The clinical utility of fetal cell sorting to determine prenatally fetal E/e or e/e Rh genotype from peripheral maternal blood

OBJECTIVE: This study was undertaken to determine the fetal E/e or e/e Rh g enotype prenatally from peripheral maternal blood by examining sorted fetal cells from alloimmunized and nonalloimmunized pregnancies. STUDY DESIGN: Eighteen maternal peripheral venous blood samples were obtain ed before amniocentesis from 15 pregnant women who were homozygous for the e allele. Five were not alloimmunized and 10 were alloimmunized. The mononu clear cell layer was isolated from the maternal blood and enriched for feta l nucleated red blood cells by flow cytometry with monoclonal antibodies to CD36 or CD71 and to glycophorin A. Eight samples were treated with CD45 mo noclonal antibody-coated magnetic beads before they were sorted to deplete the maternal sample of leukocytes (GD45(+) cells). We defined the positive fetal cell fractions as the monoclonal antibody positive-sorted cells deriv ed from the maternal samples. These included sorted cells that were CD36(+) /glycophorin A(+), CD71(+)/glycophorin A(+) and CD45(-) cells that were sor ted to become CD45(-)/CD36(+)/glycophorin A(+) or CD45(-)/CD71(+)/glycophor in A(+). The negative fractions were the cells that were negative for eithe r CD36/glycophorin A or CD71/glycophorin A or were the CD45(+) cells. Deoxy ribonucleic acid was isolated from all fractions and amplified by polymeras e chain reaction with allele-specific primers for the for e Rh genes. Gel e lectrophoresis was performed to detect fetal E/e or e/e Rh genotype. The fe tal E/e or e/e Rh genotype was confirmed by serologic and deoxyribonucleic acid testing. The accuracy of E/e or e/e Rh genotype determination from the positive cell fractions was compared with that of E/e or e/e Rh genotype d etermination from the negative fractions. RESULTS: Fetal E/e or e/e Rh genotype was determined correctly in 17 of 18 of the fetal cell enriched positive fractions (94%), Fetal E/e or e/e Rh ge notype was determined correctly in 11 of 14 of the maternal samples in the negative unselected cell fractions (79%). Fetal E/e or e/e Rh genotype was determined correctly in 15 of 18 sample fractions that underwent magnetic b ead separation with CD45 and were subsequently sorted into positive and neg ative fractions (94%). Fetal E/e or e/e Rh genotype was determined correctl y in 13 of 13 of the samples obtained from the alloimmunized pregnancies (1 00%). CONCLUSIONS: The use of monoclonal antibodies for cell sorting or for magne tic separation predicted fetal E/e or e/e Rh genotype from peripheral mater nal blood correctly in,as many as 100% of alloimmunized pregnancies. Thus n oninvasive fetal E/e or e/e Rh genotyping can be performed by polymerase ch ain reaction amplification of the rare fetal cells in maternal blood. The c orrect prediction of fetal E/e or e/e Rh genotype from the cell population not selected by the monoclonal antibodies suggests that there are fetal cel l types other than fetal nucleated erythrocytes that can also be used as a source of fetal deoxyribonucleic acid for noninvasive genetic diagnosis. Im proved technology may provide methods less laborious than cell sorting to a ccurately determine fetal Rh type from different fetal cell types that circ ulate in maternal blood.