Few cases of complete hydatidiform mole with a coexisting fetus have been reported over the last two decades. This broad term can be classified into three major types: (1) twin gestation in which one twin is a diploid fetus with a normal placenta (46 chromosomes, 23 maternal and 23 paternal) and the other twin is a complete hydatidiform mole (46 chromosomes of paternal origin) with no fetus (this is applied to our case report). (2) Singleton gestation consisting of a triploid fetus with partial hydatidiform mole placenta (69 chromosomes, 23 maternal and 46 paternal). (3) Twin gestation in which one twin is a diploid fetus with normal placenta (46 chromosomes, 23 maternal and 23 paternal) and the other twin is a triploid fetus with partial hydatidiform mole placenta (69 chromosomes, 23 maternal and 46 paternal) . Categorization of the case is essential for proper management. Unlike partial hydatidiform mole that is commonly associated with multiple fetal anomalies and is managed by immediate termination of pregnancy , reported cases of twin pregnancy with complete hydatidiform mole (including our case) are not associated with fetal anomalies in the coexisting fetus; in some cases the mother has even given birth to fetuses that survived .
However, twin pregnancy with complete hydatidiform mole has a higher risk of maternal complications than partial hydatidiform moles; the same risk also applies to twin pregnancy with complete hydatidiform mole. These complications include early onset pre-eclampsia, thyrotoxicosis and persistent trophoblastic disease (PTD). Because of case rarity, the incidence of these complications cannot be exactly assessed in comparison with complete molar pregnancy; however, the incidence of PTD in reported case series varies from 19% to 50% [2, 7]. The reason for this high incidence has not yet been clarified.
Accordingly, the management of these cases remains problematic: the fair possibility of fetal survival is weighed against the expected risk of maternal complications, and for this reason, many reported cases were managed by immediate termination. However, some authors support the option of conservation under strict hospital-based observation and follow-up. In this case, we support the latter policy for the following reasons. First, conservative management of some reported cases was successful, with fetal survival . Second, most expected complications can be diagnosed by strict follow-up and clinical observation. Third, although the risk of PTD cannot be excluded during conservative management, this risk does not seem to increase with advanced gestational age . According to these findings, we suggest that conservative management should always be a choice in such patients.
Amniocentesis is expected to be beneficial in decision making; a triploid fetus would be expected to be severely malformed and thus, termination of pregnancy would be recommended. A diploid fetus (46XX or 46XY of maternal and paternal origins) indicates a viable fetus with a normal placenta and therefore, pregnancy can be allowed to continue . With regard to our patient, the option of amniocentesis was refused and we were obligated to manage our patient conservatively according to her will. Unfortunately, our patient experienced serious complications during conservation and hysterotomy was indicated; the fetus was alive but did not survive. However, as our patient was managed conservatively as long as possible and she did not experience any adverse post-operative outcomes including gestational trophoblastic disease, these findings may support the relative safety of conservation of these cases.
The falsely low result for serum β-hCG represents another interesting point in our patient's case; this is described as the 'high-dose hook effect'. This effect is not specific for β-hCG. It occurs when there is an inordinate amount of substance being measured by an immunoassay; this causes the formation of incomplete antibody-antigen complexes. Accordingly, below a certain threshold concentration, the assay will reflect the concentration of the substance correctly. Above this concentration, the assay will record falsely lower results as the concentration rises higher. This can be corrected by dilution of a sample and this was exactly what we did in this case .
It is infeasible to correlate between the occurrence of complete hydatidiform mole with a coexisting fetus and ovulation induction, particularly with clomiphene citrate, because the number of reported cases is limited. However, the association between complete hydatidiform mole and ovulation induction has not been confirmed; the incidence of complete hydatidiform mole following ovulation induction with clomiphene citrate was 1:659 [9, 10]. However, Piura et al. reported 30 cases of twin pregnancy with complete hydatidiform mole and a co-existent fetus. Of these, nine cases (30%) were preceded by ovulation induction with either hMG/hCG (eight cases) or clomiphene citrate (one case). Five out of eight patients who received human menopausal gonadotropin (hMG)/human chorionic gonadotropin (hCG) were exposed to assisted reproductive technologies (in vitro fertilization/embryo transfer (IVF/ET) in three women, intra-cytoplasmic sperm injection/embryo transfer (ICSI-ET) in two women) . This correlation may need further study and explanation.