Use of a simplified maldi-TOF workflow without DNA purification for mass-scale blood donor typing

In our blood service, we use simplified limited quality molecular typing (SLQMT) to identify blood donors with useful antigen pattern including rare donors. SLQMT is devised for an optimal information to effort ratio, simplicity, non-calls and rare miscalls are accepted, because relevant results are confirmed by serology before unit release. We evaluated the impact of a switch from pooled capillary electrophoresis (PCE) to MALDI-TOF (mTOF) in 2019.

A single well mTOF multiplex PCR was devised predicting 24 SNV including those determined previously. DNA was prepared using Sigma Extract-N-Amp reagents, amplifed in plates ( 2 µl DNA + 3 µl primer/Taq mix + 2 µl of extension primer mix) and analyzed on a mass array Dx analyzer 4 with chip prep module (Agena Bioscience). Initial results were cross-validated with PCE-typed samples; "new" SNV were validated by sequencing. Percentage of no-calls, number of donors tested, rare donors detected, donated units with predicted antigen typing results (including PCE) and discrepancies to PCE or serology results were monitored.

The number of SNV investigated by mTOF (n=24) increased (PCE: n=9). Donors tested per year were similar (mTOF: 14817 (mean, range: 14474 to 15377), PCE: 20861). No call rates were JK^a 2.2% (PCE: 4.6%), Fy^a 2.9% (13.6%), M 10.2% (4.9%), S 9.9% (8.4%), Co^a 2.2% (2.7%), Lu^b 8.3% (1.7%), Yt^a 2.1% (3.3%), Kp^b 8.1% (6.7%). We detected 558 rare donors, most frequenly Yt(a-) (n=192) and Co(a-) (n=166). "New" types of rare donors included Lu:-8 (n=48), Sc:-1 (n=4) and LW(a-) (n=3). Neither Js(a+b-) nor In(a+b-) donors were found consistent with heterozygote prevalences of 0.051% (n=33) for Js(a+b+) and 0.003% (n=2) for In(a+b+). In 2023, the 31% of donated units had typing results for Jk^a (Fy^a : 31%, M 30%, S 30%). Rare units numbered 1163, including 474 Yt(a-) and 370 Co(a-). Among 135 discrepancies detected since 29-Aug-2023, 67 were due to incorrect "old" PCE-typing, 41 to incorrect serologic results and 27 to incorrect mTOF predictions, most frequently antigen N (n=10). Only one misprediction was caused by a silent allele (JK*02N[c.1000G>T]).

Our simplified mTOF workflow extends the SLQMT concept and allows long-term sustained molecular donor typing. Most relevant antigens are predicted with sufficient accuracy at limited cost. Adding more SNV is unlikely to give more information without disproportionately high increases in workload.

no COI to declare