P020 Evaluation of 5-color multiplex short tandem repeat method for chimerism/engraftment study

Abstract

Aim Performance of multiplex STR method (5-color) was evaluated for (1) number of informative loci that that discriminate donor and recipient alleles in related and unrelated stem cell transplant settings; (2) accuracy of %Donor calculated from each single locus or an average from 5 informative loci. Methods PowerPlex Fusion STR kit (Promega) amplifies 24 polymorphic DNA markers (22 autosomal and 2 X/Y-markers) per reaction. Fragment analysis used 3130xl Genetic Analyzer and GeneMapper software. STR markers, Type 1 (no alleles shared between Donor and Recipient) and Type 2 (only one allele shared) were selected as informative markers for this study, which are likely to provide more reliable information to calculate % donor. We tested three types of donor-recipient pairs (unrelated N = 22, siblings N = 10, and ASHI/CAP surveys N = 4). Numbers of available informative loci between three types of donor-recipient pairs were compared by t -test. %Donor determined by a single locus vs. %Donor determined by average from five loci were compared for accuracy, using peer means of mixed chimerism samples ( N = 4) as a gold standard from the ASHI survey 2015 EMO-1. Results Number of informative loci (Type 1 and 2) was significantly smaller in sibling pairs (mean: 8.89; range: 5–11) than unrelated pairs (mean: 17.34; range: 14–21) and ASHI/CAP survey pairs (mean: 16.0; range: 11–20) ( p p = 0.0003, respectively). For 2015 ASHI EMO-1 survey, we found 18 informative loci. Discrepancy from the mean of ASHI peer results was significantly smaller when %Donor was determined as an average from 5 loci (mean: 0.30%; range: 0–1.1%), compared to the discrepancy when %Donor was determined by single locus (mean: 3.73%; range: 0–11.7%) ( p = 0.015). Conclusions Suitable informative loci are harder to find for sibling donor-recipient pairs. However, testing of 22 autosomal STR markers provided minimum of 5 informative loci (Type 1 and 2) for all donor and recipient types. Availability of multiple informative loci allows calculating average %Donor, which improves accuracy of chimerism/engraftment study compared to a %Donor from a single locus.