A New Score to Determine the Probability of Finding an HLA Identical Unrelated Donor: A Validated Efficient Time and Cost Saving Method

Abstract

AbstractAbstract 604 IntroductionWe recently demonstrated that a delayed time to find an unrelated HLA compatible donor and also to proceed to allogeneic-HSCT can worsen the transplantation outcome (Michallet et al. ASH 2010). The French transplant registry has recently developed a software called “Easy Match” that predicts the number of HLA compatible donors for a given patient. The goal of our study is to validate this new program comparatively to a known score method (Tiercy et al. BMT 2007, 40; 515–522) and to improve the cost and search delay. AimsTo accelerate the process of donor search using the results of EasyMatch program and define a new score for donor finding probability, in order to be time- and cost-efficient. Material and methodsFor the first step, we retrospectively analyzed 217 patients transplanted between 2009 and 2011 after finding an unrelated donor (identical or not) or cord blood units. We used the EasyMatch software which realized a “qualitative” analysis that consisted on checking that each HLA recipient phenotype was found among all possible pair wise combinations of 2 haplotypes of the different sets of haplotypes. Various “quantitative” analyses calculated the likelihood associated to each recipient phenotype for a given set of HLA genes, in a given population, at low versus high resolution typing. The EasyMatch software gave for each patient a number of potential donors sharing the same phenotype as the patient. We validated the probability given by the EasyMatch program with the known published score method.For the second step, we used the new defined score prospectively for 62 new patients, directing the search in donor or cord blood files as stipulated by the probability to find a suitable 10/10 donor. We, therefore, analysed the number of complementary typing asked for each group of patients (before and after new score), and the delay between registration of the patient and identification of the donor or cord blood. ResultsOur 217 patients were classified in 3 different categories according to the combined results of the EasyMatch software and the HLA score (Tiercy):- Score 0 with low probability to find a suitable 10/10 donor (96% of transplantations performed with a non 10/10 donor in our study). The choice of the source will be defined considering the HLA characteristics of the recipient; in case of class I rare allele or rare HLA-BC linkage disequilibrium, a cord blood unit will be easier and more rapidly available. A complementary help should be given by an associated analysis with 4-digit haplotypes using the HaploStats program (http://www.haplostats.org/home.do).- Score 1 with a median probability to identify a suitable 10/10 donor (67% of transplantations performed with a non 10/10 donor in our study).- Score 2 with high probability to find a suitable 10/10 donor (95% of transplantations with a 10/10 donor in our study).The number of supplementary HLA typing and the delay to identify a suitable donor or cord blood was significantly reduced when using the new score, in each of the three categories (table).The EasyMatch program provides us (1) easy tracking of mismatches (2) estimation of the number of potential donors (3) selection of population following ethnic origin of patients and a high prediction when number of potential donors is higher than 5 or <1. Its use could be improved when coupled with a second scoring system.In conclusion, the use of this new scoring system allows time and cost spare. In case of low chance to find a donor, physicians can switch to searching for unrelated HLA mismatched donor or a CBU, or to use another alternative treatment in order to keep an optimal result.TableScoreBefore scoreAfter scorep0Median nb of supplementary typingn=75n=31<0.001Delay (days)7.86.5144371Median nb of supplementary typingn=39n=120.003Delay (days)7.24.7561282Median nb of supplementary typingn=103n=190.05Delay (days)2.51.744630 Disclosures:Nicolini:Novartis, Bristol Myers-Squibb, Pfizer, ARIAD, and Teva: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.