Circulating cell-free fetal DNA for the detection of RHD status and sex using reflex fetal identifiers.

OBJECTIVE: To determine the sensitivity and specificity of circulating cell-free fetal DNA in determining the fetal RHD status and fetal sex. METHODS: Maternal blood was collected in each trimester of pregnancy from RhD negative nonalloimmunized women. Whole blood was centrifuged, separated into plasma and buffy coat, and frozen at -80°C. DNA analysis was conducted via allele-specific primer extensions for exons 4, 5, and 7 of the RHD gene and for a 37-base pair insertion in exon 4 (RHD pseudogene; psi) three Y-chromosome sequences (SRY, DBY, and TTY2), and an extraction control (TGIFL-like X/Y). RhD serotyping on cord blood and gender assessment of the newborns were entered into a Web-based database. RESULTS: One hundred twenty women were enrolled. The median gestational age at the first venipuncture was 12.4 (range: 10.6-13.9) weeks with 120 samples drawn; 118 samples were drawn at 17.6 (16-20.9) weeks; and 113 samples at 28.7 (27.9-33.9) weeks. Overall accuracy for RHD was 99.1%, 99.1%, and 98.1% for each trimester and was 99.1%, 99.1%, and 100% for fetal sex determination. CONCLUSIONS: Fetal RHD genotyping and sex can be very accurately determined in all three trimesters using circulating cell-free fetal DNA in the maternal circulation.

PTPN11 analysis for the prenatal diagnosis of Noonan syndrome in fetuses with abnormal ultrasound findings.

Noonan syndrome (NS) is an autosomal dominant disorder characterized by short stature, congenital heart defects and distinctive facies. The disorder is genetically heterogeneous with approximately 50% of patients having PTPN11 mutations. Prenatally, the diagnosis of NS has been suspected following certain ultrasound findings, such as cystic hygroma, increased nuchal translucency (NT) and hydrops fetalis. Studies of fetuses with cystic hygroma have suggested an NS prevalence of 1-3%. A retrospective review was performed to assess the utility of PTPN11 testing based on prenatal sonographic findings (n = 134). The most commonly reported indications for testing were increased NT and cystic hygroma. Analysis showed heterozygous missense mutations in 12 fetuses, corresponding to a positive test rate of 9%. PTPN11 mutations were identified in 16% and 2% of fetuses with cystic hygroma and increased NT, respectively. Among fetuses with isolated cystic hygroma, PTPN11 mutation prevalence was 11%. The mutations observed in the three fetuses with hydrops fetalis had previously been reported as somatic cancer mutations. Prenatal PTPN11 testing has diagnostic and possible prognostic properties that can aid in risk assessment and genetic counseling. As NS is genetically heterogeneous, negative PTPN11 testing cannot exclude the diagnosis and further study is warranted regarding the other NS genes.

First- and second-trimester Down syndrome screening markers in pregnancies achieved through assisted reproductive technologies (ART): a FASTER trial study.

OBJECTIVE: To determine whether first- and second-trimester Down syndrome screening markers and screen-positive rates are altered in pregnancies conceived using assisted reproductive technologies (ARTs). METHODS: ART pregnancies in the multicenter FASTER trial were identified. Marker levels were evaluated for five types of ART: in vitro fertilization with ovulation induction (IVF-OI), IVF with OI and egg donation (IVF-OI-ED), IVF with ED (IVF-ED), and intrauterine insemination with OI (IUI-OI) or without OI (IUI). Each group was compared to non-ART controls using Mann-Whitney U analysis. RESULTS: First-trimester marker levels were not significantly different between ART and control pregnancies, with the exception of reduced PAPP-A levels in the IUI-OI group. In contrast, second-trimester inhibin A levels were increased in all ART pregnancies, estriol was reduced and human chorionic gonadotropin (hCG) was increased in IVF and IUI pregnancies without ED, and alpha-fetoprotein (AFP) was increased in ED pregnancies. Second-trimester screen-positive rates were significantly higher than expected for ART pregnancies, except when ED was used. CONCLUSIONS: These data show that ART significantly impacts second-, but not first-, trimester markers and screen-positive rates. The type of adjustment needed in second-trimester screening depends on the particular type of ART used.

Circulating hematopoietic progenitor cells in a fetus with alpha thalassemia: comparison with the cells circulating in normal and non-thalassemic anemia fetuses and implications for in utero transplantations.

Our aim was to evaluate the number of progenitor cells circulating in an alpha-thalassemic fetus during its infusion in utero with paternal CD34(+) and adult red cells and to compare those values with those circulating in normal and non-thalassemic anemic fetuses of matched gestational age. The treatment of the alpha-thalassemic fetus has been described elsewhere. Fetal blood was obtained from normal and anemic fetuses by fetal blood sampling for diagnostic or therapeutic purposes according to a protocol approved by the human subject committee. The number of progenitor cells in fetal blood was estimated on the basis of the number of colonies they gave rise to in semisolid cultures. The alpha-thalassemic fetus, as did the other fetuses analyzed, contained high numbers (10(6)-10(7) depending on the age) of progenitor cells, values which were higher than the number (10(4)-10(5)) of paternal progenitor cells being transplanted. Progenitor cells with adult characteristics (adult kinetics of differentiation) were detected rapidly (10 min) after the CD34(+) cell infusion, but were not detectable 2-3 weeks after the transplant. These results indicate that adult progenitor cells do not have a numerical advantage when transplanted into alpha-thalassemic fetuses.

Multifetal pregnancy reduction.

Multifetal pregnancies have increased dramatically since the introduction and improvement in assisted reproductive techniques. Multifetal pregnancy reduction (MPR) is a technique developed over the past 15 years to deal with the sequelae of higher order multiple gestations resulting from infertility treatment. With increasing operator experience, loss rates for patients undergoing MPR have declined dramatically. The present review addresses recent data on the outcomes of patients undergoing MPR, as well as recent information on the natural history of higher-order multiple gestations in individuals not undergoing MPR. New developments in this area such as the use of chorionic villus sampling before MPR, as well as MPR to a single fetus, and information on the psychological follow-up of MPR patients are also discussed.

Chorionic villus sampling before multifetal pregnancy reduction.

OBJECTIVE: This study was undertaken to determine the technical feasibility and accuracy of chorionic villus sampling before multifetal pregnancy reduction and to determine whether sampling increases the pregnancy loss rate after the reduction procedure. STUDY DESIGN: Between January 22, 1986, and January 20, 2000, a total of 1183 patients underwent first-trimester multifetal pregnancy reduction at Mount Sinai Medical Center. Chorionic villus sampling was attempted in 86 patients before the reduction procedure. Information on the technical success and accuracy of chorionic villus sampling, as well as pregnancy outcome, was collected on all patients. Pregnancy loss rates before 24 weeks' gestation in patients undergoing chorionic villus sampling before multifetal pregnancy reduction were compared with rates in patients not undergoing sampling. RESULTS: Chorionic villus sampling was successfully completed in 85 (98.8%) of 86 patients in whom sampling was attempted. Of 166 fetuses, 165 (99.4%) were successfully sampled. Of 165 fetuses, 3 (1.8%) had karyotypic abnormalities. Sampling errors were probably made in 2 (1.2%) of 165 fetuses. Of the 73 patients who have been delivered or are beyond 24 weeks' gestation, only 1 patient (1.4%) had a pregnancy loss after the multifetal pregnancy reduction. CONCLUSIONS: Chorionic villus sampling before multifetal pregnancy reduction is technically feasible and accurate, with an acceptably low sampling error rate. Chorionic villus sampling before multifetal pregnancy reduction appears to be safe and does not increase the risk of loss after the reduction procedure.

Gastrointestinal disorders of the fetus.

The approach to developing a differential diagnosis to an abnormal gastrointestinal ultrasound finding should include consideration of the lesion's location, size, shape, and echogenicity; fetal gender; relationship to and understanding of adjacent structures; and the presence of associated anomalies. Once a differential diagnosis has been established, ancillary testing, such as amniocentesis, maternal serology, or fetal echocardiography should be undertaken to refine further the diagnosis. A multidisciplinary team approach should be taken in the prenatal management of a suspected fetal anomaly. This may include collaboration with specialists in the fields of perinatology, prenatal ultrasound, genetic counseling, neonatology, pediatric surgery, and patient support groups.

Successful pregnancy in severe methylmalonic acidaemia.

Methylmalonic acidaemia is an inborn error of metabolism characterized by recurrent episodes of life-threatening ketoacidosis. With improved and intensive treatment, these patients are living into adulthood, but many experience late-onset disease complications such as chronic renal failure, chronic pancreatitis and osteopenia. We report the successful delivery of a healthy baby to a 20-year-old woman with vitamin B12-unresponsive methylmalonic acidaemia who has these late-onset manifestations of the disease and had plasma methylmalonic acid concentrations of 1900 mumol/L during the first trimester of pregnancy.

Controversies in the intrapartum management of twin gestations.

Advances in reproductive endocrine technology have helped to make twin gestations commonplace; however, as this article suggests, many unanswered questions and areas of controversy about the intrapartum management of twin gestations remain. Continued research in this area and the performance of prospective studies will shed further light on many of these topics.

One hundred consecutive cases of selective termination of an abnormal fetus in a multifetal gestation.

OBJECTIVE: To determine whether transabdominal selective termination of one or more abnormal fetuses in a multifetal pregnancy with dichorionic placentation is a safe and effective procedure. METHODS: One hundred consecutive selective termination procedures were performed by transabdominal injection of potassium chloride into the heart or umbilical vein of an anomalous fetus in a multifetal pregnancy. All of the abnormal fetuses were presumed to have dichorionic diamniotic placentas, based on an ultrasound evaluation before the procedure. Follow-up data were obtained for each patient regarding the development of postprocedural complications, laboratory or clinical evidence of a coagulopathy, maternal or neonatal morbidity, gestational age at delivery, and birth weight of the infants. RESULTS: Ninety-one sets of twins were reduced to singletons, six sets of triplets were reduced to twins, two sets of triplets were reduced to singletons, and one set of quadruplets was reduced to triplets. The anomalous fetus or fetuses were identified correctly and terminated in each case. Three patients spontaneously aborted, and one women electively terminated her pregnancy 2 weeks after the procedure. The mean gestational age at delivery of the 96 patients who delivered surviving infants was 36.8 weeks, and 85.4% delivered at 32 weeks or later. Three women developed laboratory evidence of a coagulopathy, but there were no cases of clinically evident disseminated intravascular coagulation. CONCLUSION: This procedure, performed at a single institution by a small number of operators using a common protocol, accomplished its objective in all cases, was accompanied by a low spontaneous loss rate, and resulted in the birth of healthy infants at or near term in the vast majority of cases. This series suggests that selective termination is a reasonable option to consider when one abnormal fetus is found in a multifetal pregnancy with dichorionic placentation.

Circulating hematopoietic stem cell populations in human fetuses: implications for fetal gene therapy and alterations with in utero red cell transfusion.

Circulating progenitor cell populations in normal human fetuses and fetuses with various hematological problems were evaluated. Thirty blood samples from 21 human fetuses (17-36 weeks of gestation) were assayed for erythroid, myeloid, and mixed-cell progenitor cells. The mean number of progenitor cells/10(4) blood mononuclear cells in the normal fetal population was 103 +/- 47. Granulomonocytic and mixed progenitor cells (capable of giving rise to both erythroid and myeloid progeny) were the predominant progenitor types in these samples, with pure erythroid progenitors barely detectable. The frequency of progenitor cells in the samples from fetuses with hematological disorders was within the range of normal in all but 1 fetus infected with parvovirus in whom very few progenitor cells were detected. The frequency of progenitor cells in the blood did not change after intravascular red cell transfusion for alloimmunization despite the large volumes transfused, indicating that transfusion may have triggered a release of progenitor cells into the circulation. Progenitor cells in human fetal blood are present in distributions similar to those commonly detected in cord blood. Their total number in the circulating blood is in the same order used for pediatric and adult bone marrow transplantation. These results can be used to calculate the number of colony-forming cells which could be obtained from a fetus by in utero apheresis and which could be made available for autologous fetal gene therapy.

Circulating progenitor cells in human ontogenesis: response to growth factors and replating potential.

We have measured the number of progenitor cells circulating in fetal (17-32 weeks of gestation), perinatal (36 weeks of gestation), and adult (30-50 years old) blood. The progenitor cells at each ontogenetic stage were also characterized in terms both of the minimal combinations of growth factors they required to form maximal numbers of colonies in vitro and of their self-replication potential, as measured by the number of secondary and tertiary progenitor cells each could generate. The number of progenitor cells circulating in fetal and perinatal blood can be measured by directly plating the unfractionated blood. In this assay, fetal blood contains half the number of progenitor cells detected in perinatal blood (18.0 +/- 16.4 versus 40.88 +/- 0.63, p < 0.01), and the number of progenitor cells in adult blood is below the level of detection of the assay (< 1/8 microliter of blood). To compare the number of progenitor cells in all three stages of human development, progenitor cell counts were performed on blood mononuclear cells enriched by density separation. In this case, the light density cell fractions from fetal and neonatal blood contained the same number of progenitor cells (300/10(5) cells), numbers that were 10-fold higher than those observed with adult blood (30/10(5) cells). Circulating fetal-neonatal erythroid and multipotential progenitor cells were found to differ from their adult counterparts in terms of their response to growth factors and their self-renewal ability. In fact, the number of cytokines required to observe maximal colony formation increased with the ontogenetic stage of the cells. No differences were found in the frequency of primary colonies containing progenitor cells or in the mean number of secondary progenitor cells per primary colony in cultures of fetal, neonatal, or adult blood. Differences between the three ontogenetic stages, however, were found with respect to the number of sequential replatings that were possible. In fact, although both secondary granulocyte-macrophage (GM) and mixed-cell colonies derived from fetal cells gave rise to tertiary colonies, only perinatal secondary mixed-cell colonies grew in tertiary cultures, and no growth was observed in tertiary cultures of adult cells. These results suggest that the greater amplification of progenitor cells observed in liquid culture of fetal/neonatal versus adult blood is due both to a higher proliferative capacity of neonatal progenitor cells (up to two replatings versus one) and to a higher frequency in these samples of mixed-cell colony-forming cells (CFC) (37.7 +/- 7.3 versus 2.0 +/- 0.7/10(5) light density cells, respectively). Because of the high numbers of progenitor cells circulating in the fetus, as well as their high proliferative capacity, it is predicted that if blood could be harvested directly in utero, fetal blood would be as good a source of stem cells for transplantation as perinatal placental/cord blood. Circulating fetal stem cells would, therefore, represent an ideal target for gene therapy and in utero autologous transplantation.

Conflicts between physicians and patients in non-elective cesarean delivery: incidence and the adequacy of informed consent.

A study was undertaken in 372 consecutive patients undergoing non-elective cesarean delivery to explore the incidence and nature of conflicts between physician and patient surrounding the decision to undergo non-elective cesarean delivery; to examine the adequacy of informed consent at the time of non-elective cesarean delivery; and to describe the importance of a preventive ethics approach to non-elective cesarean delivery. During a 6-month interval, all patients who underwent non-elective cesarean delivery and their physicians were asked to take part in a survey in the early postpartum period concerning their response to recommendations for cesarean delivery. The survey included demographics as well as questions pertaining to informed consent and the presence and nature of patient-physician conflict. Of the 326 patients who were interviewed, 319 (98%) agreed to the recommendation for non-elective cesarean delivery and 7 patients (2%) initially disagreed. Reasons for disagreeing included: feared surgery (4 of 7), needed husband's approval (1 of 7), and questioned the medical necessity of surgery (2 of 7). In all 7 cases of initial disagreement, cesarean delivery was eventually performed with the patient's consent. The mean age of patients who initially disagreed was younger (24.7 +/- 6) than that of those who agreed (31.0 +/- 4 [p < 0.05]). Conflicts were present in 7 of 113 clinic patients and 0 of 213 private patients (p < 0.05). Of those surveyed, 26 (8.7%) indicated that they did not have adequate input in the decision for non-elective cesarean delivery. Patients with inadequate input expressed significantly more concerns with regard to the effect of surgery on their own health (p < 0.05) as well as its effect on the baby (p < 0.05). Our findings suggest that even though the incidence of physician-patient conflict about non-elective cesarean delivery was quite low, a significant number of patients (1 in 12) may have reservations concerning the informed consent process at the time of non-elective cesarean delivery. Patients with reservations are more likely to have greater concerns with regard to maternal and fetal risks, suggesting that a more detailed risk disclosure prior to the procedure is warranted for all pregnant patients. Perhaps by incorporating the preventive strategies discussed, the adequacy of informed consent and therefore the patient's autonomy could be enhanced, thus diminishing patient reservations and preventing physician-patient conflict in the intrapartum period.

Antiplatelet antibody testing in thrombocytopenic pregnant women.

OBJECTIVE: The purpose of the study was to attempt to distinguish pregnant women with gestational thrombocytopenia from those with idiopathic immune thrombocytopenia by eight different platelet antibody assays. STUDY DESIGN: Sera from pregnant women with presumed gestational thrombocytopenia (n = 160) and idiopathic immune thrombocytopenia (n=90) were prospectively tested for indirect and platelet-associated immunoglobulins G and M and complement C3, as well as for serotonin release. After the results were analyzed, a subset of patients were subsequently analyzed for circulating antiplatelet antibody directed against platelet membrane glycoprotein GPIIb/IIIa. RESULTS: Indirect immunoglobulin G was significantly greater in the 85 women with idiopathic immune thrombocytopenia than in the 129 women with gestational thrombocytopenia (p<0.001). Platelet-associated immunoglobulin G was elevated in the majority of women, both those with gestational thrombocytopenia and those with idiopathic immune thrombocytopenia. There were also no statistically significant difference in the values for platelet-associated C3 or indirect immunoglobulin M and C3. Levels of platelet-associated immunoglobulin M showed a tendency to be higher in women with gestational thrombocytopenia (p=0.04), as did the values in the serotonin release assay (p=0.06). CONCLUSION: Our data demonstrate that patients with gestational thrombocytopenia had surprisingly high levels of platelet-associated immunoglobulin despite mild thrombocytopenia. Comparison of a relatively large number of patients with idiopathic immune thrombocytopenia and gestational thrombocytopenia indicates that women with idiopathic immune thrombocytopenia cannot be distinguished from those with gestational thrombocytopenia by means of one or more of the prototypic platelet antiglobulin tests currently in use. Our preliminary data with glycoprotein-specific assays indicate that they may be more useful.

Neonatal hepatitis and excessive hepatic iron deposition following intrauterine blood transfusion.

The management of hemolytic disease has undergone a number of significant changes over the past few decades. Intrauterine transfusion therapy, particularly intravascular transfusions, have significantly reduced the morbidity and mortality associated with isoimmunization. This therapy results not only in the transfusion of blood, but also in the transfusion of iron. The long-term consequences of iron loading in the fetus are unknown. We report a case of a newborn with Rh hemolytic disease who was treated with in utero transfusions and subsequently developed liver disease consistent with iron overload.

Long-term generation of human mast cells in serum-free cultures of CD34+ cord blood cells stimulated with stem cell factor and interleukin-3.

The generation of murine mast cells is supported by several cytokines, and mast cell lines are frequently established in long-term cultures of normal murine marrow cells. In contrast, growth of human mast cells was initially dependent on coculture with murine fibroblasts. The growth factor produced by murine fibroblasts and required to observe differentiation of human mast cells is attributable in part to stem cell factor (SCF). However, other factors are likely involved. We have previously shown that the combination of SCF and interleukin-3 (IL-3) efficiently sustains proliferation and differentiation of colony-forming cells (CFCs) from pre-CFC enriched from human umbilical cord blood by CD34+ selection. With periodic medium changes and the addition of fresh growth factors, five consecutive cultures of different cord blood samples gave rise to differentiated cells and CFCs for more than 2 months. Although differentiated cells continued to be generated for more than 5 months, CFCs were no longer detectable by day 50 of culture. The cells have the morphology of immature mast cells, are Toluidine blue positive, are karyotypically normal, are CD33+, CD34-, CD45+, c-kit-, and c-fms-, and die in the absence of either SCF or IL-3. These cells do not form colonies in semisolid culture and are propagated in liquid culture stimulated with SCF and IL-3 at a seeding concentration of no less than 10(4) cells/mL. At refeedings, the cultures contain a high number (> 50%) of dead cells and have a doubling time ranging from 5 to 12 days. This suggests that subsets of the cell population die because of a requirement for a growth factor other than SCF or IL-3. These results indicate that the combination of cord blood progenitor and stem cells, plus a cocktail of growth factors including SCF and IL-3, is capable with high efficiency of giving rise in serum-deprived culture to human mast cells that behave like factor-dependent cell lines. These cells may represent a useful tool for studies of human mast cell differentiation and leukemia.

Rapid exclusion of chromosomal aneuploidies by fluorescence in situ hybridization prior to fetal surgery for obstructive uropathy--a case report.

Ultrasound of a fetus at 17 weeks gestation revealed posterior urethral valve syndrome with anhydramnios. Fluorescence in situ hybridization (FISH) to detect aneuploidies of chromosomes 13, 18, 21, X and Y was performed on transitional cells from the fetal bladder obtained at percutaneous vesicocentesis, followed by conventional cytogenetics. Fetal urine was chosen due to unavailability of amniotic fluid for karyotypic analysis. A nonlethal (disomic) karyotype was suggested by FISH, and thus placement of a vesicoamniotic shunt was performed. The ability to prognosticate in cases of obstructive uropathy is not absolute, and fetal surgery for relief of urinary obstruction is best performed at the earliest possible gestational age. Thus, all available means for rapidly ruling out lethal congenital anomalies should be undertaken in cases of obstructive uropathy prior to any decision regarding fetal surgery.

Stem cell factor and the amplification of progenitor cells from CD34+ cord blood cells.

We have studied the frequency of colony-forming cells (CFC) in fetal and neonatal blood in comparison with adult blood and marrow. Fetal or neonatal blood contains at least as many CFC as adult marrow and higher numbers of the more primitive CFC--those CFC (mixed-cell CFC) giving rise to colonies composed of erythroid and myeloid cells. CD34+ cord blood cells (selected by one of several means) proliferate in culture over time and generate more CFC (from pre-CFC) and differentiated cells in response to stem cell factor (SCF) plus different hematopoietic growth factors. For its effect, SCF requires the synergistic action of erythropoietin (Epo), granulocyte colony-stimulating factor (G-CSF), or interleukin-3 (IL-3). In the presence of Epo or G-CSF, CFC and differentiated cells are generated for 15 days and are mainly erythroid or granulocytic, respectively. In contrast, SCF plus IL-3 generate multilineage CFC and differentiated cells for more than 1 month. When the conditions for these long-term suspension cultures were optimized, CFC and differentiated cells were generated for more than 2 months. At this time, CFC were no longer detectable, but cells continued to be generated, and the cells had a mast cell phenotype. These cells have been maintained and propagated for more than 8 months in the presence of IL-3 and SCF and may represent a useful tool to study human mast cell differentiation.

Long-term generation of colony-forming cells (CFC) from CD34+ human umbilical cord blood cells.

Human umbilical cord blood cells represent a potential alternative to bone marrow as a source of stem and progenitor cells for allogeneic transplantation. Therefore, many studies are underway to evaluate the number of cord blood stem cells and their amplification potential. We analyze here the amplification potential of CD34+ cord blood cells in liquid cultures stimulated with stem cell factor (SCF) in combination with interleukin-3 (IL-3), erythropoietin (Epo) or granulocyte colony-stimulating factor (G-CSF) under serum-deprived conditions. We report that under certain circumstances (stimulation with SCF and IL-3, replacing of the medium and growth factors every 3-4 days, no change of the initial culture flask, 37 degrees C as incubation temperature), CD34+ cells give rise to differentiated cells and progenitor cells for more than two months. During this period, more than 10(10) differentiated cells and 10(6) progenitor cells are generated from 0.25-1 x 10(4) CD34+ cells in the absence of a stromal layer. These data highlight the high proliferative and differentiative potential of cord blood stem cells and, because the culture procedures are relatively simple and do not require a stromal layer, open the way to the clinical use of ex vivo stem cell expansion.

Aspects of the biology of the neonatal hematopoietic stem cell.

We have studied the frequency of colony forming cells (CFC) in fetal and neonatal blood in comparison with adult blood and marrow. Fetal/neonatal blood contains at least as many CFC as adult marrow and higher numbers of the more primitive CFC--those CFC giving rise to colonies composed of erythroid and myeloid cells. CD34+ cord blood cells (selected either by sorting, panning or affinity chromatography) proliferate in culture over time and generate more CFC (from pre-CFC) and differentiated cells in response to Steel factor plus different hematopoietic growth factors. Steel factor is unable to stimulate cell growth by itself under serum-deprived conditions and requires the synergistic action of erythropoietin (Epo), granulocyte colony stimulating factor (G-CSF) or interleukin 3 (IL-3). In the presence of Epo or G-CSF, CFC and differentiated cells are generated for 15 days and are mainly erythroid or granulocytic, respectively. In contrast, Steel factor plus IL-3 generates multilineage CFC and differentiated cells for more than one month. When the conditions for these long-term suspension cultures were optimized (37 degrees C, regular refeeding with fresh growth factors and media without changing the flask), CFC and differentiated cells were generated for more than two months. At this time, CFC were no longer detectable and all cells had a mast cell phenotype. These cells have been maintained and propagated for more than eight months in the presence of IL-3 and Steel factor and may represent a useful tool to study human mast cell differentiation.(ABSTRACT TRUNCATED AT 250 WORDS)