Limited feasibility of routinely analyzing fetal cells from maternal blood by using magnetic activated cell sorting and polymerase chain reaction for prenatal diagnosis.

Abstract

Background: To assess the feasibility of analyzing fetal cells from maternal circulation by using magnetic activated cell sorting (MACS) and polymerase chain reaction (PCR) for prenatal diagnosis.

Methods: Thirty-one high-risk (either advanced maternal age or abnormal serum Down screening) pregnant women (14-22 weeks) were enrolled. Twenty ml of venous blood from each woman after amniocentesis were pretreated with density gradient centrifugation and sorted by MACS with monoclonal antibodies: anti-CD71 (n = 26) or anti-GPA (n = 5). Nested PCR with Y-specific probes--Y1.5-Y1.8 (n = 10) and Amelogenin (n = 21) were then applied to the sorted nucleated red blood cells (NRBCs) for fetal sex determination. These results were compared with cytogenetic data. To assess the sensitivity of PCR, different proportions of known male and female cultured amniocytes were mixed and amplified for gender identification.

Results: Karyotypes were normal in all fetuses (18 females and 13 males). The proportions of NRBCs (in total cells) sorted by MACS--anti-GPA or anti-CD71 were 50% (2000 +/- 1500) and 85% (350 +/- 280), respectively. Accuracies of sex determination by PCR-Amelogenin or Y1.5-Y1.8 were 76.2% (16/21) and 50% (5/10), respectively. Three cases resulted in PCR failure. Assay of nested PCR inferred that after cell sorting, existence of at least 20% of male fetal cells mixed in maternal blood circulation was required for prenatal diagnosis under current methodology.

Conclusions: We confirmed the existence of fetal NRBCs in maternal blood during pregnancy. The low accuracy of sex determination (76.2%) may be attributed to contamination of either maternal NRBCs or non-NRBCs. No conclusive data, however, so far demonstrates the ideal marker to identify the origin of NRBCs. Without specific fetal cell marker and more sophisticated fetal cell analysis methodologies, in our experience, the feasibility of routinely analyzing fetal cells from maternal blood for prenatal diagnosis is limited.