1033 LETTERS to the EDITOR First trimester prenatal diagnosis of trisomy 21 in fetal cells from maternal blood SiR,—More than 20 years ago Walknowska et aP proposed analysis of fetal cells in maternal blood for prenatal diagnosis. We2 and others3--7 have shown that fetal cells are present in the maternal circulation during the first and second trimesters. With the polymerase chain reaction, Y-specific DNA sequences have been identified in peripheral venous blood from most pregnant women carrying male fetuses.’ We also reported prenatal detection of a fetal,aneuploidy (47,XY, + 18) by analysis of fetal nucleated red blood cells (NRBCs) from a maternal blood sample drawn before chorionic villus sampling (CVS).8 (Blood samples drawn after invasive procedures could contain fetal cells because of an iatrogenic fetomaternal bleed). The sample was flow-sorted on the basis of four criteria: forward angle light scatter (denoting cell size); side scatter (cell granularity); presence of CD71 (transferin receptor); and presence of glycophorin-A, a cell surface sialoglycoprotein. The fetal chromosome complement was assessed by interphase fluorescence in-situ hybridisation (FISH) with chromosome- specific DNA probes for 18 and 21. We now report prenatal diagnosis of a female fetus with trisomy 21, again based on analysis of maternal blood drawn before CVS. The patient was a 42-year-old primigravida referred for prenatal diagnosis because of age. Ultrasonographic fetal crown-rump length was 33 mm (mean for 10-3 weeks’ gestation). Before transabdominal CVS, 20 ml antecubital venous blood was collected into vacutainers containing acid citrate dextrose solution A. Cytogenetic analysis of chorionic villi revealed 47,XX,+21. The patient decided on termination, after which the diagnosis was confirmed in the abortus. The methods used were similar to those described:8,9 106 cells were analysed by flow cytometry, 105 cells were flow-sorted, and 3-7 x 101 nucleated cells were plated onto two microscope slides. We used simultaneous hybridisation and dual colour detection of the 18 probe (digoxigenin/antidigoxigenin-fluoreseein isothiocyanate) Hybridisation of chromosome 21 specific probe set9 to fetal NRBCs. Three hybridisation signals can be seen in each nucleus. Digital imaging microscopy with Zeiss epifluorescence microscope and cooled CCD camera (Photometrics PM512) controlled by personal computer Grey scale images were captured with filter sets for DAPI and fluorescein (pseudocoloured and merged for display). and the 21 probe (biotin/avidin-Texas Red) (figure). The mean signal distributions of the slides scored by three independent observers, all unaware of the fetal karyotype, were: Chromosome-specific Hybridisation signals Mean no probes 1 2 3 4 nuclei scored 18 1 98 2 0 101 21 0 25 74 1 100 At the same time we analysed five 20 ml blood samples drawn before CVS (10-12 weeks’ gestation) from women subsequently found to be carrying chromosomally normal fetuses. A mean of 7-5% (range 0-12-4%) nuclei displayed three signals with the 21 probe. In the only other fetal trisomy 21 cases in which we had the opportunity to analyse maternal blood before CVS (10-1 weeks’ gestation), 3 of 104 nuclei (2-9%) displayed three signals with the 21 probe. These preliminary data are encouraging, and consistent with our previous findings and those of others who studied samples drawn after invasive procedures However, sensitivity and specificity must be evaluated in trials. If these issues are successfully addressed, we see this method being used initially as a screening test for fetal aneuploidy with confirmation of abnormal results by conventional invasive techniques, such as CVS or amniocentesis. Eventually fetal cells isolated from maternal blood might be used for definitive fetal diagnosis. Such methods would then enable couples to elect prenatal diagnosis without subjecting their fetuses to risk. We thank L. P. Shulman and 0. P. Phillips (University of Tennessee, Memphis) for their collaboration. Funded in part by National Institute of Health and Human Development NOI-HD-2904. Department of Obstetrics and Gynecology, Microbiology and Immunology, and Medicine, University of Tennessee, Memphis, Tennessee 38103, USA SHERMAN ELIAS JAMES PRICE MICHAEL DOCKTER STEPHEN WACHTEL AVIRACHAN THARAPEL JOE LEIGH SIMPSON Integrated Genetics Laboratories, Inc Framingham, Massachusetts, USA 1. Walknowska J, Conte FA, Grumback MM. Practical and theoretical implications of fetal/maternal lymphocyte transfer. Lancet 1969; i: 1119. 2. Wachtel SS, Elias S, Price J, et al. Fetal cells in the maternal circulation: isolation by multiparameter flow cytometry and confirmation by PCR. Hum Reprod 1991; 6: 1466. 3. Lo Y-MD, Wainscot JS, Gilmer MDG, et al. Prenatal sex determination by DNA amplification from maternal preipheral blood. Lancet 1989; ii: 1363. 4. Mueller UW, Hawes CS, Wright AE. Isolation of fetal trophoblast cells from peripheral blood of pregnant women. Lancet 1990; 336: 197. 5. Bianchi DW, Flint AF, Pizzimenti MF, et al. Isolation of fetal DNA from nucleated erythrocytes in maternal blood. Proc Natl Acad Sci USA 1990; 87: 3279. 6. Camaschella C, Alfarno A, Gattardi E, et al. Prenatal diagnosis of fetal haemoglobin Lepore-Boston disease on maternal peripheral blood. Blood 1990; 75: 2101. 7. Bruch JF, Metezeau D, Garcia-Fonknechten N, et al. Trophoblast-like cells sorted from peripheral maternal blood using flow cytometry: a multiparametric study involving transmission electron microscopy and fetal amplification. Prenat Diagn 1991; 10: 787. 8. Pnce J, Elias S, Wachtel SS, et al. Prenatal diagnosis using fetal cells isolated from maternal blood by multiparameter flow cytometry. Am J Obstet Gynecol 1991; 165: 1737. 9. Klinger K, Landes G, Shook D, et al. Rapid detection of chromosome aneuploidies in uncultured aminocytes by using fluorescence m-situ hybridisation (FISH). Am J Hum Genet 1992; 51: 55. 10. Cheuh J, Lebo R, Flandermeyer R, et al. Prenatal diagnosis using fetal cells in the maternal blood. Prenat Diag 1992; 12: S68. KATHERINE W. KLINGER