Serum free ß-HCG and alpha-fetoprotein levels in IVF, ICSI and frozen embryo transfer pregnancies in maternal mid-trimester serum screening for Down's syndrome

R. Räty1,9, A. Virtanen4, P. Koskinen2, L. Anttila1,5, J. Forsström3, P. Laitinen6, P. Mörsky7, A. Tiitinen8 and U. Ekblad1

1 Departments of Obstetrics and Gynaecology, 2 Central Laboratory, 3 Medical Informatics Research Centre in Turku (MIRCIT), University of Turku, Kiinamyllynkatu 4–8, FIN-20520 Turku, 4 Department of Central Laboratory and Rehabilitation Research Centre of the Social Insurance Institution, Peltolantie 3, FIN-20720 Turku, 5 The Family Federation of Finland, Maariankatu 3A, FIN-20100 Turku, 6 Central Laboratory, Oulu University Hospital, P.O.Box 22, 90221 Oulu, 7 Department of Central Laboratory, Tampere University Hospital, P.O.Box 2000, 33521 Tampere and 8 Department of Obstetrics and Gynaecology, University of Helsinki, Haartmaninkatu 2, P.O.X. 140, HUS, FIN-00029 Helsinki, Finland


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The aim of this study was to compare the maternal mid-trimester free ß-HCG and alpha-fetoprotein (AFP) levels in pregnancies conceived by assisted reproduction technology and spontaneous pregnancies in Down's syndrome screening. The influence of the number of embryos transferred and the amount of gonadotrophins used on the marker levels was also evaluated. METHODS: The study population consisted of 58 IVF, 32 ICSI and 26 frozen embryo transfer (FET) singleton pregnancies. The levels of ß-HCG and AFP were compared with the control group of 6548 singleton spontaneous pregnancies. RESULTS: The false positive rate (FPR) in the Down's syndrome screening was 19% overall in assisted reproductive technology pregnancies, being highest (30.8%) in the FET group. The free ß-HCG multiples of the median (MoM) values were statistically significantly elevated only in the FET group (1.33 MoM; P = 0.012). A positive correlation between the number of embryos transferred and the marker levels was observed in the IVF group. No correlation was found between the amount of gonadotrophin medication used and the marker levels. CONCLUSIONS: The present data confirm that the overall FPR in the serum screening for Down's syndrome in assisted reproduction pregnancies is high, resulting in unnecessary invasive procedures.

Key words: AFP/ß-HCG/Down's syndrome/FET/ICSI/IVF


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Several studies have shown that the false positive rate (FPR) in serum Down's syndrome screening is higher in IVF singleton pregnancies than in spontaneous pregnancies (Heinonen et al., 1996Go; Ribbert et al., 1996Go; Frishman et al., 1997Go; Wald et al., 1999Go; Maymon and Shulman, 2001Go). In IVF singleton pregnancies the levels of HCG are often elevated, whereas alpha-fetoprotein (AFP) levels are the same or slightly lower as compared with normal spontaneous pregnancies (Heinonen et al., 1996Go; Frishman et al., 1997Go; Lam et al., 1999Go; Wald et al., 1999Go; Maymon and Shulman, 2001Go). Analogously, we have recently shown that the ß-HCG levels are also significantly higher in IVF twin pregnancies than in spontaneous twin pregnancies (Räty et al., 2000Go). In contrast, it has been reported that both the HCG and AFP levels were slightly lower in IVF pregnancies than in the controls, and that the FPR was not altered between the groups (Barkai et al., 1996Go). In a recent study, it was reported that in pregnancies initiated by ICSI the second trimester serum levels of AFP were significantly lower than in controls or in IVF pregnancies (Lam et al., 1999Go).

The reason behind the marker level alterations in assisted reproduction pregnancies has remained unknown. The fact that the median age of women becoming pregnant with assisted reproduction is higher than in spontaneous pregnancies cannot alone explain the higher FPR (Heinonen et al., 1996Go; Wald et al., 1999Go). Recently, it was suggested that the high FPR in IVF and oocyte donation pregnancies could reflect factors associating with the pregnancy itself and not with the assisted reproduction procedure (Maymon and Shulman, 2001Go).

The aim of this study was to evaluate the effects of the number of embryos transferred and the amount of gonadotrophin medication for ovarian stimulation for IVF and ICSI on maternal serum levels of ß-HCG and AFP in the second trimester Down's syndrome screening. We hypothesized that the ovulation stimulation regimens may have an effect both on the ovum and the maturation of the endometrium, and thereby on the implantation of the embryo on the uterus wall. To test this hypothesis, we studied the marker levels in frozen embryo transfer (FET) pregnancies, where ovarian stimulation was not used, but had affected the ovum prior to fertilization.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The study population consisted of IVF (n = 58), ICSI (n = 32) and FET (n = 26) singleton pregnancies with known normal outcome. The ovarian stimulation protocols used included down-regulation by GnRH agonist followed by gonadotrophin stimulation. The stimulation was monitored by serum estradiol levels and by ultrasound examination of the ovaries. The oocyte retrieval was performed 36 h after HCG injection (5000–10000 IU). In the IVF group, oocytes were inseminated with motile spermatozoa, and in the ICSI group a single spermatozoon was injected into the oocyte. The embryo transfer was performed 48 h after retrieval. Progesterone (400–600 mg) was given either orally or vaginally for up to 8 weeks of pregnancy. In the FET group, the embryo transfer was performed either in a natural cycle (n = 22) or during estrogen substitution (n = 4, transdermal estrogen 100 µg/day or oral estradiol valerate 4 mg/day). In FET cycles, the progesterone substitution was similar to that in IVF/ICSI cycles.

Screening for Down's syndrome was performed by analysing maternal serum levels of ß-HCG and AFP between pregnancy weeks 14 and 18, and the risk ratio for Down's syndrome was counted with Wald's formula (Wald et al., 1994Go). The data were collected from different registers: birth register, hospital registers, malformation register, legal abortion register and laboratory registers in Finland between years 1994 and 1999. The control group consisted of 6548 singleton pregnancies with known normal outcome and no congenital anomalies. The information on the number of embryos transferred and the medication used for ovarian stimulation was collected from the patient documents. The data concerning the number of embryos transferred was missing from four IVF pregnancies and one ICSI pregnancy.

Serum AFP and ß-HCG were analysed with dual-labelled time-resolved immunofluorometric assay (hAFP/free HCG beta Dual kit; Wallac, Turku, Finland).

The differences in the levels of the two analyses between the different study groups and the control group were evaluated with the Kruskal–Wallis and Wilcoxon tests. An analysis of co-variance (ANCOVA) for log-transformed multiples of the median (MoM) values was used to adjust the age differences between the groups. Spearman's correlation coefficient was used to assess whether the number of embryos transferred or the amount of medication used had an effect on serum ß-HCG or AFP levels. SAS 6.12 for Windows was used in all calculations.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The demographic data on the study groups are shown in the Table IGo. The subjects in the assisted reproduction groups were older than controls (P = 0.0001).


View this table:
[in this window]
[in a new window]
 
Table I. The demographic data on the IVF, ICSI, frozen embryo transfer (FET) groups and the controls
 
The FPR among the controls was 5%, whereas among assisted reproduction pregnancies (n = 116) it was 19% overall, being 17.2% in the IVF group, 12.5% in the ICSI group and 30.8% in the FET group. After exclusion of the four cases on estrogen substitution, the FPR in the FET group was 22.7%. In the overall comparison, a statistically significant difference in the free ß-HCG MoM levels was seen between the three study groups and the control group (P = 0.018). No difference in the ß-HCG MoM levels between the three study groups was observed by the Kruskall–Wallis test. The age difference between study subjects and controls did not explain the difference in ß-HCG MoM values, which remained statistically significant after the age adjustment (P = 0.0017).

The post-hoc comparisons of the serum MoM levels of free ß-HCG were made between the study groups and the controls: a statistically significant difference in the ß-HCG levels was observed only between the FET group and the controls (P = 0.012) (Table IIGo). After exclusion of the four cases on estrogen in the FET group, the ß-HCG level remained the highest among the study groups, but the difference did not reach statistical significance. There were no differences in AFP levels between any of the study groups and the controls (Table IIGo).


View this table:
[in this window]
[in a new window]
 
Table II. Free ß-HCG and alpha-fetoprotein (AFP) multiple of the median (MoM) values with median differences between the study group and the controls
 
Table IIIGo summarizes the marker levels according to the number of transferred embryos. A weak positive correlation between the number of transferred embryos and marker levels was observed only in the IVF group (Table IIIGo). The amount of gonadotrophins used did not differ between the IVF (range 450–4800 IU) and ICSI (range 950–3750 IU) groups (Table IGo). No correlation was found between the marker levels and the amount of gonadotrophins used.


View this table:
[in this window]
[in a new window]
 
Table III. The serum free ß-HCG and alpha-fetoprotein (AFP) multiple of the median (MoM) values according to the number of embryos transferred in IVF, ICSI and frozen embryo transfer (FET) groups
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In many countries, screening for Down's syndrome is still performed by measuring maternal serum levels of HCG and AFP at second trimester, although first trimester screening either biochemically or by ultrasound has increased in recent years. The detection rate of Down's syndrome in serum screening is 60–66% with 5% FPR (Wald et al., 1988Go), whereas ultrasound screening detects 77–81% of the affected pregnancies (Pandya et al., 1995Go; Economides et al., 1998Go). In our material of assisted reproduction pregnancies, the FPR was 19%, being the highest in the FET group (30.8%). According to our data, the higher FPR in Down's syndrome screening in assisted reproduction pregnancies was due to elevated free ß-HCG levels. In the present material, the levels of ß-HCG were elevated both in the IVF and ICSI pregnancies, but the elevation was more obvious in the IVF group. The AFP levels remained nearly at the same level as in the controls. Our results are in line with earlier studies concerning IVF pregnancies (Heinonen et al., 1996Go; Ribbert et al., 1996Go; Frishman et al., 1997Go; Wald et al., 1999Go), but opposite to the results of Lam and co-workers who reported slightly lower HCG and significantly diminished AFP levels in ICSI pregnancies as compared with the controls (Lam et al., 1999Go). No good explanation is currently available on the differences in the marker levels among ICSI pregnancies between the present study and the study of Lam et al. Clinical characteristics of the study populations are comparable (mean maternal age, gestational age at sampling), but the number of cases in both studies was relatively small (Lam et al., 1999Go: n = 23; our study, n = 32). As the former study was performed on Asian females, and ours on Europeans, ethnic differences might be one explanation for the differences in the marker levels. However, higher HCG levels have been reported previously in Down's syndrome screening of Asian populations than of white, black African or North African populations (Muller et al., 1994Go), which makes interpretation of these data even more difficult.

To our knowledge, the effect of the number of embryos transferred on the marker levels has not been evaluated previously, although an early implantation of several embryos may interfere with the marker levels and thereby cause the false positive results in Down's syndrome screening in assisted reproduction pregnancies. We found a weak positive correlation between the number of embryos transferred and the serum marker levels only in the IVF group. However, the statistical power of this analysis is diminished by the fact that both one- and four-embryo transfers were rarely performed in Finland during years 1994–1999, which makes the sizes of the corresponding subgroups small. According to our data, the number of embryos transferred may have an influence on the marker levels and should be re-evaluated with larger data. Although an early ultrasound examination was performed around week 7 of pregnancy to exclude multiple pregnancies, early implantation of several embryos could affect HCG levels even during second trimester screening.

One of our basic hypotheses was that the dose of gonadotrophin used for ovarian stimulation may cause alterations in the marker levels in assisted reproduction pregnancies. However, we failed to find any association between the levels of maternal serum markers and the amount of ovarian stimulation medication. Unfortunately, we were unable to evaluate the effect of actual ovarian response during stimulation on the marker levels. One may speculate that in maternal mid-trimester serum screening an increased ß-HCG level can occur as a result of an excessive ovarian response, where multiple follicles and corpora lutea develop. However, the FET group, which we assumed to be the closest to the controls in marker levels, exhibited the highest serum HCG MoM levels and highest FPR (30.8%). Instead of ovulation stimulation, the mechanism behind elevated serum HCG levels in the FET group may be due to a placentation failure that could result in changes in trophoblast function and thus in HCG production. Among our study population the implantation rate was slightly lower in the FET group (49%) than in the IVF (54%) or the ICSI (51%) groups. This is in line with the higher pregnancy rate after transfers of fresh than cryopreserved embryos. Interestingly, after exclusion of subjects using estrogen, the ß-HCG levels diminished slightly. It has been suggested that some unidentified endocrine/metabolic errors might explain sub/infertility. These factors may affect the serum screening results in assisted reproduction pregnancies and may also be causative for the complications among these pregnancies (McElrath and Wise, 1997Go; Maymon and Shulman, 2001Go).

It has been speculated that the high HCG levels in assisted reproduction pregnancies may be caused by the higher progesterone levels in these pregnancies than in spontaneous pregnancies (Wald et al., 1999Go). These workers suggested that the reason for high progesterone levels could be multiple follicle development in IVF and ICSI pregnancies and subsequent formation of multiple corpora lutea, which could in turn increase the production of HCG from the placenta. The results obtained with the FET group in the present study and the recent data concerning pregnancies after oocyte donation (Maymon and Shulman, 2001Go) are confusing with regard to this theory, because in FET and oocyte donation pregnancies excessive follicles or corpora lutea do not exist. In the FET cases on estrogen replacement therapy no corpus luteum is involved, but progesterone substitution is used up to week 8 of pregnancy, as in all IVF/ICSI/FET protocols of the present study. Since the half-life of natural progesterone is only 4 h, we assumed that progesterone substitution cannot interfere the mid-trimester (week 14–18 of pregnancy) marker levels in assisted reproduction pregnancies. It was also recently shown that in IVF pregnancies serum levels of progesterone, but not HCG, are significantly higher than in normal pregnancies up to 8 weeks post-conception (Costea et al., 2000Go).

According to the present study, the FPR in Down's syndrome screening in assisted reproduction pregnancies is higher than in spontaneous pregnancies. This results in an increased need for invasive procedures, e.g. amniocentesis, which carries 0.5–1% risk of abortion. Even this small risk can be too much in these valuable pregnancies where infertility may have lasted for many years.


    Notes
 
9 To whom correspondence should be addressed. E-mail: raija.raty{at}pp.inet.fi Back

Submitted on April 2, 2001


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Barkai, G., Goldman, B., Ries, L. et al. (1996) Down's Syndrome screening marker levels following assisted reproduction. Prenat. Diagn., 16, 1111–1114.[ISI][Medline]

Costea, D.M., Gunn, L.K., Hargreaves, C. et al. (2000) Delayed luteo–placental shift of progesterone production in IVF pregnancy. Int. J. Gynaecol. Obstet., 68, 123–129.[ISI][Medline]

Economides, D.L., Whitlow, B.J., Kadir, R. et al. (1998) First trimester sonographic detection of chromosomal abnormalities in an unselected population. Br. J. Obstet. Gynaecol., 105, 58–62.[ISI][Medline]

Frishman, G.N., Canick, J.A., Hogan, J.W. et al. (1997) Serum triple marker screening in in vitro fertilization and naturally conceived pregnancies. Obstet. Gynecol., 90, 98–101.[Abstract/Free Full Text]

Heinonen, S., Ryynänen, M., Kirkinen, P. et al. (1996) Effect of in vitro fertilization on human chorionic gonadotropin serum consentrations and Down's Syndrome screening. Fertil. Steril., 66, 398–403.[ISI][Medline]

Lam, Y.H., Yeung, W.S.B., Tang, M.H.Y. et al. (1999) Maternal serum alpha-fetoprotein and human chorionic gonadotropin in pregnancies conceived after intracytoplasmic sperm injection and conventional in-vitro fertilization. Hum. Reprod., 14, 2120–2123.[Abstract/Free Full Text]

Maymon, R. and Shulman, A. (2001) Comparison of triple serum screening and pregnancy outcome in oocyte donation versus IVF pregnancies. Hum. Reprod., 16, 691–695.[Abstract/Free Full Text]

McElrath, T.F. and Wise, P.H. (1997) Fertility therapy and the risk of very low birth weight. Obstet. Gynecol., 90, 600–605.[Abstract/Free Full Text]

Muller, F., Bussieres, L., Pelissier, M-C. et al. (1994) Do racial differences exist in second trimester maternal hCG levels? A study of 23369 patients. Prenat. Diagn., 14, 633–636.[ISI][Medline]

Pandya, P.P., Snijders, R.J.M., Johnson, S.P. et al. (1995) Screening for fetal trisomies by maternal age and fetal nuchal translucency thickness at 10 to 14 weeks of gestation. Br. J. Obstet. Gynaecol., 102, 957–962.[ISI][Medline]

Räty, R., Virtanen, A., Koskinen, P. et al. (2000) Maternal midtrimester serum AFP and free ß-hCG levels in in vitro fertilization twin pregnancies. Prenat. Diagn., 20, 221–223.[ISI][Medline]

Ribbert, L.S., Kornman, L.H., De Wolf, B.T. et al. (1996) Maternal serum screening for fetal Down syndrome in IVF pregnancies. Prenat. Diagn., 16, 35–38.[ISI][Medline]

Wald, N.J., Cuckle, H.S., Densem, J.W. et al. (1988) Maternal serum screening for Down syndrome in early pregnancy. Br. Med. J., 297, 883–887.[ISI][Medline]

Wald, N.J., Densem, J.W., Smith, D. et al. (1994) Four-marker serum screening for Down's Syndrome. Prenat. Diagn., 14, 707–716.[ISI][Medline]

Wald, N.J., White, N., Morris, J.K. et al. (1999) Serum markers for Down's Syndrome in women who have had in vitro fertilisation, implications for antenatal screening. Br. J. Obstet. Gynaecol., 106, 1304–1306.[ISI][Medline]

accepted on October 3, 2001.