The dynamics of pre-existing anti-HLA antibodies and the results of kidney transplantation

Abstract

Aim. To analyze the relationship between the peak panel-reactive antibodies (peak-PRA), the value at the time of transplantation (Tx-PRA) and the results of kidney transplantation.Materials and methods. The study included 287 patients from the waiting list  with anti-HLA antibodies of I and/or II classes. 142 patients underwent transplantation of a cadaveric kidney. All patients received standard  immunosuppression: a calcineurin inhibitor, mycophenolate and steroids.  Desensitization in the preoperative period was carried out in 11 patients.  Screening and identification of antibodies was performed using multiplex technology on Luminex platform.Results. The median PRA was 47% (interquartile range – 29%; 65%).  Depending on the dynamics of PRA, we identified several groups of patients:  stable PRA (83 patients), increasing (77 patients) or decreasing value (96  patients), or variable dynamics (31 patients). The change in PRA was  accompanied by a change in the mean fluorescence intensity (r = 0.787, r2 =  0.59, p < 0.0001). In the univariate analysis, each 5% of peak-PRA and Tx-PRA  increased the relative risk (RR) of humoral graft rejection (1.09 (95%CI 1.06;  1.17), p < 0.001; 1.17 (95%CI 1.09; 1.26), p < 0.001 respectively), and ΔPRA  decreased a RR (0.932 (95% CI 0.861; 0.967), p = 0.009). In multivariate analysis (adjusted for sex and age of recipient, duration of dialysis, number of  HLA mismatches), we observed a similar scenario: peak-PRA 1.14 (95% CI 1.07; 1.19), p < 0.001; Tx-PRA 1.13 (95% CI 1.09; 1.22), p < 0.001; ΔPRA 0.949  (95% CI 0.871; 0.981), p = 0.017. In the univariate analysis increases in peak- PRA and Tx-PRA increased aRR of graft loss (1.1 (95% CI 1.05; 1.14), p <  0.001; 1.09 (95% CI 1.05; 1.15), p < 0.001 respectively), and increase in ΔPRA  decreased RR (0.952 (95% CI 0.891; 0.97), p = 0.011). In the adjusted  multivariate model, Tx-PRA did not increasea RR of graft loss (1.04 (95% CI,  0.95; 1.1), p = 0.098), while peak-PRA and ΔPRA remained significant factors  (1.1 (95% CI, 1.17; 1.24), p < 0.001; 0.931 (95% CI, 0.855; 0.954), p =  0.007, respectively). Conclusion. In the selection donor-recipient pair it is necessary to take into account the spectrum of antibodies at the point of peak values of PRA. A  decrease in PRA may hide antibodies that have a specificity to donor antigens or on certain epitope.