OR35 Haploidentical matched donor or unrelated mismatched?

Abstract

Blood from a 64 year-old Caucasian male patient was HLA typed using NGS. He was determined to be homozygous for A*02:01, and this was confirmed on SSO. Buccal swab testing of the patient on SSO and SBT also confirmed the A locus homozygosity. The patient’s extracted DNA from the original blood sample was also run on high resolution SSP, confirming he was a homozygote for A*02:01. Three of the family member’s haplotypes were extrapolated by a haploidentical sibling’s HLA typing result. A blood sample from the patient’s CMV-matched, ABO-matched 29-year old son was a haploidentical match for all of the HLA loci except for the A locus. The son was a homozygote for A*03:01, missing the paternal A allele. We suspected the difference was due to a crossover phenomenon, but we could not reliably determine that due to the limited number of family members tested. The other probability was the patient had an allelic dropout. A week later, we tested the 26-year old daughter who was typed as identical to her brother. To determine the cause of the paternal missing allele in the children, we typed the patient’s second sibling. Though typing the additional family member established all 4 haplotypes in the patient’s family, the A*03:01 was not present. The hospital confirmed all family members’ biological relationship and confirmed the patient did not have a prior transplant. One more blood sample from the patient was submitted and run on all platforms, obtaining the same result. At this time, we had ruled out technology/manufacturer-related allelic dropout, crossover phenomenon, and hematological malignancy driven LOH. Although NGS produced a sequence for the entire gene of the A locus, we wondered if there could be unrecognized (blank) allele that accounted for the patient’s and children’s homozygosity? Was there possibly a germ-line mutation in a primer binding region? To ensure the two children were suitable haplo-donors for the patient, we decided it was in the patient’s best interest to run STR to confirm biological relationships. Our reasoning was that if related, the two children would be the most suitable donors considering the blank allele would be present in all three. Eleven out of 21 polymorphic genes throughout the genome excluded the patient as the biological father leading us to conclude that the two children were mismatched unrelated donors.