Applying a test system for discriminating fetal from maternal cells

Objective The objectives of this study were to enhance and apply a simple system capable of testing the capacity Of putative, gender-independent fetal cell markers. individually and in combination, to discriminate between fetal and maternal cells. Methods Chorionic villi tissue obtained from 25 male pregnancies at 10 to 12 weeks' gestation served as the experimental group. Following removal of villi pieces for clinical use, unattached cells were collected by centrifugation of the CVS fluid, fixed in the tube, and used as a source of mixed fetal and maternal cells. Blood obtained from a fetus at 13 weeks' gestation served as a positive control. Peripheral blood from two adult males served as negative controls. Antibodies to three possible fetal markers were tested using immunohistochemical techniques: anti-Flk-1, anti-epsilon globin, and anti-CD71. Each antibody was used alone and in combination in conjunction with fluorescent in situ hybridization (FISH) of X and Y chromosomes to confirm that positively stained cells were in fact fetal in origin. Results On CVS samples, the average predictive value for anti-Flk-1 was 35.8%, 76.2% for anti-CD71, and 90.5% for anti-epsilon. The combination of anti-epsilon and anti-CD71 antibodies identifying a fetal cell was 87.2% and the combined use of single and double antibodies gave a value of 82.7%. The combination of anti-epsilon globin and anti-CD71 increased the sensitivity of identifying pure fetal blood cells from 63%, for anti-epsilon alone. and 67%. for anti-CD71 alone, to 86%. Conclusion Although anti-Flk-1 has been reported to be a successful marker of fetal cells, the results in this test system did not support this finding. This work supports the use of CVS washings containing both fetal and maternal cells as a viable test system for assessing antigenic markers. The combination of anti-CD71 and anti-epsilon as fetal identifiers may increase the chances of identifying a fetal cell without compromising the predictive value. Copyright (C) 2003 John Wiley Sons, Ltd.