Genotyping patients' and donors' blood groups for efficient blood therapy.

Dear Sir, The applications of molecular genetics in the immunohaematology laboratory are increasing day by day and are spreading worldwide on the basis of the knowledge that most of the blood group polymorphisms depend on single nucleotide polymorphisms1. Many molecular techniques which combine both DNA analysis and bioinformatics software are now available for genotyping red blood cells (RBC): automated, high throughput platforms allow extended genotyping of donors or patients in blood banks or in immunohaematology reference laboratories2 in order to minimise transfusion complications, such as alloimmunisation and haemolytic transfusion reactions. Within a few months a Rare Blood Group Bank will be started in the Transfusion Service of our Hospital using a molecular biology approach. This new project, supported by Regional Blood Centre of Emilia-Romagna (Italy), will involve (i) identifying donors of rare blood group, using high throughput automated genotyping, and then (ii) collecting and cryopreserving the rare phenotype blood units. Furthermore, the same technology will also be applied to predict RBC phenotype in chronically transfused, alloimmunised patients who have a positive direct antiglobulin test (DAT), with the aim of administering fully compatible blood units, with better RBC recovery. We evaluated the accuracy, reliability and flexibility of RBC genotyping using a BLOODchip IDCore+ kit (Progenika Biopharma, Derio, Spain) and XMAP Luminex® technology (Luminex Corporation, Austin, TX, USA) on samples from patients and donors. We performed genotyping on DNA samples from 45 adult and paediatric patients affected by various haemoglobinopathies and haematological disorders: beta-thalassaemia (n=33), sickle cell disease (n=9), microdrepanocytosis (n=1), Blackfan-Diamond anaemia (n=1), and aplastic anaemia (n=1). Five out of the 33 patients with beta-thalassaemia patients and two out of the nine with sickle cell disease are alloimmunised. The alloantibody specificities are: anti-c+E, anti-E, anti-C+e, anti-Kell+Wra, anti-Wra, anti-Kpa, and anti-M. The phenotype of all patients’ RBC was reported in their clinical records and was determined by serology. Thirty-two of the samples from the 45 patients were also typed with HEA BeadChip (BioArray Solutions, Warren NJ, USA) to evaluate the performance of the microarray molecular biology technique on recently transfused patients’ samples. Divergent results were solved by polymerase chain reaction with sequence specific primers (PCR-SSP) (Innotrain, Kronberg/Taunus, Germany). Ninety-nine repeat blood donor samples were also genotyped and results were compared with those obtained with serological typing (Sanquin, Amsterdam, the Netherlands) (Table I). Table I Samples and typing methods. DNA was extracted from buffy coat samples using magnetic separation (Mag maxi PLUS kit, LCG Genomics GmbH, Berlin, Germany) on an automated platform (Microlab®Sampler SD, Hamilton Company, Bonaduz, Switzerland). RBC genotyping was performed with the BLOODchip IDCore+ Kit which can identify polymorphisms of the RHCE, KEL, FY, JK GYPA, GYPB, DI DO, CO, YT genes responsible for 33 RBC antigens. The products of multiplex polymerase chain reaction (PCR), using biotinylated deoxycytidine triphosphate (dCTP), are hybridised onto 100 sets of coloured micro-beads: each set exhibits a unique fluorescence signature of red and infrared dye and is bound to a specific probe which hybridises to a DNA complementary region that specifies a single nucleotide polymorphism. The hybridisation product is then labelled with streptavidin-conjugated phycoerythrin. BIDS software (Progenika Biopharma, Derio, Spain) generates a work list and sends it to the Luminex platform. Phycoerythrin and bead fluorescences are detected by Luminex® laser and quantified signals are analysed by BIDS software which produces genotype results and converts them into predicted phenotypes for the RBC antigens tested. BLOODchip IDCore+ obtained a complete genotype in all RBC genes in all samples with zero “no calls” or “not valid” results (repetition rate=0%). Internal and external controls were valid for all tests. Among the patients’ samples the genotype-predicted phenotype was different from the serological typing in 9/45 and 3/45 patients’ for RH and KEL antigens, respectively: such discrepancies between genotyping and serological phenotyping probably depend on old serological results reported in clinical records. Recent transfusions impaired serological typing, so we used PCR-SSP to solve discrepancies: PCR-SSP genoptype confirmed RHCE and KEL genotyping results of BLOODchip IDCore+ and HEA BeadChip. In patients’ samples typed with both BLOODchip IDCore+ and HEA BeadChip, as an alternative test, concordance was 100% for all shared antigens of the Rh, Kell, Duffy, Kidd, MNS, Diego Dombock, Colton and Yt systems (Table II). Table II Discrepant results in patients’ samples (n =45). The HEA BeadChip genotype-predicted phenotype was concordant with the results of other genotyping methods, BLOODchip and PCR-SSP, for all shared antigens. In 4/32 samples we had “not valid” tests for contamination of negative controls, probably because of the poor experience with this technology. Retyping non-valid tests gave complete results in all four samples. HEA BeadChip, even if tested on only 32 patients’ samples, seemed simple and fast and, thus, a high throughput technology suitable for genotyping of a large cohort of donors and patients. The minimum of eight samples per batch could be a limitation for urgent patient typings. r’s variant associated with weak C expression was detected in one patient and one donor, both of whom were Black Africans. Special antigens related to the Black African population were identified: (i) VS+V+ expression in 2/99 donors and 5/45 patients and (ii) Fy (a−b−) phenotype (FYB-33/FYB-33 genotype) in one donor and nine multiply transfused patients with sickle cell disease, respectively. We did not find discrepancies between BLOODchip genomic and serological Rh, Kell Kidd, Duffy, MNS system antigen typing in any of the samples from donors. Haemagglutination is the gold standard method for RBC typing, but there is no doubt that blood group genotyping can be a useful tool for blood banks and immunohaematology reference laboratories, both for patients and donors. In our region, especially in the Po Delta area, where malaria was endemic in the past, there is a large group of patients with beta-thalassaemia and immigration waves are bringing a growing number of West African patients affected by sickle cell disease who need periodic RBC transfusions. Some of them are already alloimmunised and the interethnic differences between patients and donors, for example in Duffy phenotype, are a further hurdle to overcome in order to minimise transfusion complications. Extended serological typing of donors for antigen-negative blood is difficult to apply on a large scale because of the time required, weak reactivity and the fact that some antisera are limited or no available at all; furthermore, it is not useful in recently transfused and DAT-positive patients2. In this context, a high throughput automated system is essential to perform extended RBC genotyping. The Progenika BLOODchip technology is easy to use and flexible: it takes less than 5 hours for the overall procedure and adapts to a variable number of samples (up to 48 per run) with a minimum increase in time spent. BIDS software provides assistance in a step-by-step manner, from generating the work-list to single nucleotide polymorphism genotype and printing the predicted phenotype in a final report. This technology is reliable because it allows a wide range of DNA concentrations and has a very low contamination rate. Results obtained by BLOODchip IDCORE+ analysis concurred with those obtained with genotyping and phenotyping in all cases, except in samples from multiply transfused patients previously characterised by serology and with rare variants not detectable with conventional haemagglutination tests. The high level of concordance obtained shows that an easy high throughput assay such as BLOODchip IDCORE+ has the potential to be applied: (i) to predict the phenotype of donors to increase inventories of antigen-negative blood (especially for high prevalence antigens)3 and of RBC available for antibody identification, and (ii) to determine blood group genotype of chronically transfused patients. These patients could benefit from receiving molecularly-matched RBC units4,5, preventing post-transfusion complications (alloantibody formation and subsequent haemolytic transfusion reactions) depending on antigenic differences between donors and patients. Extended genomic typing of donors and patients of various ethnic origins will increase blood safety by providing better-matched, compatible homologous blood products to patients who are chronically transfused or at increased risk of alloantibody formation.