Intracytoplasmic sperm injection (ICSI) was introduced in the early 1990s to treat cases of severe male factor infertility.¹ From the very beginning there has been concern about increased health risks in children conceived by ICSI. Large follow up studies have shown that the risk of major malformations, low birth weight, and pregnancy complications is increased after in vitro fertilisation, but independent of the use of ICSI.^2–5

Recent case reports on three children who were conceived by ICSI and have Angelman syndrome (AS) as the result of an imprinting defect on chromosome 15 have suggested that artificial reproduction may increase the risk of imprinting defects.^6,7 Similar observations have been made for children with Beckwith-Wiedeman syndrome.^8,9,10

AS is a rare neurogenetic syndrome characterised by microcephalus, jerky movements, absence of speech, abnormal EEG pattern, severe mental retardation, and frequent laughing (incidence, 1 in 15 000 newborns). It is caused by the loss of function of the maternal UBE3A allele on chromosome 15. UBE3A is subject to genomic imprinting and, in brain, is expressed on the maternal chromosome only. Approximately 70% of patients have a deletion on the maternal chromosome, 10% have a mutation in the UBE3A gene, 4–7% have two paternal copies of chromosome 15 (uniparental disomy), and 3–4% have an imprinting defect (ID) that silences the maternal UBE3A allele. In the remaining cases, the patients have some other, hitherto unknown defect.¹¹ The relative frequencies of the different genetic defects do not vary between different ethnic groups.

To investigate a possible correlation between infertility treatment and imprinting defects, we performed a cohort study in Germany using data from the German Angelman Syndrome Support Group.

METHODS

All members of the German Angelman Syndrome Support Group were contacted by a letter and asked to provide, on a voluntary basis, anamnestic data and also information on the method of conception and the time to pregnancy (TTP). The study was approved by the ethics review board of the Ärztekammer Hamburg. Patients gave written informed consent. Parents who stated that TTP exceeded 2 years and/or who had undergone infertility treatment were asked to provide a blood sample (10 ml EDTA) or buccal smear from themselves and the child.

The methylation status of the SNURF-SNRPN gene was determined by bisulphite treatment of genomic DNA and methylation specific PCR.¹² For segregation analysis of microsatellite loci along chromosome 15, fluorescence tagged PCR products were analysed on an ABI Prism 3100 Genetic Analyzer, using GeneScan and Genotyper software (ABI, Foster City, CA, USA). In accordance with established diagnostic criteria, patients carrying an unmethylated maternal SNURF-SNRPN allele were classified as having an ID. In all patients with an ID, a familial imprinting centre (IC) deletion was excluded by quantitative real time PCR analysis of the critical IC elements (Buiting and Horsthemke, unpublished).

RESULTS

Of 270 members, 82 (30%) replied. One child was adopted; two other parents refused to give detailed information. Thus, 79 valid questionnaires were returned. Sixteen children (20%) were born to subfertile couples (defined as having had a TTP >2 years and/or infertility treatment). This is a higher rate than that expected in the general population (∼10–15%) and indicates a reporting bias. However, this bias does not affect our results, because we have compared the percentage of IDs in this group with the percentage of IDs in all patients with Angelman syndrome.

Four of the sixteen children (25%) were found to have a sporadic ID (table 1). Assuming that an ID accounts for 4% of AS patients (see above), the relative risk (RR) was significantly increased in this group of patients (table 2). The RR was increased by the same factor in the untreated subgroup of couples with TTP >2 years (n = 8) and in the subgroup of couples who underwent ICSI or hormone treatment (n = 8), although the increase did not reach statistical significance. We noted that there was one ID child in each of the ICSI (n = 3) and the hormone treatment only (n = 5) groups. RR was highest in couples with TTP >2 years who had also undergone infertility treatment (n = 4). This increase was statistically significant.

DISCUSSION

In the present study we found that subfertile couples have an increased risk of conceiving a child with an imprinting defect causing AS. The increased RR in untreated couples with TTP >2 years suggests that imprinting defects and subfertility may have a common cause. This observation is supported by data showing that an increased TTP alone is a significant risk factor for the development of pre-eclampsia or intrauterine growth retardation.^13–15 This risk of pregnancy complications associated with an increased TTP is similiar to that observed in pregnancies conceived after ICSI⁴ or assisted reproduction in general.¹⁶

Interestingly, early embryonic development and intrauterine growth are critically dependent on the proper function of imprinted genes.¹⁷ Based on these findings, we propose that there is some genetic predisposition, possibly of a heterogeneous nature, to epigenetic instability of gametes or early embryonic cells. Depending on the number and type of affected loci in these cells, epigenetic instability can lead to early developmental failure, which may manifest as subfertility, intrauterine growth retardation, or congenital anomalies.

As the RR was twice as high in couples with TTP >2 years who had undergone hormonal stimulation or ICSI, infertility treatment may further increase the risk of conceiving a child with ID. One possible explanation is that these couples suffered from a more severe form of subfertility that necessitated infertility treatment. Alternatively, hormonal stimulation, which is also used for ICSI, may lead to the maturation of “poor quality” oocytes that would not have been ovulated without treatment, or that a too rapid maturation provoked by the hormonal stimulation procedure disturbs the process of DNA methylation in the oocyte. A lower oocyte quality following ovarian stimulation procedures has already been suggested from studies in the mouse.¹⁸

As our cohort of patients does not contain children conceived by conventional in vitro fertilisation without ICSI or intrauterine insemination, we do not know whether gamete and embryo culture or embryo manipulation increases the risk for an ID, as has been shown in animal studies.¹⁹

In summary, we have demonstrated for the first time that the prevalence of imprinting defects in patients with AS born to subfertile couples is significantly increased. Our data suggest that imprinting defects and subfertility may have a common cause, and that superovulation, rather than ICSI, may further increase the risk of conceiving a child with an imprinting defect. However, the absolute risk remains small.