Congenital malformations after assisted reproduction: risks and implications for prenatal diagnosis and fetal medicine

Abstract

The number and proportion of children born after assisted conception are increasing, both in the UK and internationally. Worldwide there have already been over 3 500 000 births resulting from assisted reproductive technology (ART), and with falling fertility in some countries rates are likely to rise1. In 2006, 1.7% of all children born in England and Wales were conceived after ART, compared with 0.5% in 19922. Whilst the development of ART has evidently been of great benefit for millions of couples worldwide, the growing numbers of children born as a result make it increasingly important that potential risks to these children are investigated. ART has developed rapidly since the birth in 1978 of Louise Brown, the first infant conceived by in-vitro fertilization (IVF). The diverse range of techniques now available includes gamete intrafallopian tube transfer (GIFT), oocyte donation, embryo cryopreservation, intracytoplasmic sperm injection (ICSI), preimplantation genetic diagnosis and blastocyst culture (extended embryo culture) as well as standard IVF. The possibility of an association between ART and congenital malformations was raised as early as 1987, when Lancaster6 reported a greater than expected number of children with neural tube defects and transposition of the great arteries born to mothers after ART. Subsequent initial prospective studies reported inconsistent results7, 8. This inconsistency may have been the result of limitations in the methodology of these studies, most notably small sample size and hence low statistical power, particularly when investigating relatively rare outcomes. In addition, differing definitions of congenital malformations, inclusion or otherwise of spontaneous and induced abortions and ascertainment bias have contributed to the difficulty in reaching consensus between studies. In more recent years, as the number of children born after ART increases, larger studies with more robust methodologies have suggested that children born after assisted conception do have an increased risk of congenital malformation when compared with children conceived spontaneously. Data from recent meta-analyses suggest that the overall risk of major congenital malformation in children born after ART is around 30% higher than that in children conceived spontaneously9, 10. Prospective data support the same conclusion. In a study of a large cohort of children born after standard IVF and ICSI (n = 2840 and 2955, respectively) the rate of major congenital malformations was around 4%11, and another large prospective study comparing children born after ICSI with controls conceived spontaneously reported a relative risk of 1.24 (95% CI, 1.02–1.50)12. A large European multicenter prospective study similarly reported an increased risk of major congenital malformations after ART, with a slightly higher risk for both ICSI and IVF (odds ratio (OR), 2.77; 95% CI, 1.41–3.46 and OR, 1.80; 95% CI, 0.85–3.81, respectively)13. Finally, recently published data from population-based registries also support these conclusions. In a study of 16 280 children conceived after ART, identified from population-based Swedish registries, there was a 42% excess rate of congenital malformations compared with the unselected Swedish population over the same time period14, and another registry-based study using a retrospective cohort design reported a similar increased malformation rate (OR, 1.3; 95% CI, 1.0–1.67)15. Considering all congenital anomalies together as a single group could mask stronger associations with specific defects or types of defects. A number of studies have therefore focused on individual congenital malformation types. The most recent, and among the largest, of these was carried out in the USA, and suggested an association between ART singleton births and several types of malformation after adjusting for confounders such as maternal age, parity and prematurity16, including those affecting a range of organ systems, as discussed below. Whilst this study is among the most comprehensive to date, there remain potential methodological problems. The study was a case–control design, and it was subject to potential recall bias as information was collected retrospectively from mothers up to 2 years after delivery; thus incomplete medical records may have contributed further to inaccuracy. The association of cardiovascular defects and ART was first suggested in a registry-based study published more than 20 years ago, in which the number of cases of transposition of the great arteries in ART pregnancies was more than four-fold greater than expected. Other subsequent studies have similarly reported an approximate four-fold increase in the frequency of cardiac defects after ART17, 18. This increased risk was not limited to multiple pregnancies but the difference was less marked when only singletons were considered18. This study had the advantage of following children for 3 years, thus probably achieving a more complete ascertainment of cardiac defects compared with other studies, but individual numbers were small and the study included live-birth data only18. Another study reported an increased OR of 1.7 (95% CI, 1.5–2.0) for all cardiovascular defects in children born after ART compared with population expected values, with OR of 2.1 for all major cardiovascular defects (95% CI, 1.6–2.8) and 2.6 for cardiac septal defects (95% CI, 2.2–3.1)14. Other studies have supported this more modest increased risk15, 16, especially an association between ART singleton births and cardiac septal defects (OR, 2.1; 95% CI, 1.1–4.0)16. Increased risks of musculoskeletal malformations after ART have been reported by a number of studies15-19. Neurological defects, specifically neural tube defects such as spina bifida, have also been widely reported as increased after ART; studies including both the first6 and the largest14 published in this area suggest an approximately five-fold increased risk of spina bifida in children born after ART. There is some speculation that part of this increased risk could arise because women who have undergone ART may be less likely to terminate affected pregnancies than are women who conceive naturally, an issue potentially relevant to all types of congenital malformation following ART. However, even if this is the case, it does not adequately explain the extent of the increased risk observed because only approximately 50% of cases of spina bifida are terminated in the general population14. Two large studies reported a significantly increased incidence of esophageal and anorectal malformations associated with ART14, 16, when considering both isolated and multiple malformations. The risk of a cleft palate was reported as increased after ICSI in one study17, but these results remain controversial since the data had previously been presented by the original study group who concluded that the incidence was within the expected range20. However, one further study also reported an increase in the risk of orofacial clefts after ART (OR, 2.4; 95% CI, 1.0–3.1)14. There is convincing evidence to support the association of ART and urogenital malformations13, 14, 19, 21-23. Many of these studies have reported a specific increased risk of hypospadias in male infants born after ICSI13, 14, 21, 22. For example, the increased risk of congenital abnormality found in children conceived after ICSI compared with those conceived after IVF in a large, international, prospective follow-up study was largely due to the increase in frequency of hypospadias in male ICSI-conceived infants13. This association, now reported repeatedly, was not identified in some earlier studies, likely as a consequence of the small numbers included in these subgroup analyses11, 19, 23. Several syndromes, such as Beckwith–Wiedemann, Prader–Willi and Angelman syndromes, are known to be associated with defects in the expression of imprinted genes. A number of recent publications have reported unexpectedly high proportions of children born after assisted conception with such imprinting disorders24-29. Furthermore, those with Beckwith–Wiedemann syndrome and Angelman syndrome who were conceived by ART have a higher than expected rate of epigenetic mutations causing the phenotype24. It is not clear whether these associations are related to the underlying genetic predisposition of subfertile couples and/or to the interference of specific aspects of ART with epigenetic reprogramming during gametogenesis and early embryonic development. However, previous evidence of the influence of ART on the epigenetic make-up of gametes and embryos was provided by animal studies. Large offspring syndrome, observed after in-vitro culture of sheep embryos was found to be associated with reduced maternal methylation30. Further animal studies demonstrated that conditions during ART may also interfere with normal genetic programming, resulting in aberrant cardiovascular physiological status31. In addition to the influence of embryo culture on the epigenome, evidence is accumulating that ovarian hyperstimulation with gonadotropins, the transfer of embryos to the uterus and exposure to light may all be associated with epigenetic disturbance, resulting in possible imprinting disorders among other potential effects32, 33. The vast majority of evidence demonstrates that, in general, there is no difference in the rates and types of congenital malformation when comparing ICSI and standard IVF pregnancies7, 10, 11, 15, 19, 34. A recent meta-analysis including more than 5000 children born following ICSI and more than 13 000 born after standard IVF supports this conclusion both for all malformations and for certain types of defect, such as cardiovascular, musculoskeletal and neural tube defects34. The main exception to this finding, as mentioned above, is the marked association of urogenital defects, specifically hypospadias, with ICSI13, 14, 21, 22. This association is highly reproducible and has been reported to a similar degree in large methodologically sound studies11, 19, 23. There is little evidence reporting congenital malformations after other newer forms of ART. One of the very few published studies examining outcome following use of these newer techniques reports similar outcomes between two different methods of sperm extraction prior to ICSI35. However, as these newer techniques become more widely used it will become increasingly important to assess their safety for potential offspring, with regard to both congenital malformations and other outcomes. The underlying mechanisms of the increased rates of congenital malformations after ART is largely unknown. Some of the observed effects could be accounted for by the increased rates of multiple pregnancies, including monochorionic twinning, following ART, such multiple pregnancies being associated with increased risk of structural malformations. However, whilst this could explain some of the increased risk of congenital malformations, it cannot explain all of the risk and many of the studies discussed reported increased risks for singletons only, having excluded multiple births14, 16. Other more likely possibilities include the effects of medications used in ART cycles and specific factors associated with ART procedures themselves, including environmental factors such as culture conditions, and physical stresses placed upon gametes during ICSI. Finally, it is difficult to determine how much of the effect is related to factors associated with the parental characteristics of the underlying subfertility. The majority of studies investigating the association between ART and congenital malformations cannot differentiate the effects of ART itself from the effects of underlying subfertility36-39. Zhu et al.40 are among the few to attempt to address this fundamental issue. They found that the reported increased risk of congenital malformations associated with ART conception is also present, to a lesser extent, in previously subfertile couples who have then conceived spontaneously, suggesting a significant underlying parental component. This concept is also supported by data from another study investigating ART pregnancy outcomes, in which the increased risks of congenital malformations disappeared after adjustment for the parental characteristics of maternal age, parity, known years of unwanted childlessness and smoking14. The practical implications for a fetal medicine service providing antenatal screening and prenatal diagnosis for couples conceiving after ART are as follows. First, there should be increased awareness that the overall risk for congenital malformations is increased, particularly for specific subtypes (Table 1). These data should prompt specific examination of certain organ systems, such as for the prenatal detection of cardiac defects. Second, the potential implications of certain sonographic findings should be modified. For example, since a significant proportion of fetuses with apparently isolated omphalocele may have underlying Beckwith–Wiedemann syndrome41, the increased background frequency of such imprinting defects following ART will make the adjusted risk greater in the context of an abnormal ultrasound examination in this setting. Finally, it should be noted that the parental decision-making process for couples who have undergone ART may differ from that associated with pregnancies following spontaneous conception, due to a combination of the effects of the pre-existing difficulty in conceiving and the subsequent wish to avoid prenatal diagnosis procedures which may cause miscarriage; women with ART pregnancies are significantly more likely to attend for first-trimester screening but are much less likely to undergo invasive diagnostic procedures compared with spontaneously conceived pregnancies, especially those aged over 35 years, both with singleton and multiple pregnancies42, 43. If screening for chromosomal abnormalities is undertaken, nuchal translucency thickness measurement-based protocols are suggested: dating using crown–rump length appears equivalent in spontaneous and ART pregnancies for accurately assessing gestation, while median NT thickness is apparently very slightly, but significantly, increased in the ART group44, 45. Finally, it should be remembered that whilst the risk of congenital malformations is slightly, but significantly, increased following ART, the risks of other pregnancy complications, in particular multiple pregnancy and its consequences and preterm delivery, remain far more common obstetric complications in this patient group46. In summary, the available evidence suggests an association between ART and a slightly increased frequency of structural congenital malformations, specifically congenital heart defects, neural tube defects, facial cleft, gastrointestinal malformations, genitourinary malformations and imprinting disorders. The underlying mechanisms are unclear and the risks associated with emerging ART techniques are yet to be determined.