35-P

Abstract

Aim Innovative single and combined marrow and solid organ transplants used to treat patients with complex disorders and diseases are now being performed by transplant physicians. This study evaluated a method for calculating the probability of finding a suitably HLA antigen-matched deceased donor for a patient being considered for such an organ transplant. Methods A Monte Carlo sampling method was developed to calculate all possible matches for a three locus HLA system (HLA-A, B and DRB1) and estimate the probability of any individual finding a N/6 HLA antigen match (N = 0, 1, … , 6) in the general population. The study began with an existing data set of 250,366 high-resolution HLA haplotypes derived from the HLA typing data of the 21 populations contained in the NMDP registry. We converted each high-resolution haplotype to an antigen equivalent haplotype, thereby creating a reduced data set of 69,750 haplotypes. By selecting an extremely large set of genotype pairs at random, we estimated the number of N/6 matches for each of the 21 populations represented in the NMDP registry. Results The bioinformatics strategy utilized in this study allowed us to produce estimates of N/6 match rates for all 21 populations represented in the NMDP registry. We extended the application of this population-based approach and HLA match rate findings by creating an NMDP web-based tool that would allow physicians to input the HLA typing and race of their patient and calculate the N/6 match rate estimates specifically related to their patient. Conclusions The estimation of N/6 HLA match rates reported in this study will be of particular interest to HLA population scientists and transplant physicians. Availability of the bioinformatics tool will provide much needed support for transplant physicians as they consider the benefits and the risks of a transplant with an N/6 matched donor, set the patient expectations regarding transplant waiting time and outcomes, and determine if alternative treatments should be considered.