Objective
The purpose of this study was to analyze the correlation between the genetic constitution and the phenotype in triploid pregnancies.
Study Design
One hundred fifty-eight triploid pregnancies were identified in hospitals in Western Denmark from April 1986 to April 2010. Clinical data and karyotypes were collected retrospectively, and archived samples were retrieved. The parental origin of the genome, either double paternal contribution (PPM) or double maternal contribution (MMP) was determined by an analysis of methylation levels at imprinted sites.
Results
There were significantly more PPM than MMP cases ( P < .01). In MMP cases, the possible karyotypes had similar frequencies, whereas, in PPM cases, 43% had the karyotype 69,XXX, 51% had the karyotype 69,XXY, and 6% had the karyotype 69,XYY. Molar phenotype was seen only in PPM cases. However, PPM cases with a nonmolar phenotype were also seen. For both parental genotypes, various fetal phenotypes were seen at autopsy. Levels of human chorionic gonadotropin in maternal serum were low in MMP cases and varying in PPM cases, some being as low as in the MMP cases.
Conclusion
In a triploid pregnancy, suspicion of hydatidiform mole at ultrasound scanning, by macroscopic inspection of the evacuated tissue, at histology, or because of a high human chorionic gonadotropin in maternal serum level each predict the parental type PPM with a very high specificity. In contrast, the sensitivity of these observations was <100%.
Triploidy, which is one of the most common chromosome aberrations, is estimated to occur in 1-2% of all human conceptions.
Triploid pregnancies can be classified according to the genetic constitution. The genome in a triploid pregnancy is either digynic, which consists of 2 maternal and 1 paternal set of chromosomes (MMP), or diandric, which consists of 2 paternal and 1 maternal set of chromosomes (PPM). Extremely varying frequencies of the 2 parental types have been reported (the frequency of PPM cases ranges from 20–73%).
Triploid pregnancies can also be classified according to their phenotype. By the use of ultrasound scanning (US), it has been found that the placenta can appear normal, small, or enlarged and with or without vesicular changes. If a fetus is present, it can exhibit malformations but can also appear normal. Growth restriction appears to be a common, but not a consistent, phenomenon. In studies that have been focused on levels of serum maternal human chorionic gonadotropin (MS-hCG), low, normal, and high levels have been reported. In some studies that involved postmortem examinations of triploid pregnancies, 2 distinctive phenotypes have been suggested: type 1, relatively well-grown fetuses with proportionately sized body parts and large placentas with the morphologic condition of a partial hydatidiform mole; type 2, fetuses with severe growth retardation, an uneven development of various body parts that typically results in a relative macrocephaly and small, noncystic placentas.
However, a systematic description and classification of a large cohort of triploid pregnancies has not been published. We correlate the parental origin of the genome with karyotype and phenotype of the triploid pregnancy and with the concentration of MS-hCG in a series of 158 triploid pregnancies.
Materials and Methods
The study was approved by the Regional Committees on Health Research Ethics in Southern Denmark and the Central Region Denmark and by the Danish Data Protection Agency.
We aimed at including all triploid pregnancies in Western Denmark (Jutland) that were diagnosed from April 1986 to April 2010. In this period, triploid pregnancies were registered either in The Danish Mole Project (DMP), The Danish Cytogenetic Central Registry (DCCR), or both.
As part of the DMP, gynecologists in Jutland forward fresh placental tissue from pregnancies that are suspected to be hydatidiform moles. Triploidy is identified by karyotyping uncultured and/or cultured cells and/or by measurement of the nuclear DNA contents by flow cytometry of unfixed nuclei with the use of 2 external controls.
The DCCR has held data on all karyotypes that have been made for clinical purposes in Denmark since 1960. All women in Jutland who have been registered in the DCCR with a triploid pregnancy in the study period were identified. With the use of the women’s Civil Registration System number, their place of residence was found, and the laboratory that performed the genetic analysis was identified.
Frozen cell cultures and/or DNA were retrieved from the departments of clinical genetics at Aarhus University Hospital and Vejle Hospital and from the DMP.
Determination of the parental origin of the genome
DNA was purified with QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). The triploid samples were analyzed with methylation-specific multiplex ligation–dependent probe amplification. In brief, the level of methylation at 2 imprinted loci in the 11p15 region was measured with the SALSA MLPA ME030 BWS/RSS probemix (Kit ID MRCH 30808 and MRCH 36554; MRC-Holland, The Netherlands). Because imprinted loci are methylated in a parent-of-origin–specific manner, we previously have shown that the methylation level is an indicator of the proportion of maternally vs paternally inherited genomic material.
Clinical data
Information regarding the first day of last menstrual period, the maternal serum concentration of the beta subunit of human chorionic gonadotropin (MS-beta hCG; if tested in the first trimester), the maternal serum concentration of total human chorionic gonadotropin (MS-total hCG; if tested later in pregnancy), and the results from histopathologic examination/autopsy of the pregnancy were retrieved from medical records. Data on the morphologic findings that were observed by US were retrieved from Astraia (a database used by the Danish Departments of Gynecology and Obstetrics).
Gestational age was calculated from last menstrual period (GA LMP ) by Naegele’s rule.
Classifications and definitions
The placenta was categorized as molar if the sonographer described the placenta with cystic spaces or molar appearance, if the placenta fulfilled the criteria for inclusion in the DMP, or if a histopathologic diagnosis of hydatidiform mole was made.
Fetal abnormalities detected by US were defined as any fetal malformation and/or oligohydramnios and/or intrauterine growth restriction (IUGR). With regards to abnormalities of the fetus that were found after termination, the morphologic condition that was described by the obstetrician after evacuation, and the autopsy findings were retrieved.
For the MS-beta hCG and MS-total hCG, the multiples of the median (MoMs) were calculated by expressing the absolute concentrations relative to the median for the relevant GA LMP . For MS-beta hCG, medians were calculated by the algorithm described by Spencer. For MS-total hCG, medians were calculated by the exponential regression. Low levels of MS-hCG were defined as <0.5 MoM; high levels were defined as >3.0 MoM.
Categoric variables were compared with the use of the chi-square test, Fisher’s exact test (Microsoft Excel 2010; Microsoft Corporation, Redmond, WA), or unpaired t -test (Graph Pad Prism, version 5.01; GraphPad Software, Inc, La Jolla, CA). GA LMP s were compared with the use of a 2-tailed Mann-Whitney test (Graph Pad Prism software version 5.01; GraphPad Software, Inc). Significance level was set at a probability value of < .05.
Results
Data on each triploid pregnancy can be seen in the Appendix ( Supplementary Table ).
A total of 183 triploid pregnancies had been registered: 140 in the DMP, 41 in the DCCR, and 2 who were registered in both the DMP and the DCCR. We succeeded in retrieving samples from a total of 158 cases: 121 cases from the DMP, 35 from the DCCR, and the 2 cases who were registered both in the DMP and the DCCR. Clinical data were available in all 158 cases.
GA LMP at termination was available in 115 cases (21 MMP and 94 PPM) and ranged from 10 + 5 to 30 + 3 weeks’ gestation (median, 16 + 6 weeks). The difference between the median GA LMP at termination for PPM and MMP cases (17 + 0 and 15 + 5 weeks gestation, respectively) was not significant ( P = .7).
Parental origin of the genome
One hundred thirty-four of the 158 included cases (85%) had the parental type PPM (all of the 121 cases from the DMP, 11 cases from the DCCR, and the 2 cases registered in both registers), and 24 cases (15%) had the parental type MMP (all 24 were found in the DCCR). The difference between the 2 proportions was statistically significant ( P < .01).
The parental type of 105 cases (103 PPM and 2 MMP) has been presented previously.
Karyotype
A karyotype was registered in 137 cases ( Table 1 ): among the 115 cases with the parental type PPM, 49 cases (43%) had the karyotype, 69,XXX; 59 cases (51%) had the karyotype 69,XXY, and 7 cases (6%) had the karyotype 69,XYY. All cases with the karyotype 69,XYY were classified as hydatidiform moles by the histopathologist.
| Karyotype | Total, n | Double paternal contribution/double maternal contribution, n |
|---|---|---|
| 69,XXX | 61 | 49/12 |
| 69,XXY | 69 | 59/10 |
| 69,XYY | 7 | 7/0 |
| Total | 137 | 115/22 |
Among the 22 cases with the parental type MMP, 12 cases (55%) had the karyotype 69,XXX, and 10 cases (45%) had the karyotype 69,XXY.
Placenta morphologic findings
Suspicion of a molar pregnancy based on US was raised in 75 of 103 PPM cases, but in none of 22 MMP cases. All cases included in the DMP had the parental type PPM. The reports from histopathologic examination of the placentas were available in 105 cases (98 PPM cases and 7 MMP cases). At the histopathologic examination of the cases with the parental type PPM, 74 of the placentas (76%) were described as molar; 12 placentas (12%) were diagnosed with hydropic degeneration, and 12 placentas (12%) were described as normal placentas. Histopathologic observations were available for 7 of the placentas from pregnancies with the parental type MMP. None of these cases were classified as hydatidiform moles; in 6 cases, the placentas were described as normal, and in 1 case, hydropic degeneration was diagnosed.
Fetal morphologic findings
By US before termination of the pregnancy, a living fetus was registered in 10% of the PPM cases (11/103) and in 95% of the MMP cases (18/19). Among 29 cases with a living fetus at US, fetal abnormalities were registered in 14 cases. Of these, 3 cases had the parental type PPM, and 11 cases had the type MMP; the difference was not significant ( P = .16). Autopsy data on the fetuses are shown in Table 2 . A wide variety of phenotypes that ranged from near normal to highly abnormal was seen.
| Identification no. | Parental type | GA LMP at termination in wk + d | Histopathologic morphologic condition of placenta | Histopathologic observations in fetus |
|---|---|---|---|---|
| 70 | PPM | 12 + 2 | Normal | Intrauterine growth restriction; no external malformations |
| 40 | PPM | 13 + 6 | Pregnancy product | CRL, 7 cm |
| 71 | PPM | 18 + 3 | Normal | No external malformations, renal cysts bilaterally, dysplasia of the right kidney, and possible hypoplasia of the thymus |
| 77 | PPM | 19 | Partial mole | CRL, 12 cm; feet, 20 mm, which indicated gestational age at 16 weeks; sacral skin defect 8 mm; low set ears; bifid uvula; hyperplasia of Leydig cells, proliferation of adrenal cord |
| 120 | PPM | 20 + 1 | Partial mole | Total length, 19 cm; feet, 20 mm; cleft lip and palate |
| 81 | PPM | 22 | Partial mole | CRL, 3.5 cm; weight, 3.5 g; long body small head; adhesion of left upper arm to facial structures; adhesion between upper and lower leg; agenesis of hands and feet; anus abnormally placed in columnar; organs degenerated |
| 122 | PPM | 24 + 5 | Partial mole | Intrauterine growth restriction and deformities of feet (not specified) |
| 123 | PPM | 26 + 2 | Partial mole | Syndactyly of 3th and 4th finger and toe bilaterally and micro-penis; pulmonary hypoplasia; multicystic renal dysplasia; atresia of proximal ureters |
| 23 | MMP | 30 + 3 | Normal | Square-shaped cranium; low-set ears; narrow nose; syndactyly of the 3th and 4th finger on the left hand and of the 4th and 5th toe on both feet; migration defect in cerebrum; atrophic thymus; and perforated small intestine |
| 18 | MMP | 15 + 2 | Normal | Asymmetric growth retardation; low-set ears; increased distance between eyes; micrognathia, syndactyly of 3th and 4th fingers on the left hand and 2th and 3th toe on the right foot |
| 19 | MMP | 15 + 3 | Normal | No external malformations |
| 10 | MMP | 16 + 1 | NA | Asymmetric growth retardation a |
| 12 | MMP | 16 + 6 | NA | Small of date possibly anencephalic a |
| 14 | MMP | 18 + 4 | NA | Large head; low-set ears; cleft palate; myelomeningocele over spinal column; omphalocele a |
| 21 | MMP | 20 + 5 | Normal | Hydrocephaly; micrognathia; syndactyly of 3th and 4th finger on the left side and 3th and 4th toe on the right side; possible hypoplasia of the lungs |
MS-hCG
In 7 of 24 cases with the parental type MMP, data on MS-beta hCGs were retrieved (median, 0.17 MoM; range, 0.03–0.33 MoM). Among the 134 cases with the parental type PPM, MS-se-total hCGs were retrieved in 75 cases (median, 13.5 MoM; range, 0.25–1556.9 MoM; Figure ).
Results
Data on each triploid pregnancy can be seen in the Appendix ( Supplementary Table ).
A total of 183 triploid pregnancies had been registered: 140 in the DMP, 41 in the DCCR, and 2 who were registered in both the DMP and the DCCR. We succeeded in retrieving samples from a total of 158 cases: 121 cases from the DMP, 35 from the DCCR, and the 2 cases who were registered both in the DMP and the DCCR. Clinical data were available in all 158 cases.
GA LMP at termination was available in 115 cases (21 MMP and 94 PPM) and ranged from 10 + 5 to 30 + 3 weeks’ gestation (median, 16 + 6 weeks). The difference between the median GA LMP at termination for PPM and MMP cases (17 + 0 and 15 + 5 weeks gestation, respectively) was not significant ( P = .7).
Parental origin of the genome
One hundred thirty-four of the 158 included cases (85%) had the parental type PPM (all of the 121 cases from the DMP, 11 cases from the DCCR, and the 2 cases registered in both registers), and 24 cases (15%) had the parental type MMP (all 24 were found in the DCCR). The difference between the 2 proportions was statistically significant ( P < .01).
The parental type of 105 cases (103 PPM and 2 MMP) has been presented previously.
Karyotype
A karyotype was registered in 137 cases ( Table 1 ): among the 115 cases with the parental type PPM, 49 cases (43%) had the karyotype, 69,XXX; 59 cases (51%) had the karyotype 69,XXY, and 7 cases (6%) had the karyotype 69,XYY. All cases with the karyotype 69,XYY were classified as hydatidiform moles by the histopathologist.
| Karyotype | Total, n | Double paternal contribution/double maternal contribution, n |
|---|---|---|
| 69,XXX | 61 | 49/12 |
| 69,XXY | 69 | 59/10 |
| 69,XYY | 7 | 7/0 |
| Total | 137 | 115/22 |
Among the 22 cases with the parental type MMP, 12 cases (55%) had the karyotype 69,XXX, and 10 cases (45%) had the karyotype 69,XXY.
Placenta morphologic findings
Suspicion of a molar pregnancy based on US was raised in 75 of 103 PPM cases, but in none of 22 MMP cases. All cases included in the DMP had the parental type PPM. The reports from histopathologic examination of the placentas were available in 105 cases (98 PPM cases and 7 MMP cases). At the histopathologic examination of the cases with the parental type PPM, 74 of the placentas (76%) were described as molar; 12 placentas (12%) were diagnosed with hydropic degeneration, and 12 placentas (12%) were described as normal placentas. Histopathologic observations were available for 7 of the placentas from pregnancies with the parental type MMP. None of these cases were classified as hydatidiform moles; in 6 cases, the placentas were described as normal, and in 1 case, hydropic degeneration was diagnosed.
Fetal morphologic findings
By US before termination of the pregnancy, a living fetus was registered in 10% of the PPM cases (11/103) and in 95% of the MMP cases (18/19). Among 29 cases with a living fetus at US, fetal abnormalities were registered in 14 cases. Of these, 3 cases had the parental type PPM, and 11 cases had the type MMP; the difference was not significant ( P = .16). Autopsy data on the fetuses are shown in Table 2 . A wide variety of phenotypes that ranged from near normal to highly abnormal was seen.
| Identification no. | Parental type | GA LMP at termination in wk + d | Histopathologic morphologic condition of placenta | Histopathologic observations in fetus |
|---|---|---|---|---|
| 70 | PPM | 12 + 2 | Normal | Intrauterine growth restriction; no external malformations |
| 40 | PPM | 13 + 6 | Pregnancy product | CRL, 7 cm |
| 71 | PPM | 18 + 3 | Normal | No external malformations, renal cysts bilaterally, dysplasia of the right kidney, and possible hypoplasia of the thymus |
| 77 | PPM | 19 | Partial mole | CRL, 12 cm; feet, 20 mm, which indicated gestational age at 16 weeks; sacral skin defect 8 mm; low set ears; bifid uvula; hyperplasia of Leydig cells, proliferation of adrenal cord |
| 120 | PPM | 20 + 1 | Partial mole | Total length, 19 cm; feet, 20 mm; cleft lip and palate |
| 81 | PPM | 22 | Partial mole | CRL, 3.5 cm; weight, 3.5 g; long body small head; adhesion of left upper arm to facial structures; adhesion between upper and lower leg; agenesis of hands and feet; anus abnormally placed in columnar; organs degenerated |
| 122 | PPM | 24 + 5 | Partial mole | Intrauterine growth restriction and deformities of feet (not specified) |
| 123 | PPM | 26 + 2 | Partial mole | Syndactyly of 3th and 4th finger and toe bilaterally and micro-penis; pulmonary hypoplasia; multicystic renal dysplasia; atresia of proximal ureters |
| 23 | MMP | 30 + 3 | Normal | Square-shaped cranium; low-set ears; narrow nose; syndactyly of the 3th and 4th finger on the left hand and of the 4th and 5th toe on both feet; migration defect in cerebrum; atrophic thymus; and perforated small intestine |
| 18 | MMP | 15 + 2 | Normal | Asymmetric growth retardation; low-set ears; increased distance between eyes; micrognathia, syndactyly of 3th and 4th fingers on the left hand and 2th and 3th toe on the right foot |
| 19 | MMP | 15 + 3 | Normal | No external malformations |
| 10 | MMP | 16 + 1 | NA | Asymmetric growth retardation a |
| 12 | MMP | 16 + 6 | NA | Small of date possibly anencephalic a |
| 14 | MMP | 18 + 4 | NA | Large head; low-set ears; cleft palate; myelomeningocele over spinal column; omphalocele a |
| 21 | MMP | 20 + 5 | Normal | Hydrocephaly; micrognathia; syndactyly of 3th and 4th finger on the left side and 3th and 4th toe on the right side; possible hypoplasia of the lungs |
MS-hCG
In 7 of 24 cases with the parental type MMP, data on MS-beta hCGs were retrieved (median, 0.17 MoM; range, 0.03–0.33 MoM). Among the 134 cases with the parental type PPM, MS-se-total hCGs were retrieved in 75 cases (median, 13.5 MoM; range, 0.25–1556.9 MoM; Figure ).
Comment
In Denmark, the health care system is public, and all citizens have a unique Civil Registration System number. Therefore, a systematic retrieval of samples and clinical data was possible. Ascertainment bias was minimized because cases were included consecutively and prospectively both in the DMP and in the DCCR.
The real frequency of triploidy is not known. An estimate of 1-2% of all clinically recognizable pregnancies, however, is accepted widely. Within the study period, there were a total of 606,512 live births in Western Denmark. The rate of induced and spontaneous abortions totals 60% of the birth-rate, giving an estimate of approximately 1.5 million recognizable pregnancies. We observed 183 triploid pregnancies, which equals 0.1% of all estimated clinically recognizable pregnancies and suggests that the frequency of triploid pregnancies previously has been overestimated. However, some triploid pregnancies are most likely missing in our cohort. In our study, no pregnancy terminated earlier than GA LMP 10 weeks was included. In addition, it is likely that not all triploid pregnancies that were terminated at higher gestational ages were included. Therefore, our data can by no means be taken as representative for “all triploid pregnancies.” However, because the triploid pregnancies diagnosed in Western Denmark are either registered in the DMP or the DCCR (or both), our observations are representative of triploid pregnancies that have been detected in a clinical setting.
The frequency of pregnancies with the 2 parental types
Reported frequencies of pregnancies with the 2 different parental types have varied pronouncedly ( Table 3 ). This variation most likely is due to differences in ascertainment. In the present study, we have found a striking predominance of cases with the parental type PPM (85%), but it should be pointed out that our estimate can be affected also by ascertainment bias, despite the fact that we have tried to ensure that we included all cases of triploidy from a defined area of Denmark by ascertaining from both the DMP and the DCCR.
| Study | Cohort | Wks’ gestation, range (median) | Population | Period | No. registered | No. examined | Method | Frequency of parental type: double paternal contribution, % |
|---|---|---|---|---|---|---|---|---|
| Szulman et al | Karyotyped spontaneous abortion | 6–27 (NA) | Laboratories Centres dÉtudes de Biologie Prénatale Paris, France | 1968-1971 | NA | 92 | Ratios between gonosomes | More frequent than double maternal contribution |
| Jacobs et al | Karyotyped spontaneous abortion | 7–31 (14) | One hospital, Honolulu, HI | 1978-1982 | 106 | 78 | Cytogenetic comparison of Q and C band heteromorphisms with those in parents | 73 |
| Uchida and Freeman | Karyotyped spontaneous abortion; abortions, liveborn, and stillborn infants suspected of chromosomal abnormality | NA | McMaster University of Health Sciences Center, Hamilton, Canada | 1977-1984 | 105 | 76 | Cytogenetic comparison of Q and C band heteromorphisms with those in parents | 62 |
| McFadden et al | Karyotyped specimens with intact fetus | 10–29 (NA) | Embryo Pathology Laboratory and Autopsy Service at British Colombia Children’s’ Hospital, Canada | 1987-1990 | 110 | 8 | Comparison of DNA markers with those in both parents | 25 |
| Miny et al | Chorionic villus samples; karyotyped because of maternal age or abnormal sonography | 8–33 (18) | Münster Chorionic Villus Sampling and Amniocentesis Program, Germany | NA | NA | 17 | Comparison of DNA markers with those in both parents | 29 |
| Baumer et al | Karyotyped 22 fetuses; karyotyped 3 liveborn probands | 8–36 (20) | NA | NA | NA | 25 | Comparison of DNA markers with those in both parents | 20 |
| Zaragoza et al | Karyotyped spontaneous abortion; abortions ascertained because of pathologic observations | 5–19 (12) | University Hospitals, Cleveland, OH, and Pittsburg, PA | NA | 91 | 87 | Comparison of DNA-markers with those in 1 or both parents | 69 |
| McFadden and Langlois | Karyotyped spontaneous abortion; abortions ascertained because of pathologic observations | NA–30 (NA) | Embryo pathology Laboratory and Autopsy Service at British Colombia Children’s’ Hospital, Canada; 1 case from another institution in the United States | NA | 41 | 38 | Comparison of DNA markers with those in 1 or both parents | 37 |
| Daniel et al | Karyotyped spontaneous abortion; abortions ascertained because of pathologic observations; 1 dysmorphic child | NA | NA | NA | 19 | 18 | Phenotype or comparison of DNA-markers with those in 1 or both parents or chromosome 15 methylation test | 61 |
| Present study | Triploid pregnancies registered in the Danish Cytogenetic Central Register a or the Danish Mole Project b | 10–30 (16) | Western Denmark | April 1986-April 2010 | 183 | 158 | Methylation sensitive multiplex ligation–dependent probe | 85 |
a Holds data on all karyotypes made for clinical purposes in Denmark since 1960
b Holds data on and samples of fresh placental tissue from conceptuses that gynecologists suspected to be hydatidiform moles.
Karyotypes
In pregnancies with the parental type MMP, we found no significant difference between the frequencies of the 2 possible karyotypes 69,XXX and 69,XXY ( P = .36). Similar results have been found by others.
In triploid pregnancies with the parental type PPM, it generally is believed that the cause behind these is dispermy. If this is true, it would be expected that the frequency of pregnancies with the various karyotypes would be 25% 69,XYY, 50% 69,XXY, and 25% 69,XXX. We found the karyotype 69,XYY in 6.1% of the pregnancies with the parental type PPM. This frequency is significantly higher than the frequency of 1% found by others and at the same time very far from the expected frequency of 25%. Interestingly, the frequency in our study is similar to the frequency that has been observed in preimplanted human triploid embryos. The difference in the frequencies of pregnancies with the karyotype 69,XYY observed by us and others may be explained by a difference in the frequencies of hydatidiform moles in the cohorts because all of the 7 cases with the karyotype 69,XYY in our study were classified as hydatidiform moles. Therefore, one could speculate that preimplantation growth and the growth of a hydatidiform mole require the function of fewer genes compared with the growth of a more differentiated conceptus that is less sensitive to the suggested selection against pregnancies with a deficiency of X chromosomal genes relative to autosomal genes in triploid pregnancies.
However, disregarding the assumed selection against pregnancies with the karyotype 69,XYY, dispermy should still result in twice as many pregnancies with the karyotype 69,XXY as pregnancies with the karyotype 69,XXX. In our study, pregnancies with these 2 karyotypes had similar frequencies, which raises the question whether dispermy is the only mechanism that causes PPM triploidy.
Placenta morphologic findings
Combining the data on triploid pregnancies of both parental types, molar appearance was found on US in 75 of 103 cases (73%). Our results contrast to the observation of molar appearance on US in 30% in triploid pregnancies, which was reported by Jauniaux et al. Our cohort consisted of all triploid pregnancies that were detected in a clinical setting in a well-defined geographic region. However, because the Departments of Gynecology participated in a research project on hydatidiform moles, the gynecologists may very well have been focused especially on forwarding samples with a molar morphologic finding for genetic analyses. This may have resulted in an overestimation of the proportion of molar placentas from triploid pregnancies in our study. In the study by Jauniaux et al, patients who were referred with placental abnormalities were excluded. This should result in an underestimation of the proportion of placentas with molar appearance. Thus, it seems plausible that the proportion of triploid conceptuses with molar appearance on US is somewhere between 30% and 59%.
Suspicion of a molar pregnancy based on US was raised in 75 of 103 PPM cases and in none of 22 MMP cases. Thus, in our cohort, the specificity and sensitivity for the detection of diandric triploids was 100% and 70%, respectively. All cases that fulfilled the DMP inclusion criteria had the parental type PPM, which indicates that this selection is highly specific for the parental type PPM.
At histopathologic examination of the 105 placentas, 74 were diagnosed as molar; they all had the parental type PPM. Among the 98 cases with the parental type PPM, 76% were molar; 12% were diagnosed with hydropic degeneration, and 12% were normal placentas. This is in agreement with the results from histopathologic examinations of spontaneous abortions with the parental type PPM reported by Zaragoza et al. In our cohort of triploid pregnancies, a histopathologic diagnosis of molar pregnancy would have had a specificity of 100% and a sensitivity of 76% for the parental type PPM.
The correlation between morphologic findings and parental type may have been influenced by the high frequency of PPM cases in our cohort. However, it seems safe to conclude that the specificity of suspicion of molar pregnancy by any of the morphologic methods for the prediction of diandric triploidy is high.
In contrast to others, we did not find a correlation between small placentas and the parental type MMP. However, we had information only on histopathologic morphologic findings of the placenta in 7 cases with the parental type MMP; hence, our results show only that normal placentas can be observed in triploid pregnancies with the parental type MMP.
Fetal morphologic findings
The frequency of fetal abnormalities detected by US in triploid pregnancies with a living fetus has been stated to be 93%, 66%, or 86%. In our study, fetal abnormalities at the initial US were registered in 48% (95% confidence interval, 29–69%) of the triploid pregnancies with a living fetus. The mean gestational age is available for 2 of the earlier studies: 11 and 20 weeks, respectively. The mean GA LMP in our study was 15 weeks. Thus, it is not likely that differences in gestational age explain the differences in the observed frequencies of triploid pregnancies with fetal abnormalities that are detected by US. However, ascertainment bias cannot be excluded as a cause of the discrepancies.
A variety of fetal phenotypes that ranged from nearly normal to highly abnormal were seen. No malformations that are pathognomonic for triploid fetuses were seen, which correlates well with the findings in previous studies.
Two distinct phenotypes have been proposed to be associated with triploidy after the gestational age 10 + 0 weeks: type 1 (with the parental type PPM) consists of relatively well-grown fetuses with proportionately sized body parts, and the associated placentas are abnormally large with the phenotype of partial hydatidiform mole; and type 2 (with the parental type MMP) consists of growth-retarded fetuses with pronounced wasting of the body and therefore relative macrocephaly; the associated placentas are noncystic and abnormally small.
A substantial fraction of our cases, which all had GA LMP above 10 + 0 weeks, did not comply with this categorization. In 6 of 8 PPM cases (75%) we found fetuses with IUGR; in 3 cases (38%), the placenta was normal. In 2 of 7 MMP cases (29%), no IUGR was found; in 4 of the 7 cases (57%), the placenta was normal.
In our study, a fetus was classified as subject to IUGR if the gestational age according to ultrasound biometric parameters was <7 days shorter than that calculated from the last menstrual period. Others have defined IUGR using development stage of the fetus that was seen in autopsy. Therefore, comparison of IUGRs must be made with caution. However, the combination of results from placental examination and IUGR data prompts us to suggest that the association of 2 distinct phenotypes with the 2 parental types is not entirely consistent.
MS-hCG
Most studies have reported high MS-total hCG in cases with the parental type PPM. Although we observed high values in many PPM cases, we also observed some PPM cases with low MS-total hCG values, which corroborates the observations in previous case reports. Like others, we found a low MS-beta hCG in all the triploid pregnancies with the parental type MMP. Thus, the use of a high MS-beta hCG value in a triploid pregnancy as a predictor of the parental type PPM has a high specificity, but a lower sensitivity.
By the correlation of triploid pregnancy and either suspicion of hydatidiform mole at US, at macroscopic inspection of the evacuated tissue, or at histologic examination and a high level MS-beta hCG, one can predict the parental type PPM with a very high specificity. In contrast, the sensitivity of these observations as predictors of this parental type is <100%.
Acknowledgments
The authors are grateful for the assistance from Professor Steen Kolvraa, DMSc, MedDr, the Department of Clinical Genetics, Vejle Hospital, and Institute of Regional Health Research, University of Southern Denmark for discussions and assistance in constructing the manuscript, and from Associate Professor Niels Tørring, PhD, at the Department of Clinical Biochemistry, Aarhus University Hospital-Skejby, for providing algorithms for calculations of MoM values. We are also grateful for the help supplied from the Departments of Gynecology and Obstetrics in Jutland when old medical records had to be retrieved. Finally, we would like to express our gratitude to the Departments of Clinical Genetics in Aarhus and Vejle for their help in retrieving triploid samples.
Appendix
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
