For personal use only. Reproduce with permission from The Lancet Publishing Group. 962 THE LANCET • Vol 357 • March 24, 2001 CORRESPONDENCE Authors’ reply Sir—An absolute increase in fetal-cell number by in-vitro expansion has been seen in clinical samples. 1 This approach exploits the growth potential of presumed circulating stem or progenitor cells, whose presence can be inferred from the colony formation in culture. Sorting cells, such as by CD34+ enrichment, can increase the target-cell concentration, but inevitably loses cells. Campagnoli and colleagues used red-cell lysis, followed by fluoresence-activated cell sorting for two markers (CD45+ and CD34+), cell depletion, or magnetic sorting. We collected mononuclear cells from 9 mL unsorted antecubital venous blood. We cultured two to eight samples (median four) of 510 5 cells per patient within hours of sampling. Fetal cells in maternal blood probably comprise non-expandable as well as colony-forming cells (stem or progenitor cells). Expandable cells include CD34+ cells, but CD34 cells have produced haemopoietic progenitor cells in a fetal sheep model. 2 A novel cell-surface antigen on CD34+ and CD34 cells capable of multilineage differentiation has been described. 3 CD34 and Clonal culture of fetal cells from maternal blood Sir—Boris Tutschek and colleagues (Nov 18, p 1736) 1 report the successful clonal culture of fetal cells from an unspecified volume of maternal blood for non-invasive prenatal diagnosis. Several groups, including ours, have investigated this and similar approaches, none of which has led to an independently reproducible technique in clinical samples, 2 despite clear selective expansion of fetal over maternal progenitors in mixing experiments. This disparity in findings is not surprising given that there are only around 24 fetal nucleated cells in a 20 mL blood sample, 3 and that only 0·05% of first trimester fetal blood mononuclear cells are CD34+. 4 Such calculations if correct suggest that several litres of maternal blood may be needed to ensure culturing a single fetal haemopoietic progenitor. Tutschek and colleagues identified fetal colonies in four of 12 mothers by use of standard culture media, but only 0·3% of colonies were fetal. Three times as many colonies contained mixed fetal and maternal cells. How such contamination occurred, since each colony is clonally- derived from one cell, is not clear. First, they claim to have used micro- manipulation but provide no details; as a mechanical technique for harvesting individual cells, this should yield pure colonies. Second, they used PCR rather than fluorescence in-situ hybridisation which is less sensitive to contamination and would have shown the individual cell source. If the explanation is crowding of colonies, surely Tutshek and colleagues thought of increasing the size of the plate or reducing the plating density. They present no mixing experiments to validate their technique. Such experiments would have shown that before CD34+ enrichment, preliminary liquid culture and shorter harvest times increase the chance of pure fetal colonies. Although clonal fetal progenitor expansion is much slower than enrichment of committed fetal cells in maternal blood, its whole rationale is the yield of pure, not mixed, genetic material. In model systems, we have never encountered overlapping maternal and fetal colonies; first trimester fetal Burst-Forming Units erythroid are larger, appear earlier, and are more haemoglobinised than their maternal counterparts. The fundamental obstable to this expansion approach is similar to that of the enrichment approach: lack of a cell type seen only in fetal blood and not endogenously in maternal blood. We have identified a readily expandable fetal mesenchymal stem cell, which is only in first-trimester fetal blood. 5 The dilutional mathematics of rare fetal cell types in maternal blood, however, pose a considerable challenge to their application for non-invasive prenatal diagnosis. This work is supported by Wellbeing and Action Research. Cesare Campagnoli, *Irene Roberts, Sailesh Kumar, Phillip R Bennett, Nicholas M Fisk Institute of Reproductive and Developmental Biology and *Department of Haematology, Imperial College School of Medicine, Hammersmith Hospital Campus, London W12 0NN, UK 1 Tutschek B, Reinhard J, Kögler G, Wernet P, Niederacher D. Clonal culture of fetal cells from maternal blood. Lancet 2000; 356: 1736–37. 2 Chen H, Griffin DK, Jestice K, Hackett G, Cooper J, Ferguson-Smith MA. Evaluating the culture of fetal erythroblasts from maternal blood for non-invasive prenatal diagnosis. Prenat Diagn 1998; 18: 883–92. 3 Bianchi DW, Williams JM, Sullivan LM, Hanson FW, Klinger KW, Shuber AP. PCR quantitation of fetal cells in maternal blood in normal and aneuploid pregnancies. Am J Hum Genet 1997; 61: 822–29. 4 Campagnoli C, Fisk N, Overton T, Bennett P, Watts T, Roberts I. Circulating hematopoietic progenitor cells in first trimester fetal blood. Blood 2000; 95: 1967–72. 5 Campagnoli C, Fisk NM, Kumar S, Bellantuono I, Bennett PR, Roberts IAG. Identification of mesenchymal stem cells in human first trimester fetal blood, liver and bone marrow. Blood 2000; 96: 279A. weakly CD34+ cells would be lost in a CD34+ selection. In Campagnoli and colleagues’ study, fetal progenitor cells peaked in second- trimester samples and after 2 weeks of culturing. As little as 50 mL fetal blood in the entire maternal circulation would yield 0·8 fetal progenitor cells per mL maternal blood. We harvested the colonies after 14 days in semi-solid media-promoting colony formation. We took via pipette 2 L semi-solid medium in which the colonies had grown under an inverted microscope. We expected neighbouring cells or even free DNA carried over to be isolated and amplified along with the target cells. We termed colonies with more than two parental PCR peaks mixed. Our aim was not strictly clonal expansion, but to explore the feasibility of growing fetal colonies from maternal blood with a view to prenatal diagnosis. Clearly, the techniques for culturing and micromanipulation would have to be refined to avoid the mixing of maternal or fetal cells with plucked colonies. We used the term clonal knowing that, in the strict sense of the word, we did not and could not prove the clonal nature of the colonies. The PCR we used detects minority DNA or cells when present in excess of 2·5%. 4 This sensitivity can explain the occurrence of the high number of mixed colonies, which might pose a diagnostic difficulty, as we discussed. With fluoresence in-situ hybridisation, such as for the Y chromosome, theoretically all cells from a colony could have been assessed based on the fetal sex. The fetal origin of individual male fetal cells could be identified, but the large number of colonies required to find the few fetal colonies (almost 3000 individual colonies were screened) precluded the use of this technique. *Boris Tutschek, Joscha Reinhard, Gesine Kögler, Peter Wernet, Dieter Niederacher Departments of *Obstetrics and Gynaecology and Transplantation Diagnostics and Cell Therapeutics, University Hospital, 40225 Düsseldorf, Germany; and University College London Medical School, London, UK 1 Valerio D, Altieri V, Cavallo D, Aiello R, Antonucci FR. Detection of fetal trisomy 18 by short-term culture of maternal peripheral blood. Am J Obstet Gynecol 2000, 183: 222–25. 2 Zanjani ED, Porada CD, Crapnell KB, et al. Production of human hepatocytes by human LIN2 CD34+/ cells in vivo. Blood 2000; 96: 2127 (abstr). 3 Giesert C, Kuci S, Almeida-Porada G, Zanjani ED, Kanz L, Bühring J. The monoclonal antibody W7C5 detects a novel cell surface antigen expressed on CD34+ and CD342 stem cell subsets. Blood 2000; 96: 2853 (abstr). 4 Adinolfi M, Sherlock J, Soothill P, Rodeck C. Molecular evidence of fetal- derived chromosome 21 markers (STRs) in transcervical samples. Prenatal Diagn 1995; 15: 35–39.