SEARCH

SEARCH BY CITATION

Keywords:

  • Diploidisation;
  • genetics;
  • recurrent molar pregnancy;
  • therapy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

Hydatidiform moles of two women, each with three molar pregnancies, were examined in order to study their origin. Multiple recurrences have previously been associated with women who have biparental complete hydatidiform moles (CHM). However, all the moles examined in this study were androgenetic CHM (AnCHM), indicating that recurrent (>2) moles, particularly in the absence of a positive family history, may be androgenetic rather than biparental. These data suggest that some women have a specific liability for having AnCHM. Making the distinction between a biparental or an androgenetic origin of recurrent moles is of relevance for counselling and when considering therapeutic options. Therefore, we propose that all recurrent moles should be investigated using molecular techniques.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

Hydatidiform moles (HM) form a heterogeneous group of disorders, with an incidence in the Western World ranging from 1 in 500 to 1 in 1500 pregnancies. The unifying feature of all moles is the absolute or functional overrepresentation of the paternal genome in the conceptus. On gross morphological, histological and genetic grounds, a distinction between partial hydatidiform moles (PHM) and complete hydatidiform moles (CHM) can be made. This is important since CHM have a far greater potential for developing into a choriocarcinoma compared with PHM. PHM are diandric triploids (one maternally derived and two paternally derived chromosome sets). The great majority of these triploids are caused by dispermic fertilisations.1 CHM are diploids. They can be subdivided into homozygous or heterozygous androgenetic CHM (AnCHM) and biparental CHM (BiCHM). Homozygous AnCHM are the most frequent type of CHM (∼80%); two identical paternal contributions are present, derived by duplication of a single paternal chromosome complement. They are female; 46,YY moles have never been observed. In heterozygous AnCHM, which can be either male or female, there are two different paternal contributions (Table 1).

Table 1.  Different types of HM
Molar pregnancy
PHM 3n (two pcc + one mcc), ∼60%CHM 2n, ∼40%
  1. pcc, paternal chromosome complement; mcc, maternal chromosome complement.

 AnCHM; two pcc, no mccBiCHM; one pcc + one mcc (rare)
 Homozygous, ∼80%Heterozygous, ∼20% 
69,XXX69,XXY69,XYY (rare)46,XX46,XX or 46,XY46,XX or 46,XY

BiCHM are an infrequent condition. In BiCHM, there is a maternal as well as a single paternal contribution to the molar genome. However, women with BiCHM are unable to establish a normal female imprinting pattern in their oocytes, resulting in a functional over-representation of the paternal genome at certain imprinted loci after fertilisation. The imprinting defect is caused by mutations in an as yet unidentified gene at 19q13.4,2 and inherited in an autosomal recessive fashion. BiCHM are generally diagnosed in women with multiple moles, usually in the absence of normal pregnancies. Consistent with the autosomal recessive inheritance, family history is usually positive for HM and consanguinity in the parents of these women has frequently been reported.

In order to make a definite distinction between PHM and CHM, ploidy has to be determined, for example by flow cytometry. Alternatively, absence of immunostaining with an antibody for the P57KIP2 protein, which is expressed only from the maternal allele, can be regarded as evidence for CHM. However, androgenetic and biparental subtypes of CHM can only be distinguished using polymorphic DNA markers.

In this study, moles of two couples, each with three CHM but a negative family history, were investigated for the possibility of being BiCHM. The prior probability of BiCHM in this situation is unknown. However, since women with BiCHM have a close to 100% chance of adverse pregnancy outcome, this is clearly of clinical relevance.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

The couples were ascertained at a university clinic for genetic counselling (case 1) and a fertility clinic of a large regional hospital (case 2).

Case 1 concerns a healthy 38-year-old woman of Dutch Caucasian descent, who had a total of five pregnancies. At age 32, she had a healthy son. At age 33, she had a molar pregnancy, followed by treatment with methotrexate because of persisting elevation of human chorionic gonadotrophin (hCG) levels. All the subsequent pregnancies were conceived after hCG level was normalised for at least 6 months after a molar event. At age 35, the woman had a miscarriage at around 10 weeks of gestation, and at age 36, a mucinous cystadenoma was removed laparoscopically. At ages 37 and 38, two more molar pregnancies followed, which spontaneously resolved after evacuation. All the moles were initially diagnosed as PHM, but this diagnosis was revised to CHM after ploidy had been determined. Both partners had normal karyotypes. Family history was negative for HM, recurrent miscarriage, infertility or known genetic disease. The proband has one brother, who has fathered no children of his own. Genealogical investigations demonstrated unexpected distant consanguinity in the parents of the proband, leading to a 0.2% chance of homozygosity by descent at any given locus (chance that the index patient has two identical copies of any allele, derived from the common forefather).

Case 2 is a healthy 29-year-old woman of Turkish descent, with her nonconsanguineous 30-year-old husband. No consanguinity in the parents of the woman was reported. The family history is negative for HM. The proband had a total of three pregnancies at ages 22, 24 and 26, all of which were CHM on pathological examination. HCG levels dropped soon after evacuation in all pregnancies. In 2003, the couple elected to try to become pregnant through intracytoplasmatic sperm injection (ICSI), although they were aware that this would not necessarily prevent a recurrence. Out of 34 oocytes retrieved, 26 could be injected. Two were scored with 3 pronuclei (PN), 11 with 2 PN, 5 with 1 PN and 8 were either damaged or showed no PN. Five zygotes with one PN is slightly above average. Because of imminent, severe ovarian hyperstimulation syndrome, two embryos were initially cryopreserved. One month later, they were transferred, but this did not result in a pregnancy. Tissue from the last molar pregnancy was available for investigation.

For both cases, fluorescent microsatellite genotyping was carried out as previously described, using DNA prepared from parental blood and pathological sections of molar tissue.3

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

Results of the genotyping are shown in Table 2. In all four moles investigated, only a single marker, derived from the father, was present for each informative markers, indicating that all four moles were AnCHM. There was no evidence of a maternal contribution to the genome for any informative markers. In addition, P57KIP2 immunostaining was negative in all cases, consistent with a diagnosis of CHM. Testing of an additional five informative markers in case 1 showed all three moles to be homozygous AnCHM (data not shown).

Table 2.  Microsatellite genotyping
Molecular analysis
Case 1Case 2
MarkersFatherMotherCHM1CHM2CHM3FatherMotherCHM3
  • *

    Not informative.

Mfd50177–189191189189177177190–192177
D10S179115–123113–131123115123115–129113–131115
D11S1999112124112112112108–112104–116112
D1S1656131–153144–153131153*131148144–156148
APOC2146–148122–146146*146*146*144–146130–151144
D20S481224–238230–238224224238230–242222–234242

Although only one out of three moles was investigated in case 2, it is extremely unlikely that the previous moles were BiCHM. Other types of moles have not been reported in individuals with BiCHM.

Despite the fact that BiCHM could be ruled out, the couple of case 1 decided not to become pregnant again. The couple of case 2 had not yet reached a final decision on future pregnancies.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

Different pathogenetic mechanisms lead to molar pregnancies. Fertilisation errors have been shown to underlie PHM, and an autosomal recessive single-gene defect has been implicated in BiCHM. In homozygous AnCHM, normal fertilisation of an anucleate oocyte is postulated, followed by duplication of the single paternal chromosome set. Heterozygous AnCHM result after a dispermic fertilisation of an anucleate oocyte.

Pregnancies at both extremes of the reproductive age have an increased risk of being molar. However, the most important determinant of the HM risk seems to be a previous molar pregnancy. Most studies report recurrence risks after one HM, ranging from 0.5 to 2%.4 This figure may be acceptably low to most prospective parents. However, given the causal heterogeneity in HM, it may well be that the observed small increase in risk is largely due to high risks in a small subset of women. Indeed, 3 of 27 women with CHM, who had a recurrence in the series of Sebire et al.,4 went on to have a third molar pregnancy. Recurrence risk after two moles has been estimated to be 10–28%,4,5 possibly due to inclusion of BiCHM cases with a very high risk of recurrence. It has been demonstrated that women with three moles may have BiCHM even in the absence of a positive family history3 and may be included in series of recurrent HM. Although the gene defect predisposing to BiCHM is likely to be very rare, the magnitude of the contribution of BiCHM to the observed risk increase remains to be determined. Our cases indicate that the large increase in risk after a second molar event is not solely due to the admixture of these rare BiCHM cases.

This study has shown that women with more than two CHM may have homozygous AnCHM rather than BiCHM, even where the parents of the woman are related. It is likely that women with multiple AnCHM have a specific liability for developing the disorder. The finding of distant consanguinity in the parents of the proband in case 1 may be coincidence. Alternatively, it may point in the direction of a recessive gene effect (different from that, which plays a role in BiCHM) contributing to the liability for recurrent AnCHM. A consanguineous family with recurrent HM in several family members has been reported, which was unlinked to the known BiCHM locus at 19q13.4,6 suggesting a second autosomal recessive locus for familial HM. Unfortunately, it is not clear whether the moles in this family were BiCHM or AnCHM.

A possible explanation for recurring AnCHM, as described in this study, would be an increased tendency for ovulating anucleate oocytes, caused by a defect in the female meiotic apparatus. However, since proof is lacking that anucleate oocytes are ovulated regularly and capable of being fertilised in vivo, this concept has recently been challenged as a cause for AnCHM. Postzygotic diploidisation of triploid (PDT) conceptuses offers an alternative explanation.7 This concept is based on the experimental observation that dispermic triploid conceptions are unstable at first mitosis and may frequently give rise to daughter cells that could potentially develop to become AnCHM. Not all peculiarities of molar genetics can be directly explained by the PDT concept. Especially the observed 4 to 1 difference in frequency between homozygous and heterozygous AnCHM remains to be clarified. However, PDT is an attractive hypothesis for different reasons. In the first place, triploid conceptions are very frequent, whereas ovulated anucleate oocytes are a theoretical concept. Second, it would follow that all non-BiCHM can be caused by one single mechanism—a dispermic fertilisation. This would explain the puzzling fact, noted by different authors, that a CHM increases the risk of both CHM and PHM in subsequent pregnancies and vice versa.

Making the distinction between the anucleate oocyte and PDT concepts is of relevance when studying multiple AnCHM cases. If moles indeed arise through PDT, research should focus on fertilisation errors, such as defects in the function of the zona pellucida, rather than on female meiosis. From a therapeutic point of view, ICSI alone would be sufficient to prevent recurrences in women with recurrent AnCHM, if they indeed arise through PDT. Since ICSI was used in our second case instead of regular in vitro fertilisation, this hypothesis could unfortunately not be evaluated.

In summary, we demonstrated for the first time that women with multiple (>2) CHM do not necessarily have BiCHM. We propose that women with repeated AnCHM do have an increased recurrence risk, although it is likely to be much lower than in women with BiCHM. Making the distinction between BiCHM and AnCHM is relevant to genetic counselling and treatment options. Women having BiCHM have close to 100% risk of adverse pregnancy outcome, while (based on empirical data)4,5 women with recurrent AnCHM will probably have a reasonable chance of having healthy children. Regardless of the exact pathogenesis, recurrences in women with AnCHM could be prevented if ICSI were offered in combination with preimplantation genetic diagnosis (PGD), using informative DNA markers to assure the presence of a maternal contribution in the conceptus. In BiCHM cases, only egg donation can be considered as an option. We therefore suggest that in future cases of women with two or more HM, the moles should be investigated with molecular techniques. Analysis of the fertilisation products of those women who, after two or more AnCHM, choose ICSI and PGD treatment may be helpful in clarifying the mechanism by which AnCHM arise.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References