Methods.

All primary, solitary renal transplants registered at the USRDS between January 1995 and June 1997, on initial immunosuppression that included either MMF or AZA were followed until June 1998. Primary study end points were graft and patient survival. Kaplan-Meier analysis was performed to compare AZA vs. MMF treated patients by HLA mismatch. Cox proportional hazard models were used to investigate the interaction between HLA mismatch and AZA versus MMF therapy on the study endpoints. All multivariate analyses were corrected for 13 potential confounding pretransplant variables including intention to treat immunosuppression.

Results.

A total of 19,675 patients were analyzed (8,459 on MMF and 11,216 on AZA). Overall three year graft survival was higher in the MMF group when compared to the AZA group (87% vs. 84% respectively P< 0.001). For both AZA and MMF three-year graft survival improved with fewer HLA donor-recipient mis-matches. Comparing zero antigen mismatches to six antigen mismatches, the relative improvement was comparable for both patients on AZA (92.4% vs. 80.6%) and MMF (95.2% vs. 82.9%). By Cox proportional hazard model the relative risk for graft loss decreased significantly in both the AZA and MMF treated patients with increased HLA matching.

Conclusion.

The use of MMF does not obviate the benefits of HLA matching, while HLA matching does not minimize the benefits of MMF on long term graft survival. Our study would suggest that HLA matching and MMF therapy are additive factors in decreasing the risk for renal allograft loss.