Introduction: Selection using epitope mismatches (MM) has good potential in recipients with pre-existing antibodies, and also in kidney allocation in network of transplant centres with a large common waiting list. Is the use of epitope MM advantageous compared to the use of broad antigen mismatches in a single transplant centre?
Materials: We have analysed the link between the risk of losing a transplant and epitope or broad antigen MM. The results of 789 recipients were analysed. HLA-genotyping of recipients was performed using the SSP method.
The five-year survival rate decreased as the number of broad antigen MM increased (log rank, Breslow p<0.001). Univariate analysis showed a higher number of HLA-А, -В и -DR mismatches to be associated with an increased risk of transplant loss (p=0,015; p=0,001; p<0,0001). However, multivariate adjusted Cox proportional hazard regression showed HLA-В and -DR mismatches (p=0,001; p<0,0001) to be associated with higher risk, but not the HLA-А (p=0,146). There was a linear relationship between the number of broad antigen MM and the risk of transplant loss: 1-1, 2-1,46 (p=0.523), 3-2,69 (0.012), 4-3,81 (p=0.001), 5- 5,63 (p=0.0001), 6 -  8,08 (p<0.0001). As opposed to broad antigen mismatches, increased epitope mismatches were exponentially, not linearly related to the risk of transplant loss: 5-1, 10 – 1.16 (p=0.891), 20-1.35 (p=0.02), 30-2.14 (p=0.001), 40-4.9 (p=0.0001), 50-9.9 (p<0.0001). Nevertheless, the number of epitope mismatches remains a risk factor, even when adjusted for the number of broad antigen mismatches in the multivariate analysis (p<0.0001).
Epitope mismatches allow for more accurate prediction of the risk of transplant loss: the area under the ROC curve was 0,812 [0,766; 0,858] versus 0,649 [0,59; 0,707] in broad antigen mismatches. This approach allows to identify patients at high risk of transplant loss (over 20 epitope mismatches) amongst patients, who were previously considered to have low immunological risk (i.e. 1-2 broad mismatches).
Conclusion: Even in routine practice of a single transplant centre with a small waiting list the use of epitope mismatches may be beneficial. It will allow to choose the optimal option –  the minimal number of epitope mismatches amongst a few potential recipients with the equal numbers of broad antigen mismatches. Thus, allowing to increase the transplants survival rate and reduce the risk of its loss.