsies from our transplant center, CD16+/T-bet+ NK cells are markers of antibody mediated rejection and correlate with graft failure. We asked whether CD16+/T-bet+ NK cells also predicted outcome in surveillance biopsies at 1-year after transplantation in grafts with stable function.

*Methods: 241 surveillance renal biopsies from 2011-15 were simultaneously immunohistochemically stained for CD16 (brown, membrane), T-bet (blue, nuclear), and CD34 (red), and slides were digitized. Activated NK cells (blue+brown) were identified within peritubular and glomerular capillaries (red), quantified, and normalized to cortical and glomerular cross-sectional area. Medical records were reviewed for demographic and laboratory data, including graft function. The primary endpoint was graft failure measured by return to dialysis or retransplantation, with censoring for loss of follow-up and death with functioning graft. A secondary endpoint was the change in eGFR (MDRD4 study equation) per year from biopsy to censoring.

*Results: 11 (4.6%) grafts failed, with a median follow-up time of 4.8 years after biopsy. 17 recipients died with functioning grafts, and 6 recipients transferred away. 32 grafts had elevated peritubular capillary NK cells (>10 per mm²), of which 2 (6.3%) failed, compared with 9 failure from 209 grafts (4.3%) without elevated NK cells (p=0.644, Fisher’s exact). Biopsies segregated by glomerular NK cell levels showed similar results. Surviving grafts with elevated NK cells at 1 year had an average eGFR trend of – 0.2 ml/min/year, compared to -1.6 ml/min/year in surviving grafts without elevated NK cells (p=0.48, t-test).

*Conclusions: Unlike indication biopsies, CD16+/T-bet+ NK cells did not predict graft failure or decreased function in 1-year surveillance biopsies. The rate of graft failure in the study cohort was low, raising the possibility of selection bias in the study cohort. Furthermore, the median time to initial de novo alloantibody detection at our transplant center is 2-3 years after transplant. The development of antibody mediated rejection subsequent to the biopsies studied in this cohort likely masked any effects on survival of subclinical antibody rejection at 1 year after transplant.

« Back to 2019 American Transplant Congress