C∗08:02/16:02 versus C∗05:01/16:01, ambiguous pairs particularly susceptible to typing error

Abstract

We present four individuals initially determined to be C∗08:02:01:01 & 16:02:01 but subsequently typed as C∗05:01:01:01 & 16:01:01. These consist of a patient typed in our laboratory as well as three unrelated donors (URDs). The URDs had initial typing performed at National Marrow Donor Program (NMDP) laboratories and confirmatory typing performed in our lab. Typing for our patient was performed by sequence specific oligonucleotide probe (SSOP, One Lambda) and sequence based typing (SBT, Conexio) techniques. SBT included both the C∗08:02/16:02 and C∗05:01/16:01 types as ambiguous pairs. These were resolved by SSOP showing the apparent patient C∗ locus types as 08:02 and 16:02. The patient had a diagnosis of mantle cell lymphoma and an unrelated donor search was conducted based on the type of A∗02:01/29:02, B∗44:02/44:03, C∗08:02/16:02, DRB1∗04:01/07:01. Samples from three potentially matched URDs were received. All were listed in the NMDP registry as C∗08:02/16:02. Confirmatory typing on the URDs in our laboratory by SBT showed the same ambiguous pair as expected. However, SSOP appeared to rule out the C∗08:02/16:02 combination listed by the NMDP and indicated that C∗05:01/16:01 was the correct type for all three donors. To better confirm the C∗ locus types, sequencing with an allele-specific sequencing primer (HARP) was performed on all four individuals. C∗ locus typing by sequence specific primers was also performed for three. These methods identified C∗05:01 and 16:01 as the correct type in the patient and all three URDs. Review of the initial patient typing showed a single false negative DNA probe. This probe in the One Lambda SSOP kit separates the ambiguous pairs and is susceptible to false negative results. It was near the cut-off value in our patient as well as one of the donors. Laboratories using the techniques of One Lambda SSOP with SBT for high resolution typing should be aware of the possibility of an incorrect type when the C∗08:02/16:02 pair is identified. Close attention to B∗/C∗ associations is also warranted as the C∗05:01/16:01 pair is expected to be considerably more frequent than C∗08:02/16:02 in association with B∗44:02/44:03. Additional potential C∗08:02/16:02 URDs were listed in the initial search. The NMDP was notified that investigation of those donors might be useful.