There is now a realistic alternative to termination of a pregnancy in which a fetus is diagnosed as having a potentially fatal inherited disorder. In utero transplantation (IUT), very soon after a diagnosis is confirmed, could offer therapeutic benefit before a disorder's pathology becomes evident. The rapidly expanding hematopoietic tissue in the early-gestation fetus readily accepts transplanted cells. If IUT is performed during the first trimester, when the immune system is undeveloped, even mismatched tissue cannot provoke a rejection response. Sustained (i.e., lifelong) engraftment can be achieved only through use of stem cells, those primitive cells that can generate all lineages within the bone marrow. For IUT, stem cells are the ideal material: relatively few are needed to correct a disorder and to provide lifelong normal hematopoiesis. The Second International Meeting on In Utero Stem Cell Transplantation and Gene Therapy, held September 1-2, 1997, in Nottingham, United Kingdom, brought together researchers in this field to discuss the questions to be addressed before this novel therapeutic approach is more widely accepted.

Myrtle Gordon (Royal Postgraduate Medical School, Hammersmith Hospital, London) explained the mathematics underlining the continued generation of both quiescent stem cells and committed progeny. There is around 1 stem cell per 1,000,000,000 total cells in bone marrow, so there is undoubtedly a process of self-renewal. The source of stem cells determines the number of primitive, self-renewing quiescent cells that can be generated. By inference, the fetus contains the greatest number of these primitive cells.

Dr. Colin Casimir (Imperial College School of Medicine at St. Mary's, London) then gave an overview of the gene therapy field and a critique of the relative lack of success thus far. Retroviruses can be packaged with gene constructs and are the ideal vehicle to transfer these to target cells. However, the logical targets are stem cells, and their quiescent nature renders them impregnable to viral infection (retroviruses require cycling cells). Much effort is being directed toward breaking this impasse, such as better strategies for harvesting quiescent stem cells and reliable methodologies to modify these cells to allow retroviral entry. Esmail Zanjani (VA Medical Center, University of Nevada at Reno) outlined the advantages of using the fetal sheep as a model for IUT (large animal, long gestation, resilience to surgery). Cotransplantation of stem cells with bone marrow stromal cells and postnatal boosting with additional (autologous) donor cells improves levels of donor cell engraftment in this model. There is extensive data on the successes of in utero procedures such as amniocentesis and coelomocentesis, and we can use this information to estimate the relative risk for IUT. Eric Jauniaux (University College London Medical School) hypothesized that injecting donor stem cells into the coelomic cavity would allow them to recirculate into the fetal vasculature.

There are compelling reasons for using both bone-marrow- or fetal-liver-derived tissue as a source of stem cells, but the optimum may well be determined by the disorder being treated. Fetal stem cells, being predominantly erythroid in their developmental commitment, are ideally suited to in utero transplantation for disorders such as the hemoglobinopathies. Bone-marrow-derived stem cells are appropriate for immunodeficiency disorders, where repeated in utero transplants might be needed and further donations can be obtained later in pregnancy if indicated. Rhodri Jones (University of Nottingham, United Kingdom) stressed the utility of fetal stem cells: high turnover rate, ability to generate large numbers of quiescent stem cells, and demonstrable long-term engraftment (using the sheep model). Joao Ascensao (VA Medical Center, University of Nevada at Reno) presented an overview of bone-marrow-derived stem cells. These require T-cell depletion before use (to reduce the risk of graft-versus-host disease, GvHD) but paradoxically some T cells must remain to guide engraftment. There is a plethora of T-cell depletion protocols at different centers, making comparison of clinical outcomes difficult. Angelo Tocci (Catholic University, Rome) indicated that adult-to-fetus engraftment was only likely to be successful when some form of GvHD had occurred, confirming that the presence of T cells is necessary to ensure engraftment by stem cells.

Jannine Wilpshaar (Leiden University Medical Center, the Netherlands) reported no major differences in expression of adhesion molecules on fetal stem cells between 14 and 22 weeks' gestation, although further work using the 8-to-13-week gestational age range (i.e., up to and including the age at which migration from liver to bone marrow occurs) should prove more informative. David Liu (City Hospital, Nottingham) outlined the requirements for a stem cell bank. Screening (including the ability to grow in culture, virology, karyotyping, and HLA and blood group typing) takes around three weeks so, it is imperative that a bank be established for IUT. Protocols for umbilical cord blood banking are applicable to fetal stem cell banking (Ruth Warwick, National Blood Service, Edgware, London).

François Forestier (Institut de Puericulture, Paris) and François Golfier (Hospices Civils de Lyon, Hotel Dieu, Lyon) also emphasized the utility of fetal stem cells, and also indicated that the high number present in fetal hematopoietic tissue was an advantage in choosing this source of stem cells for IUT. Morton Cowan (University of California at San Francisco) raised the prospect of cytokine/growth factor enhancement of stem cells to increase engraftment potential, although exposure of stem cells in vitro (i.e., prior to transplant) would be the recommended protocol. David Archer (Emory University School of Medicine, Department of Pediatrics, Atlanta) presented data showing that in the NOD/SCID mouse model of IUT, fetal stem cells have a nine times greater engraftment potential but that this advantage was abrogated if the recipients were irradiated prior to transplant, indicating that the recipient microenvironment is important in guiding engraftment.

The next session included updates on animal models and clinical cases. Graca Almeida-Porada (VA Medical Center, University of Nevada at Reno) showed that the level of engrafted (human) cells in sheep transplanted in utero can be increased after postnatal boosting. Although successful in around 50 percent of animals, this approach suggests that tolerance induction in utero, followed by postnatal boosting, is one likely route to enhanced engraftment by donor cells. Although postnatal boosting in the mouse model can again increase levels of donor cells, success is not evident in all animals (Ewa Carrier, University of California at San Francisco). Data from both groups indicate that evidence of microchimerism after transplant does not necessarily equate with tolerance.

Hematopoietic progenitor cells from both human fetal liver and adult bone marrow readily engraft in the NOD/SCID mouse model, and there was up to 25-fold expansion of the human cell compartment over and above the number of transplanted cells (Harv Fleming, Emory University, Atlanta), confirming the increased engraftment potential of human fetal stem cells. Fetal sheep transplanted in utero with human fetal hematopoietic cells that had been stored frozen for up to 6 months show evidence of engraftment (Elaine Anderson, Queen's Medical Centre, Nottingham) and donor cell numbers in postnatal animals could be boosted by injection of human G-CSF. Thus, human fetal stem cells can be stored frozen without impairing their engraftment capabilities. These data vindicate the choice of fetal stem cells for IUT and the establishment a cell bank.

There followed a section devoted to the experience with human IUT. Cesare Peschle (Istituto Superiore di Sanità, Rome) reported in utero transplantation in a case of beta⁰ thalassaemia. Two injections of purified, T-cell-depleted bone marrow stem cells from an HLA-identical sibling were given. At birth 4.5% donor cells were detected, but at 4 months, no donor cells were present. Magnus Westgren (Huddinge University Hospital, Huddinge, Sweden) reviewed his cases: all are hemoglobinopathies, treated in utero with fetal stem cells. There was no postnatal evidence of chimerism in any of the patients; all are transfusion dependent. George Wengler (University of Brescia, Italy) detailed two cases of severe combined immunodeficiency (SCID) treated in utero (both with purified stem cells from T-cell-depleted bone marrow). The first is healthy and shows chimerism (functional donor T cells in the peripheral blood), the other (recently born) has not engrafted and remains ill. Charles Peters (University of Minnesota at Minneapolis) reviewed the criteria for both postnatal and in utero transplantation for storage diseases. The success of postnatal bone marrow transplantation in such cases is poor, but it has been noted that where the pathology is not severe, there is evidence that a transplant can affect outcome. This bodes well for IUT, and efforts are now being directed at determining the criteria that can be used to select appropriate cases.

Daniel Surbek (University Hospital, Basel, Sweden) presented his experience with a fetoscope that has the potential to access the fetal circulation early in gestation. This technology would allow stem cell transplants to be performed early and could allow targeting of the cells more accurately. However, the fetal risk from such a procedure has yet to be determined. Jean-Louis Touraine (Hopital Edouard Herriot, Lyon) gave an overview of the cases he has treated in utero, all using fetal stem cells. There is mixed success, particularly in the hemoglobinopathies, where postnatal evidence of chimerism has not been long-lasting and all are now transfusion dependent. In the immunodeficiency cases there is engraftment with a successful outcome. Maria Sanna (Ospedale Regionale per le Microcitemie, Cagliari, Italy) presented a case of beta⁰ thalassaemia treated in utero using bone-marrow-derived stem cells. There was no evidence of engraftment.

The final scientific session concentrated on the prospects for in utero gene therapy. Chris Porada (VA Medical Center, University of Nevada at Reno) injected early-gestation sheep fetuses with engineered retroviral vector-containing supernatants and producer cells, and followed the offspring over five years. Expression of the viral product has been demonstrated in many hematopoietic lineages and the germ line has not been altered. Thus, introduction of retrovirally packaged gene constructs in utero have the potential for long term expression (although low, at present) and possibility therapeutic value. Larry Shields (University of Washington School of Medicine, Seattle) compared retroviral gene transfer into preterm and umbilical cord stem cells to assess the efficiency of the process. The data indicate that fetal stem cells will readily accept gene constructs delivered in this way and that there is no difference between preterm and umbilical cord cells. Colin Casimir used a novel technique to circumvent the problem of stem cell quiescence preventing integration of viral gene constructs. Engineered producer cells, expressing the membrane-bound form of stem cell factor, provide a growth signal to quiescent stem cells while simultaneously infecting them with a retrovirus carrying a gene of interest. In vitro experiments have achieved moderate success.

The last talk dealt with the ethics of fetal therapy (Sheila McLean, University of Glasgow). The question of a woman's right to chose termination or to keep an affected fetus was discussed, as were the legal aspects of what right a woman has to choose if an experimental therapy was available but unproven. Ethically, a woman has the right to opt for an experimental approach (IUT), but the specter of health authorities (at least in the U.K.) proscribing such an procedure on the grounds of cost cannot be ignored.

The discussion sessions flagged questions that need to be addressed with some urgency if we are to put in utero stem cell transplantation and gene therapy into the mainstream of therapeutic protocols. We need to investigate tolerance induction in utero; with current levels of engraftment, this is a worthy goal. We need some agreement on the source of stem cells to be used, and if cell banking is to become the norm, then we need to establish international standards for all. Finally we must agree on the transplantation protocol, and that is probably the greatest challenge. All information must be made available for scrutiny: IUT is not a procedure that, ethically, lends itself to clinical trial. Some delegates advocated a Web site to disseminate data, and this may well become established before the third international meeting is convened.

Further reading - bibliography of offline sources.

Long Term Bone Marrow Culture - information on how stem cells can be grown in the laboratory.

Choosing a BMT Center - one hospital's view, but the basic information is applicable everywhere.

International Bone Marrow Transplant Registry - a section on in utero transplantation is currently being developed.

The Most Powerful Cell - stem cells explained by the Why Files.

In Utero Stem Cell Transplantation: Report on the September 1996 Annual Meeting in Reno - review of the first annual meeting, divided into parts 1 and 2.