Various types of fetal cells have been found in the maternal blood, trophoblasts, lymphocytes, granulocytes, and nucleated red blood cells (RBCs). Among those, nucleated RBCs appear to have the greatest potential of being used for prenatal genetic analysis. Therefore, efforts have been made to recover fetal nucleated RBCs from maternal circulation, a relatively noninvasive procedure that obviates any fetal risk.1–3 Minimizing the risks to mothers and fetuses is essential in developing tools for prenatal genetic diagnosis.
Despite numerous attempts to collect fetal nucleated RBCs from maternal blood, information on levels of nucleated RBCs in maternal blood is far from sufficient. Nearly all existing data were derived from cross-sectional studies, which are not very powerful when addressing interindividual variability and cannot assess changes over time. Therefore, we conducted a longitudinal study to estimate the number of nucleated red blood cells in different stages of gestation.
Materials and Methods
Among the primigravidas with blood type O who visited our prenatal, genetic clinic at National Cheng Kung University Hospital between July 1, 1996 and December 31, 1997, 50 agreed to enroll in our research project at their first antenatal visits. All participants were Chinese, positive for the Rh(D) antigen, and none had a history of blood transfusions. Ten candidates completed the study protocol but were excluded from analysis because they had significantly elevated alpha-fetoprotein or β-hCG levels, received invasive diagnostic procedures, or subsequently developed various pregnancy complications, such as fetal anomalies, preterm labor, fetal growth restriction, vaginal bleeding, or hypertension. Two candidates who dropped out in mid-study for personal reasons were also excluded from analysis.
From each of the remaining 38 participants, 10 mL of peripheral venous blood was collected in heparinized tubes at 5–10, 11–14, 15–24, 25–32, and 33–39 weeks' gestation, and 6 weeks and 3 months postpartum. The first three intervals were chosen to represent the timing before invasive diagnosis (5–10 weeks); early amniocentesis or chorionic villus sampling (11–14 weeks); and second trimester amniocentesis (15–24 weeks). Of the 38 participants enrolled for data analysis, 25 completed the study protocol, and a total of 175 blood samples were analyzed. Although the other 13 participants agreed to complete the study protocol, each of them missed a sampling stage, thus only six samples were collected from each (78 total blood samples). All blood samples were drawn with the informed consent of the participants.
Nucleated RBCs were isolated using a triple-density gradient with Histopaque (Sigma Diagnostics, St. Louis, MO),4 which was developed by layering 2.5 mL of Histopaque-1119 at the bottom, 2.5 mL of Histopaque-1107 in the middle, and 2.5 mL of Histopaque-1077 on the top. The blood was diluted with two times its volume with phosphate-buffered saline. The diluted blood was overlayered onto the Histopaque-1077, then spun at 800g for 30 minutes at room temperature. Cells at the interface of 1077 and 1107 were collected, washed twice with phosphate-buffered saline, and resuspended in phosphate-buffered saline with 3% fetal calf serum. The number of live cells was counted in a hemocytometer chamber. Dead cells were identified by trypan blue staining and excluded from analysis. The average numbers of nucleated cells in a 10-mL sample before and after separation were 1.92 × 107 (range 1.07 × 107 to 2.84 × 107) and 0.29 × 107 (range 0.06 × 107 to 0.73 × 107), respectively.
After separation, blood cells were spun at 200g for 5 minutes in a cytocentrifuge (Shadon, Frankfurt, Germany). Approximately 105 to 106 nuclei were plated onto each poly-L-lysine-coated slide, and 5 to 30 cytospin preparations were made for each case. The slides were stained with Kleihauser-Betke acid stain using the modified technique described by Los et al5 to detect fetal hemoglobin.
The samples were examined microscopically by an experienced technician who was masked to the identities and clinical information of participants. The frequency of nucleated RBCs was obtained by dividing the number of nucleated RBCs by the number of all nucleated cells in the blood sample and was expressed per 107 nucleated cells.
The χ2 test with continuity adjustment and two-tailed Student t test were used to compare the differences between groups of participants (those with and without complete sampling). The changes in the frequency of nucleated RBCs over time were assessed for each individual, and repeated-measures analysis of variance was used to determine the statistical significance of the changes. If repeated-measures analysis of variance showed significant changes over time, two-tailed paired t test was used to evaluate the differences between each two consecutive stages. We also grouped the participants according to gender of the fetus and their ABO blood type compatibility with the fetus and compared the average frequency of nucleated RBCs between the subgroups by two-tailed Student t test. Repeated-measures analysis of variance followed by two-tailed, paired t test was applied to evaluate statistical significance of changes in frequency of nucleated RBCs over time.
In most participants, nucleated blood cells were detectable at the first test stage (5–10 weeks' gestation) until 6 weeks postpartum. The frequency of nucleated RBCs increased as gestation advanced in all cases except two in which nucleated RBC counts decreased slightly in the last gestation stage (33–39 weeks' gestation). After delivery, nucleated RBC levels decreased rapidly and became undetectable at 3 months postpartum in all participants except three.
The mean age was 28.4 years (standard error [SE] 0.7 years) for participants with complete sampling and 29.2 years (SE 0.8 years) for participants without complete sampling, not a statistically significant difference (P = 0.4; two-tailed Student t test). There was no significant difference between groups in proportion of male fetuses (48.0% versus 53.9%; P = 1.0 for χ2 test with continuity adjustment), the proportion of participants with ABO blood type compatibility with the fetuses (40.0% versus 38.5%; P = 1.0 for χ2 test with continuity adjustment), average frequency of nucleated RBCs in any stage, and average difference in frequencies of nucleated RBCs between any two consecutive stages. Because the groups were quite similar, their data were pooled for further analyses.
We found that increases in nucleated RBCs from one stage to the next and over the whole of pregnancy were all statistically significant and that the decreases between every two consecutive stages after the last stage of pregnancy were all statistically significant (Table 1).
Among the 38 participants, 19 had male fetuses, and 15 had compatible ABO blood types with their fetuses. The frequency of nucleated RBCs was similar between participants with male and female fetuses, and between those with compatible and those with incompatible ABO blood types with their fetuses in all seven stages (Tables 2 and 3). The changes in nucleated RBC frequency from one stage to the next were similar regardless of gender of the fetus and compatibility of ABO blood type (Tables 2 and 3). After participants were grouped by gender of fetus or compatibility of blood type, we still found statistically significant changes in nucleated RBC number over time (Tables 2 and 3).
Although it is clear that nucleated RBCs are present in maternal blood, there is no consensus on expected frequency of nucleated RBCs in normal pregnancies. One study found nucleated RBCs in maternal circulation as early as the first trimester in 10% of women.1 There was an increase in number during the second and third trimesters, with the prevalence of 20% and 40%, respectively.1 More sophisticated methods for isolating nucleated RBCs, such as fluorescence-activated cell sorting, magnetic-activated cell sorting, charged-flow separation, and density gradients, have been reported.2,3 Most of those experiments were done on maternal peripheral blood samples collected in the first or early second trimester of pregnancy. The levels of nucleated RBCs in the present study were compatible with those observed by Takabayashi et al6 in a cross-sectional study. We found an increase in frequency of nucleated RBCs as gestation advanced and a rapid decrease after delivery. Among our participants, nucleated RBCs became undetectable at 3 months postpartum, with very few exceptions, which is also consistent with findings in previous studies.1–3 In our study, the choice of simple density gradient centrifugation without subsequent sorting procedures prevented additional cell losses. The prospective study design and masking of the examiner resulted in unbiased estimation of nucleated RBC levels. Selection of primigravidas excluded any possible effect of previous pregnancies.
Many factors contribute to the presence of nucleated RBCs in the peripheral blood of pregnant women. The increase of nucleated RBCs in maternal blood should be accounted for by increases in maternal and fetal erythrocytes. During pregnancy, maternal erythropoiesis is enhanced by increasing blood volume and oxygen demand, and blood from pregnant women was found to contain a large number of erythroid progenitors.7 Fetal cells in maternal blood might increase with advancing gestation, in abortion, after invasive examinations, and in pregnancies with toxemia or abnormal fetal karyotype.1–3,8 Candidates with any of those conditions were excluded from our study, so our data represented the conditions of normal pregnant women.
We found that the frequency of nucleated RBCs was not related to compatibility of ABO blood types between mothers and fetuses. A concern was raised that Rh or ABO incompatibilities between mothers and fetuses might affect detection of fetal nucleated RBCs in maternal blood. It was found that ABO-compatible pregnancies were more likely to have circulating fetal cells in maternal peripheral blood than ABO-incompatible pregnancies, presumably due to lysis of incompatible fetal cells.1,9–11 Simpson and Elias2 argued that maternal-fetal ABO incompatibility is probably not a major factor because fetal RBCs contain fewer A or B sites than adult cells, and fetal antigens are not as specific as adult antigens. Our findings seem to support their theory. The effect of Rh incompatibility on levels of nucleated RBCs could not be tested in our study because nearly all (99.67%) Taiwanese are positive for the Rh(D) antigen.12
Attempts to differentiate between maternal and fetal nucleated RBCs have been made with special stains. For example, the Kleihauer-Betke test identifies fetal hemoglobin-containing cells, and the immunohistochemical stain used monoclonal antibodies to target fetal hemoglobins.3 The specificity of fetal hemoglobin (α2γ2) in fetal cells is mitigated by the fact that fetal hemoglobin is also expressed in a small portion (up to 1.0%) of adult erythrocytes.13 Therefore, the nucleated RBCs detected in maternal circulation might be a mixture of cells from mother and fetus. Data on proportion of fetal cells among nucleated RBCs in maternal blood are not consistent across studies. Slunga-Tallberg et al14 found that all nucleated RBCs detected in pregnant women were of maternal origin in the second trimester, which did not agree with other studies.15,16 In a cross-sectional study based on quantitative polymerase chain reaction, our research team found that a substantial proportion of nucleated RBCs were of fetal origin before 24 weeks' gestation, whereas in late gestation the majority of nucleated RBCs might be of maternal origin.17 Based on the assumption that most nucleated RBCs during the first and second trimesters are from the fetus, the proportion of fetal nucleated RBCs should be close to 1 in 106 to 1 in 107 maternal nucleated cells, which is similar to that estimated by previous studies.2,3 In the present study, we found that the proportion of nucleated RBCs varied significantly among individuals.
A major limitation of our study was that the number of participants was relatively small. It still had enough power to detect the changes in the frequency of nucleated RBCs over time, even after segregating the participants by gender of fetus and maternal-fetal ABO blood-type compatibility. None of the differences between the subgroups defined by the fetal gender or maternal-fetal blood-type compatibility was statistically significant, so study power might be a concern in making those comparisons (Tables 2 and 3). However, most of the differences were around or even less than 1 per 107 nucleated cells, which could hardly be of clinical significance. For the differences that were much larger, our study had the least power in comparing level of nucleated RBCs between participants with male and female fetuses at 25–32 weeks' gestation (Table 2). Our study still had 75% power to detect a difference of 10% of the smaller value for such a comparison (a difference of 6.92 per 107 nucleated cells), which is very close to the 80% power that is generally considered adequate. No constant trend was observed in the comparisons of subgroups defined by fetal gender or blood-type compatibility. Therefore, even though our sample was relatively small, we had adequate study power to address our concerns.
Whereas the objective of this study was to determine overall numbers of nucleated RBCs, only those of fetal origin are of prenatal diagnostic value. Therefore, more studies that isolate and identify fetal cells from maternal blood are needed before we can use nucleated RBCs in maternal blood for prenatal genetic screening.
1. Chueh J, Globus MS. Prenatal diagnosis using fetal cells in the maternal circulation. Semin Perinatal 1990;14:471–82.
2. Simpson JL, Ellas S. Isolating fetal cells from maternal blood: Advances in prenatal diagnosis through molecular technology. JAMA 1993;270:2357–61.
3. Bianchi DW. Prenatal diagnosis through the analysis of fetal cells. In: Milunsky A, ed. Genetic disorders and the fetus. 4th ed. Baltimore, MD: The Johns Hopkins University Press, 1988:931–51.
4. Bhat NM, Bieber MM, Teng NNH. One-step enrichment of nucleated red blood cells. J Immunol Methods 1993;158:277–80.
5. Los FJ, de Wolf B, Huisjes HJ. Raised maternal serum-alpha-fetoprotein levels and spontaneous fetomaternal transfusion. Lancet 1979;2:1210–2.
6. Takabayashi H, Kuwabara S, Ukita T, Yamafuji K, Ikawa K. Appearance of nucleated erythrocytes in maternal blood. Acta Obstet Gynaecol Jpn 1997;49:149–52.
7. Howells MR, Jones SE, Napier JA, Saunders K, Cavill I. Erythropoiesis in pregnancy. Br J Haematol 1986;64:595–9.
8. 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;61:822–9.
9. Creger WP, Steele MR. Human fetomaternal passage of erythrocytes. N Engl J Med 1957;256:158–61.
10. Cohen F. Zuelzer WW. Transplacental passage and postnatal survival of fetal erythrocytes in heterospecific pregnancies. Blood 1967:30:796–804.
11. Zipursky A, Hull A, White FD. Fetal erythrocytes in the maternal circulation. Lancet 1959;1:451–2.
12. Lin M. Blood groups and transfusion medicine in Taiwan. J Formos Med Assoc 1997;96:933–42.
13. Turpeinen U, Stenman UH. Determination of fetal hemoglobin by time-resolved immunofluorometric assay. Clin Chem 1992;38:2013–8.
14. Slunga-Talberg A, El-Rafai W, Keinanen M, Ylinen K, Kurki T, Klinger K, et al. Maternal origin of nucleated erythrocytes in peripheral venous blood of pregnant women. Hum Genet 1995;96:53–7.
15. von Eggeling F, Michel S, Gunther M, Schimmel B, Claussen U. Determination of the origin of single nucleated cells in maternal circulation by means of random PCR and a set of length polymorphisms. Hum Genet 1997;99:266–70.
16. Elias S, Price J, Dockter M, Watchtel S, Tharapel A, Simpson JL. First trimester prenatal diagnosis of trisomy 21 in fetal cells from maternal blood. Lancet 1992;340:1033–4.
17. Kuo PL. The frequencies of fetal nucleated red blood cells in maternal blood during different stages of gestation. Fetal Diagn Ther 1998;13:375–9.