An algorithm for cadaver kidney allocation based on a multivariate analysis of factors impacting on cadaver kidney graft survival and function

Abstract

Abstract The large imbalance between cadaver kidney supply and demand makes the implementation of equitable and effective organ allocation systems an urgent need. This has triggered a revision of the criteria used so far for cadaver kidney allocation within the North Italy Transplant program, not least in the light of the many changes that have occurred recently with respect to broader criteria for admission of patients to the waiting list, donor selection, tissue-typing methods, organ preservation and immunosuppressive protocols. We based the critical revision of our cadaver kidney allocation algorithm on univariate and multivariate analysis of a number of immunological, clinical, social and administrative factors that impacted on the transplant outcome in 2,917 patients transplanted in the 12 transplant centers operating within our organization from 1 January 1990 to 30 September 1997. This analysis indicated that younger donor age, absence of pretransplant transfusions, patient dialysis center and level of HLA match showed statistically significant positive associations with graft survival. Younger donor age and male donor gender showed a statistically significant association with excellent graft function at 4 years. The results of this analysis were used to develop a new computer-assisted version of our adult kidney allocation algorithm. It works in two steps (local pool first, then the entire waiting list) and four levels (0-1 HLA MM, PRA +; 2 HLA MM, PRA +; 0-1 MM, PRA-; 2-4 HLA MM, PRA-); within each level, selection takes into account waiting time and age difference from donor age. The evaluation of 731 transplants allocated in 19 months with the new algorithm, as against 698 transplants allocated in the preceding 19 months according to the previous algorithm, showed a significantly higher proportion of recipients who had been on the waiting list for more than 3 years (33.2% versus 22.6%). The use of the new algorithm was also associated with a significantly increased number of transplanted alloimmunized patients (18.8% versus 9.2% with the previous algorithm) and recipients with 0-1 HLA mismatches (22% versus 14.3%). Furthermore, the number of kidneys used locally has steadily increased. Differences in 6-month graft survival and percentage of patients with excellent function at 6 months were not statistically significant in recipients transplanted with the new versus the previous algorithm. Survivals were 93.7% versus 91.8%. Percentages of patients with excellent renal function were 69.9% and 71.8%, respectively. These preliminary data suggest that the new algorithm improves HLA match and reduces the number of patients on the waiting list for 3 or more years without determining significant modifications of 6-month graft survival and function. Moreover, it facilitates the achievement of a fair local balance between organs retrieved and transplanted, the compliance of operators with objective allocation rules and the documentation of the whole allocation process.