BACKGROUND: The use of hemagglutination to phenotype red cells from recently transfused patients or of red cells that are coated with IgG can be time-consuming and difficult to interpret. Because the molecular bases of many blood group antigens are known, it was investigated whether polymerase chain reaction (PCR) analysis of DNA, from white cells in blood from transfused patients, could be used to predict the blood group antigen profile of a patient. STUDY DESIGN AND METHODS: To prevent problems arising from potentially poor-quality DNA in clinical samples, primers that flanked the polymorphism of interest and that replicated a relatively short PCR amplicon were used. The PCR products, with or without digestion with the appropriate restriction enzyme, were analyzed on gels. Samples were collected from 60 patients who had received from 2 to 50 units of RBCs in the 7 days before sample collection. RBCs from some of these patients were coated with IgG. Analyses for RHD/non-D, RHE/RHe, KEL1/KEL2, FYA/FYB, FY-GATA, JKA/JKB, and GYPA M/N were performed by using assays that had been validated with DNA prepared from untransfused volunteers of known phenotype. The genotyping assays were performed without knowledge of the expected result. RESULTS: The predicted genotype after analysis of the 60 patient samples was that expected from the results of phenotyping. In all cases, the molecular analysis gave a single result; no evidence of chimerism was obtained. CONCLUSION: In each case, the molecular genotype results were in agreement with the blood group antigen as determined by historical phenotyping, phenotyping after hypotonic washing, detection of alloantibodies in the patient's serum, or elution of alloantibody(ies). Under the conditions of these assays, reliable determination of a blood group allele can be made by PCR-restriction fragment length polymorphism testing.
BACKGROUND: The use of hemagglutination to phenotype red cells from recently transfused patients or of red cells that are coated with IgG can be time-consuming and difficult to interpret. Because the molecular bases of many blood group antigens are known, it was investigated whether polymerase chain reaction (PCR) analysis of DNA, from white cells in blood from transfused patients, could be used to predict the blood group antigen profile of a patient. STUDY DESIGN AND METHODS: To prevent problems arising from potentially poor-quality DNA in clinical samples, primers that flanked the polymorphism of interest and that replicated a relatively short PCR amplicon were used. The PCR products, with or without digestion with the appropriate restriction enzyme, were analyzed on gels. Samples were collected from 60 patients who had received from 2 to 50 units of RBCs in the 7 days before sample collection. RBCs from some of these patients were coated with IgG. Analyses for RHD/non-D, RHE/RHe, KEL1/KEL2, FYA/FYB, FY-GATA, JKA/JKB, and GYPA M/N were performed by using assays that had been validated with DNA prepared from untransfused volunteers of known phenotype. The genotyping assays were performed without knowledge of the expected result. RESULTS: The predicted genotype after analysis of the 60 patient samples was that expected from the results of phenotyping. In all cases, the molecular analysis gave a single result; no evidence of chimerism was obtained. CONCLUSION: In each case, the molecular genotype results were in agreement with the blood group antigen as determined by historical phenotyping, phenotyping after hypotonic washing, detection of alloantibodies in the patient's serum, or elution of alloantibody(ies). Under the conditions of these assays, reliable determination of a blood group allele can be made by PCR-restriction fragment length polymorphism testing.