Multiplex short tandem repeat DNA analysis confirms the accuracy of p57KIP2 immunostaining in the diagnosis of complete hydatidiform mole

Human Pathology (2006) 37, 1426 – 1434

www.elsevier.com/locate/humpath

Original contribution

Multiplex short tandem repeat DNA analysis confirms the accuracy of p57KIP2 immunostaining in the diagnosis of complete hydatidiform mole Dorota A. Popiolek MDa,*, Herman Yee MD, PhDa, Khush Mittal MDa, Luis Chiriboga PhDa, Mechthild K. Prinz PhDb, Theresa A. Caragine PhDb, Zoran M. Budimlija MD, PhDa,b a

Department of Pathology, New York University School of Medicine, New York, NY 10016, USA Department of Forensic Biology, New York City Office of the Chief Medical Examiner, New York, NY 10016, USA

b

Received 26 July 2005; revised 13 April 2006; accepted 20 April 2006

Keywords: Hydatidiform mole; p57; Short tandem repeats

Summary Detailed histopathologic examination remains to be the basis for the diagnosis of hydatidiform mole (HM). However, poor sampling, necrosis, and earlier uterine evacuation can lead to uncertainty in the diagnosis. Also, the criteria are subjective, resulting in considerable interobserver variability. The p57KIP2 gene is paternally imprinted and maternally expressed, and the presence of its protein product serves as a surrogate marker for the nuclear maternal genome. Because a complete HM (CHM) is the only type of conceptus lacking a maternal contribution, p57KIP2 immunostaining is correspondingly absent, whereas it is present in CHM mimics. Although analysis of DNA microsatellite polymorphisms is a reliable method for the diagnosis and classification of HM, it is not universally available. To assess the relative accuracy of p57KIP2 immunostaining and molecular diagnosis by nuclear DNA microsatellite polymorphisms in discriminating CHM from its mimics, we analyzed archival tissue from 33 case patients (7 with a definitive diagnosis of CHM, 16 with a possible diagnosis of HM, and 10 with normal placentas) by both methods. Concordant results were obtained in all cases, and p57KIP2 immunostaining accurately identified all cases of CHM from the groups with a definitive or possible diagnosis of HM. p57KIP2 immunohistochemistry is a time- and cost-effective means of distinguishing CHM from its mimics in challenging cases. D 2006 Elsevier Inc. All rights reserved.

1. Introduction Accurate diagnosis and classification of hydatidiform mole (HM) are important because the risk of persistent * Corresponding author. Department of Pathology, New York University Medical Center, New Bellevue Hospital, New York, NY 10016, USA. E-mail address: [email protected] (D. A. Popiolek). 0046-8177/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2006.04.030

gestational trophoblastic disease, including choriocarcinoma, is significantly higher after a pregnancy affected by a complete HM (CHM; 10%-30%) or a partial HM (PHM; 0.5%-5%) [1-4] as compared with any other conception. Genetically, most CHMs are monospermic and arise by fertilization of an anucleated egg by a haploid sperm followed by endoreduplication [5-7]. Approximately 20% of CHMs are dispermic [8,9], arising by fertilization of an anucleated

Multiplex short tandem repeat DNA analysis

1427

egg by either 2 sperms [10] or a single diploid sperm, the result of a failure of the second meiotic division [6]. In either case, nearly all CHMs have a diploid karyotype that is entirely of paternal (androgenetic) origin. An exception is the rare familial case of CHM that, although diploid, is biparental, rather than androgenetic, in origin [9,11-14]. In contrast, cases of PHM are triploid and usually result from fertilization of an ovum by 2 sperms [15,16], although fertilization by a single diploid sperm cannot be excluded [17]. Although HM and its subtypes are defined genetically, most cases of HM can be identified correctly by histologic analysis if adequate tissue is submitted and appropriate criteria are used. However, there is a subset of cases in which poor sampling, necrosis, and earlier uterine evacuation can make pathologic diagnosis difficult [18,19]. This has become more evident with routine use of first-trimester ultrasound examination and maternal chorionic gonadotropin monitoring in the management of pregnancy, which has Table 1

resulted in most molar pregnancies being evacuated in the late first trimester [20], before the development of the classic histologic features [21,22]. Also, considerable overlap in the histologic appearance of molar and nonmolar pregnancies with abnormal villous morphology and that between CHM and PHM have resulted in significant interobserver variability in the diagnosis of HM and its mimics [19,23-26]. Therefore, this subset of difficult cases may require application of molecular methods to reach a definitive diagnosis [22,27-31]. The analysis of nuclear DNA microsatellite polymorphisms is especially well suited for this purpose because it is capable of determining the number and parental origin of alleles in even small amounts of fixed tissue from paraffin blocks [32-39]. However, such molecular diagnosis (MDx) methods are technically difficult, relatively costly, and not universally available. Thus, a time- and cost-effective ancillary tool that is available in most laboratories would be helpful.

Clinical and pathologic characteristics of cases designated as PosHM

Case

Clinical diagnosis

Pathologic diagnosis

Reasons for ambiguity

p57KIP2 results

MDx

1

MAB

Rare hydropic villi b2 mm; minimal focal trophoblast hyperplasia; rare inclusions

+

PHM

2

MAB

Rare hydropic villi b1 mm; minimal focal trophoblastic hyperplasia

+

PHM

3

MAB

Descriptive; features suggestive of chromosomal anomalies, including PHM Descriptive; features suggestive of chromosomal anomalies, including PHM CVs with hydropic change

+

PHM

4

MAB

+

PHM

5

MAB

+

PHM

6

MAB

+

PHM

7

MAB

+

PHM

8

MAB

9

MAB

10

MAB

Descriptive; features suggestive but not diagnostic of PHM Descriptive; features suggestive but not diagnostic of PHM Descriptive; features suggestive but not diagnostic of PHM Descriptive; features suggestive of chromosomal anomalies, including PHM Descriptive; features suggestive but not diagnostic of HM Descriptive; features suggestive but not diagnostic of HM Consistent with HM

Rare hydropic villi b1 mm; minimal focal trophoblastic hyperplasia Rare enlarged villi b2 mm; irregular outlines; no trophoblastic proliferation Rare hydropic villi b1 mm; minimal focal trophoblastic hyperplasia Rare hydropic villi b1 mm; minimal focal trophoblastic hyperplasia; rare inclusions Rare hydropic villi b2 mm; irregular outlines; minimal focal trophoblastic hyperplasia

11

IAB

Consistent with HM

12

IAB

13

HM

14 15

HM None given

16

None given

Descriptive; features suggestive but not diagnostic of HM Descriptive; features suggestive but not diagnostic of HM Consistent with HM Descriptive; features suggestive but not diagnostic of HM Descriptive; features suggestive but not diagnostic of HM

Abbreviations: MAB, missed abortion; IAB, incomplete abortion.

Few bulbous villi b2 mm; focal trophoblastic proliferation but with loss of polarity Few bulbous villi b2 mm; no trophoblastic proliferation Rare hydropic villi b2 mm; trophoblastic proliferation Few enlarged villi b3 mm; trophoblastic proliferation Hydropic villi b2 mm; focal trophoblastic proliferation Few hydropic villi b3 mm; focal trophoblastic hyperplasia Few enlarged villi of 4 mm; blood vessels noted Few bulbous villi b2 mm; focal trophoblastic proliferation but with loss of polarity Few hydropic villi b1 mm; focal and polar trophoblastic hyperplasia

CHM CHM CHM CHM CHM CHM CHM CHM CHM

1428

D. A. Popiolek et al.

The p57KIP2 is a paternally imprinted gene expressed only from the maternal allele [40]. As expected, lack of its protein product has been demonstrated in immunohistochemistry (IHC)-based studies on CHM but not in those on non-CHM, suggesting that p57KIP2 IHC can be helpful in distinguishing CHM from its mimics [29,41-46]. Most reported series assessed the usefulness of p57KIP2 IHC in morphologically unequivocal cases of CHM, and only few studies so far have compared the accuracy of p57KIP2 IHC with that of other methods, such as ploidy and other DNA techniques, and have involved only a small number of cases [29,43,45,47]. The present validation study compared the accuracy of p57KIP2 IHC with that of nuclear DNA microsatellite polymorphism in identifying CHM.

2. Materials and methods 2.1. Case selection Seven cases of morphologically unequivocal CHM, all of which were suspected clinically, and 16 cases with a histopathologic diagnosis reflecting uncertainty (ie,

Fig. 1

bcannot rule out HM, suggest hCG [human chorionic gonadotropin] follow-upQ) were selected at random from the files of the gynecologic pathology division of the New York University and Bellevue hospitals. In most of those cases, scant and/or necrotic material combined with ambiguous histologic features such as lack of uniformly swollen villi with central cisterna and/or absence of circumferential trophoblast hyperplasia accounted for the difficulty in the diagnosis (Table 1). These 16 cases were designated as bpossible HMQ (PosHM) for this study. In 14 of them, the clinical diagnosis was known, being bmissed abortionQ in 12 cases and bconsistent with molar pregnancyQ in 2. Histologic evaluation of all cases was performed on routine hematoxylin-eosin–stained sections; diagnoses were made by gynecologic pathologists using published criteria [22,25]. Ten cases of products of conception from elective termination of pregnancies with morphologically normal chorionic villi (CVs) were also included to validate the procedures used in DNA procurement and analysis.

2.2. Immunohistochemistry and DNA analysis The IHC and DNA analysis were performed on formalinfixed and paraffin-embedded tissue from 33 case patients.

Allelic profile of normal CVs illustrating heterozygous loci with 2 balanced peaks (PowerPlex16 multiplex).

Multiplex short tandem repeat DNA analysis We used StreptABC (Ventana Medical System, Tucson, AZ) and mouse monoclonal antibody against recombinant human p57KIP2 protein (clone 57P06; Lab Vision Corporation, Fremont, CA). Sections were incubated with antibody overnight at room temperature. Antigen retrieval was performed with 10 mmol/L of sodium citrate buffer, pH 6.0, for 20 minutes. Sections were counterstained with hematoxylin. Only distinct nuclear p57KIP2 staining was scored as positive. For all cases, the presence or absence of nuclear staining was assessed in 5 cell types (cytotrophoblast [CT], villous stroma [VS], syncytiotrophoblast [ST], intervillous trophoblast islands [IVTIs], and decidua [D]) as reported in the literature [42]. CVs and D were laser capture microdissected from hematoxylin-eosin–stained sections using a PixCell II LCM system (Arcturus Engineering, Mountain View, CA). The DNA extraction was performed after the standard procedure using Eppendorf phase lock gel tubes and Microcon 100 columns (Millipore, Bedford, MA). Short tandem repeat (STR) loci were evaluated from the paired-tissue DNA subfractions of each sample using a PowerPlex16 multiplex STR kit (Promega, Madison, WI), which allows amplification and fluorescent analysis of 16 loci simultaneously, in a GeneAmpPCR System 9700 instrument (Applied Biosystems, Foster City, CA). Allele designations were determined by comparison of the samples with those of the allelic ladders supplied with the original kit. The origin (maternal versus paternal) of alleles in the villous tissue was again determined by comparison with the corresponding alleles in the maternal tissue. The absence of a maternal genome (as in CHM) was indicated by the exclusion of maternal alleles in the CV tissue. Determination of the number of alleles in heterozygous loci was based on the knowledge of a normal heterozygous peak pattern (maximum, 2 peaks) observed in a typical single-source sample [48], in which the peaks correspond to the maternal/ paternal allele at a particular locus and are expressed in relative fluorescence units (RFU). In healthy individuals, the heterozygous allele peaks are balanced, which means that the RFU peak height ratio, measured by dividing the height of the lower-quantity allele peak by the height of the higherquantity allele peak in RFU, should be greater than 0.7. In addition, allele peak patterns for heterozygous loci generally have stutter products that are lower than 0.15 of the associated peak height. Thus, if the RFU peak height ratio falls between 0.15 and 0.7 (a phenomenon known as peak imbalance), then an abnormal number of alleles, such as an additional allele, is probable. Therefore, for the purpose of this study, an occurrence of peak imbalance in cases with possible molar pregnancy was interpreted as evidence of an additional overlapping allele (2 peaks but 3 alleles in a given locus). Furthermore, if 2 of those 3 alleles were of paternal origin, then an MDx of PHM was confirmed. To establish a cutoff value of the RFU peak height ratio for identification of loci with peak imbalance, we had the allelic profiles of 10 cases of products of conception with normal villous morphology determined by PowerPlex16. The mean RFU

1429 peak height ratio and standard deviation for all heterozygous loci of CVs were determined as 0.87 and 0.08, respectively, a finding in agreement with that reported in the forensic literature [48]. Based on this observation, the RFU peak height ratio cutoff value was set at 0.71 (mean, 2SD). Accordingly, any heterozygous locus with an RFU peak height ratio of 0.71 or higher was interpreted as having balanced peaks and 2 alleles (Fig. 1), whereas any heterozygous locus with an RFU peak height ratio lower than 0.71 but greater than 0.15 was interpreted as having peak imbalance and thus 3 alleles, 2 of which are identical (overlapping). Then, the origin of this extra allele in the villous tissue was determined by comparison with the corresponding maternal tissue. MDx of CHM was made if at least 1 locus with alleles (homozygous = monospermy; heterozygous = dispermy) that could not have come from the mother was identified. For an MDx of PHM, identification of at least 1 locus with 2 of 3 alleles that could not have come from the mother was required. The maternal DNA in one case was scant as well as highly degraded and so its allelic profile was quantified using a

Fig. 2 IHC staining for p57KIP2. A, Expression in normal placenta with nuclear staining of the CT and, to a lesser degree, VS (original magnification 100). B, Lack of expression in the CT and VS of unequivocal CHM (original magnification 100).

1430 method described by Nicklas and Buel [49]. The DNA extract was amplified using an AmpFISTR Identifiler kit (Applied Biosystems), and the amplified product was analyzed like the

D. A. Popiolek et al. other samples in an ABI Prism 3100 Genetic Analyzer using GeneScan and Genotyper software (Applied Biosystems) according to the manufacturer’s recommendations.

Fig. 3 Case designated as PosHM negative for p57KIP2 in which the diagnosis of CHM was based on the allelic profile of (A) CVs with all loci being homozygous and diagnostic loci (arrows) that could not have come from the mother and (B) corresponding maternal D (PowerPlex16 multiplex).

Multiplex short tandem repeat DNA analysis

1431

Fig. 4 Case designated as PosHM positive for p57KIP2 in which the diagnosis of PHM was based on the allelic profile of (A) CVs with 2 types of diagnostic loci (arrows)—(1) with 3 peaks (D21, D13, D16) and (2) with 2 imbalanced peaks (PowerPlex16 multiplex)—and (B) corresponding D (AmpFlSTR Identifiler; diagnostic loci shown in order of PowerPlex16 multiplex electropherogram).

1432

3. Results 3.1. Immunohistochemistry All 10 normal placentas strongly expressed p57KIP2 in the CT and, to a lesser degree, VS (Fig. 2A). In contrast, all 7 cases of unequivocal CHM lacked immunoreactivity for p57KIP2 in the CT and VS (Fig. 2B). Among the 16 cases of PosHM, 9 were negative and the remaining 7 were positive for p57KIP2 in the CT and VS. All 33 cases were negative for p57KIP2 in the ST, whereas the IVTIs expressed p57KIP2 in all of them. Also, maternal D was immunoreactive with p57KIP2 and served as an internal positive control.

D. A. Popiolek et al. which the RFU peak height ratio ranged from 0.34 to 0.70 (mean, 0.58), and thus were interpreted as having 3 alleles (2 of which were identical) of which 2 could not have come from the mother (Fig. 4). In 2 cases, there were 4 diagnostic loci with 3 peaks (1 locus in 1 case, 3 loci in 1 case) and diagnostic loci with 2 imbalanced peaks. In the remaining 5 cases, the MDx of PHM was based on the presence of loci with 2 imbalanced peaks. The p57KIP2 IHC technique accurately identified all cases of CHM from both groups, as judged by MDx, and there was no discrepancy between the p57KIP2 IHC and molecular analysis results ( P = .0005; Fisher’s exact test).

3.2. DNA analysis Five of the 7 unequivocal cases of CHM had adequate DNA from both the CVs and D to determine allelic profiles, and an MDx of CHM was made in all of them. The numbers of diagnostic loci in those cases were 1, 4, 5, 8, and 9 in 1 case each. Four of the 5 lesions were homozygous (CVs) in all loci, including the amelogenin (Amel) locus, which was determined as X. Therefore, they were androgenetic and monospermic in origin. The remaining case with adequate DNA was also androgenetic but contained heterozygous loci with balanced peaks and homozygous loci, a result of dispermy. An additional case in this group had all homozygous loci in the CVs but lacked the allelic profile of the matching maternal tissue, and the origin of the CV alleles (maternal versus paternal) could not be determined. The finding of homozygosity in all loci would be unexpected by chance, and this lesion most likely represents an androgenetic and monospermic CHM. The last case in this group had 7 heterozygous loci with 2 balanced peaks that had at least one paternal allele in each of them. However, the paternal origin of the other alleles in these loci could not be confirmed because the mother shared these alleles; in addition, a paternal allelic profile was needed but was not available to establish their origin. Nevertheless, the allelic profile in this case was compatible with dispermic CHM. In the group of PosHM cases, in 7 of the 9 p57KIP2negative cases in which sufficient DNA was available in the CVs and D, an MDx of CHM was made. Six of them were monospermic (Fig. 3), whereas the remaining case was dispermic (Amel locus determined as XY); the number of diagnostic loci ranged from 1 to 7 (1 in 3 cases and 2, 4, 5, and 7 in 1 case each). In 2 of the 9 p57KIP2-negative cases, an MDx could not be made because of the extent of DNA degradation. In all 7 p57KIP2-positive PosHM cases, an MDx of PHM was made, and the number of diagnostic loci ranged from 1 to 11 (1 in 2 cases, 3 in 3 cases, and 4 and 11 in 1 case each). Two types of diagnostic loci were identified: (1) those with 3 peaks, indicating the presence of 3 alleles of which only 1 came from the mother (the other 2 therefore must have been of paternal origin), and (2) those with 2 imbalanced peaks, in

4. Discussion The diagnosis and classification of HM have long been based on well-established histologic criteria (eg, degree of hydrops, trophoblastic hyperplasia, fiord-like versus smooth bulbous villous outlines, myxoid VS, evidence of embryonic development) in nearly all cases [21]. However, the routine use of ultrasound scans in early gestation has led to evacuation of most molar pregnancies in the late first trimester (8.5-12 weeks of gestation versus 16-18 weeks of gestation in the past), before the classic morphological features of CHM (cisternae and circumferential trophoblastic proliferation) are well developed [25]. Moreover, it has been recognized that a variety of other genetic abnormalities such as trisomy/monosomy may be associated with abnormal villous morphology with significant morphological overlap with early molar pregnancy [50]. Although new and expanded histologic criteria for the diagnosis of these early moles have been proposed [25,50,51], there are still considerable subjectivity and overlap of histologic features, resulting in poor interobserver reproducibility in the classification of placentas with hydropic changes in a subset of cases [26,52]. More sophisticated molecular methods, including DNA microsatellite polymorphism, have subsequently been applied in these challenging cases to resolve the diagnostic dilemma, but these assays are not available in most pathology laboratories [22,28,29,31-37,47]. On the other hand, IHC has become a widely used adjunct in diagnostic surgical pathology. The p57KIP2 gene is paternally imprinted and is expressed from the maternal allele, and the lack of its protein product, as detected by IHC, has been documented in CHM (because CHMs lack maternal genomic DNA) [41-47]. In contrast, its mimics express p57KIP2, which serves as a surrogate marker for maternal DNA. The goal of this study was to determine the accuracy of p57KIP2 IHC in the identification of CHM as compared with MDx based on DNA microsatellite polymorphism. The p57KIP2 IHC technique identified all the cases of CHM from both the morphologically unequivocal group and the ambiguous group with no discrepant result ( P = .0005; Fisher’s exact test). In agreement with most published results, p57KIP2 nuclear immunostaining was present in the CT and VS in

Multiplex short tandem repeat DNA analysis non-CHM but absent in CHM. Although genomic imprinting, a phenomenon underlying the pathophysiology of HM, explains the differential expression of p57KIP2 in the CT and VS in CHM versus non-CHM, the presence of p57KIP2 in intervillous trophoblastic islands and its absence from the ST in all cases remain enigmatic. p57KIP2, a cdk inhibitor, functions as a proliferation-limiting factor [53], and coexpression of Ki-67 and p57KIP2 (double immunostaining) was evident in the CT and intermediate extravillous trophoblast (IVTIs and implantation site trophoblast) but was absent in the ST in normal placenta [41]. This finding suggests that terminally differentiated ST, regardless of the type of gestation, has lost the expression of p57KIP2 along with the ability to divide (adaptive mechanism). The unexpected p57KIP2 immunostaining in IVTIs in CHM might be explained by a relaxation of imprinting (emergence of expression of alleles subject to repression according to their paternal origin). A similar expression pattern of another imprinted and maternally expressed gene, H19, was reported in CHM, and it was suggested that it also represents relaxation of imprinting [54]. On the other hand, Chilosi et al [41], in contrast to other investigators, reported on lack of p57KIP2 expression in CHM not only in the CT and VS but also in the extravillous trophoblast (IVTIs and implantation site with double immunostaining with cytokeratin). Also, 3 cases of gestational choriocarcinoma did not exhibit any detectable p57KIP2 by IHC. Thus, their findings were supportive of the hypothesis that deregulation of genomic imprinting, particularly the loss of cell-cycle inhibitors such as p57KIP2, is involved in abnormal androgenetic trophoblastic proliferation [41]. In this study, double immunostaining for cytokeratin and p57KIP2 was not performed; thus, the distinction of the mononuclear cells of the invasive intermediate trophoblast (implantation site) from D was difficult and loss of p57KIP2 expression in this population of cells could not be determined with certainty. Although DNA microsatellite polymorphism has been used to resolve diagnostic dilemmas centered on HM, this study applied a commercially available kit, PowerPlex16, that simultaneously tests 16 STR loci (D3S1358, THO1, D21S11, D18S51, PENTA E, D5S818, D13S317, D7S820, D16S539, CSF1PO, PENTA D, VWA, D8S1179, TPOX, FGA, and the Amel locus on chromosomes X and Y) and represents the largest number of such validated cases. The high number of loci tested simultaneously saved time and was helpful in cases with partial allelic profiles secondary to DNA degradation. In agreement with the reported results, most cases of CHM were monospermic, with approximately 20% resulting from dispermy or fertilization by a single diploid sperm. The finding that the MDx of PHM was based on the presence of 2 imbalanced peaks rather than 3 balanced peaks in most cases was unexpected but can be explained in part by allele sharing between the mother and the father as well as the fact that the products of conception (CVs) and maternal D have close relatedness.

1433 In summary, our study has shown that p57KIP2 IHC, a relatively simple technique, can be used as successfully as a very-high–technology method (DNA microsatellite polymorphism) for distinguishing CHM from its mimics in a subset of challenging cases.

References [1] Berkowitz RS, Goldstein DP. Diagnosis and management of primary hydatidiform mole. Obstet Gynecol Clin North Am 1998;15:491 - 503. [2] O’Quinn AG, Barnard DE. Gestational trophoblastic diseases. In: DeCherney AH, Pernoll ML, editors. Current obstetric and gynecologic diagnosis and treatment. 8th ed. Norwalk (Conn)7 Appleton & Lange; 1994. p. 967 - 76. [3] Rice LW, Berkowitz RS, Lage JM, et al. Persistent gestational trophoblastic tumor after partial hydatidiform mole. Gynecol Oncol 1990;36:358 - 62. [4] Scooper JT, Lewis Jr JL, Hammond CB. Gestational trophoblastic disease. In: Hoskins WJ, Perez CA, Young RC, editors. Principles and practice of gynecologic oncology. 2nd ed. Philadelphia7 LippincottRaven; 1997. p. 1039 - 77. [5] Lawler SD, Pickthall VJ, Fisher RA, Povey S, Evans MW, Shulman AE. Genetic studies of complete and partial hydatidiform moles. Lancet 1979;2:580. [6] Jakobs PA, Wilson CM, Sprenkle JA, Rosenshein NB, Migeon BR. Mechanism of origin of complete hydatidiform moles. Nature 1980;286:714 - 6. [7] Lawler SD, Povey S, Fisher RA, Pickthall VJ. Genetic studies on hydatidiform moles: II. The origin of complete moles. Ann Hum Genet 1982;246:209 - 22. [8] Fisher RA, Povey S, Jeffreys AJ, Martin CA, Patel I, Lawler SD. Frequency of heterozygous complete hydatidiform moles, estimated by locus-specific minisatellite and Y chromosome–specific probes. Hum Genet 1989;8:259 - 63. [9] Kovaks BW, Shahbahrami B, Tast DE, Curtin JP. Molecular genetic analysis of complete hydatidiform moles. Cancer Genet Cytogenet 1991;54:143 - 52. [10] Ohama K, Kajii T, Okamoto E, et al. Dispermic origin of XY hydatidiform moles. Nature 1981;29:551 - 2. [11] Jacobs PA, Hunt PA, Matsuura JS, Wilson CC, Szulman AE. Complete and hydatidiform mole in Hawaii: cytogenetics, morphology and epidemiology. Br J Obstet Gynaecol 1982;89:258 - 66. [12] Helwami MN, Seoud M, Zahed L, Zaatari G, Khalil A, Slim R. A familial case of recurrent hydatidiform molar pregnancies with biparental genomic contribution. Hum Genet 1999;105:112 - 5. [13] Fisher RA, Khatoon R, Paradinas FJ, Roberts AP, Newlands ES. Repetitive complete hydatidiform mole can be biparental in origin and either male or female. Hum Reprod 2000;15:594 - 8. [14] Sensi A, Gualandi F, Pittalis MC, et al. Mole marker phenotype: possible narrowing of the candidate region. Eur J Hum Genet 2000;8: 641 - 4. [15] Lawler SD, Fisher RA, Pickthall VJ, Povey S, Evans MW. Genetic studies on hydatidiform moles: I. The origin of partial moles. Cancer Genet Cytogenet 1982;5:309 - 20. [16] Jacobs PA, Szulman AE, Funkhouser J, Matsuura JS, Wilson CC. Human triploidy: relationship between parental origin of the additional haploid complement and development of partial hydatidiform mole. Ann Hum Genet 1982;46:223 - 31. [17] Zaragoza MV, Surti U, Redline RW, Millie E, Chakravarti A, Hassold TJ. Parental origin and phenotype of triploidy in spontaneous abortions: predominance of diandry and association with the partial hydatidiform mole. Am J Hum Genet 2000;66:1807 - 20. [18] Paradinas FJ. The histological diagnosis of hydatidiform moles. Curr Diagn Pathol 1994;1:24 - 31.

1434 [19] Genest DR. Partial hydatidiform mole: clinicopathological features, differential diagnosis, ploidy and molecular studies, and gold standards for diagnosis. Int J Gynecol Pathol 2001;20:315 - 22. [20] Berkowitz RS, Goldstein DP. The diagnosis of molar pregnancy by ultrasound: a continuing challenge. Ultrasound Obstet Gynecol 1997; 9:4 - 5. [21] Szulman AE, Surti U. The syndromes of hydatidiform mole. Morphologic evaluation of the complete and partial mole. Am J Obstet Gynecol 1978;132:20 - 7. [22] Paradinas FJ, Browne P, Fisher RA, Foskett M, Bagshawe KD, Newlands E. A clinical, histopathological and flow cytometric study of 149 complete moles, 146 partial moles and 107 non-molar hydropic abortions. Histopathology 1996;28:101 - 10. [23] Conran RM, Hitchcock CL, Popek EJ, et al. Diagnostic considerations in molar pregnancy. Hum Pathol 1993;24:41 - 8. [24] Messerli ML, Parmley T, Wodruff JD, et al. Inter- and intrapathologist variability in the diagnosis of gestational trophoblastic neoplasia. Obstet Gynecol 1987;69:622 - 6. [25] Keep D, Zaragoza MV, Hassold T, et al. Very early complete hydatidiform mole. Hum Pathol 1996;27:708 - 13. [26] Fukunaga M, Katabuchi H, Nagasaka T, et al. Interobserver and intraobserver variability in the diagnosis of hydatidiform mole. Am J Surg Pathol 2005;29:942 - 7. [27] Schorge JO, Goldstein DP, Bernstein MR, et al. Recent advances in gestational trophoblastic disease. J Reprod Med 2000;45:692 - 700. [28] Bateman AC, Hemmatpour SK, Theaker JM, et al. Genetic analysis of hydatidiform moles in paraffin embedded tissue using rapid, sequence-specific PCR-based HLA class II typing. J Clin Pathol 1997;50:288 - 93. [29] Crisp H, Burton JL, Stewart R, et al. Refining the diagnosis of hydatidiform mole: image ploidy analysis and p57kip2 immunohistochemistry. Histopathology 2003;43:363 - 73. [30] Lage JM, Mark SD, Roberts DJ, et al. A flow cytometric study of 137 fresh hydropic placentas: correlation between types of hydatidiform moles and nuclear DNA ploidy. Obstet Gynecol 1992;79:403 - 10. [31] Lai CY, Chan KY, Khoo US, et al. Analysis of gestational trophoblastic disease by genotyping and chromosome in situ hybridization. Mod Pathol 2004;17:40 - 8. [32] Fisher RA, Newlands ES. Gestational trophoblastic disease, molecular and genetic studies. J Reprod Med 1998;43:87 - 97. [33] Fukuyama R, Takata M, Kudoh J, Sakai K, Tamura S, Shimizu N. DNA diagnosis of hydatidiform mole using the polymerase chain reaction. Hum Genet 1991;87:216 - 8. [34] Fisher RA, Newlands ES. Rapid diagnosis and classification of hydatidiform moles with polymerase chain reaction. Am J Obstet Gynecol 1993;168:563 - 9. [35] Lane SA, Taylor GR, Ozols B, Quirke P. Diagnosis of complete molar pregnancy by microsatellites in archival material. J Clin Pathol 1993;46:346 - 8. [36] Abeln ECA, Cornelisse CJ, Dreef EJ, Kuipers-Dijkshoorn NJ, Hogendoorn PCW, Pancras CW. Molecular identification of a partial hydatidiform mole. Diagn Mol Pathol 1997;6:58 - 63. [37] Bell KA, Van Derlin V, Addya K, et al. Molecular genetic testing from paraffin-embedded tissue distinguishes nonmolar hydropic abortion from hydatidiform mole. Mol Diagn 1999;4:11 - 9.

D. A. Popiolek et al. [38] Arm MF, Fisher RA, Foskett MA, Paradinas FJ. Triplet pregnancy with hydatidiform mole. Int J Gynecol Cancer 2000;10:76 - 81. [39] Weaver DT, Fisher RA, Newlands ES, Paradimas FJ. Amniotic tissue in complete mole can be androgenetic. J Pathol 2000;191:67 - 70. [40] Hatada I, Mukai T. Genomic imprinting of p57KIP2, a cyclindependent kinase inhibitor, in mouse. Nat Genet 1995;11:204 - 6. [41] Chilosi M, Piazzola E, Lestani M, et al. Differential expression of p57KIP2, a maternally imprinted cdk inhibitor, in normal human placenta and gestational trophoblastic disease. Lab Invest 1998;78: 269 - 76. [42] Castrillon DH, Sun D, Weremowicz S, Fisher RA, Crum CP, Genest DR. Discrimination of complete hydatidiform mole from its mimics by immunohistochemistry of the paternally imprinted gene product p57KIP2. Am J Surg Pathol 2001;25:1225 - 30. [43] Genest DR, Dorfman DM, Castrillon DH. Ploidy and imprinting in hydatidiform moles. Complementary use of flow cytometry and immunohistochemistry of the imprinted gene product p57KIP2 to assist molar classification. J Reprod Med 2002;47:342 - 6. [44] Fukunaga M. Immunohistochemical characterization of p57KIP2 expression in early hydatidiform moles. Hum Pathol 2002;33: 1188 - 92. [45] Jun SY, Ro JY, Kim KR. p57kip2 is useful in the classification and differential diagnosis of complete and partial moles. Histopathology 2003;43:17 - 25. [46] Popiolek D, Yee H, Perle MA, Mittal K, Demopoulos RI. p57KIP2, a maternally expressed cdk inhibitor, a useful marker for separating complete mole from other gestations with changes suggestive of molar pregnancy/chromosomal abnormalities. Mod Pathol 1992;15:206A. [47] Merchant SH, Amin MB, Viswanatha DS, et al. p57KIP2 immunohistochemistry in early molar pregnancies: emphasis on its complementary role in the differential diagnosis of hydropic abortus. Hum Pathol 2005;36:180 - 6. [48] Butler JM. Forensic issues, degraded DNA, PCR inhibition, contamination, and mixed samples. In: Butler JM, editor. Forensic DNA typing. 1st ed. London (San Diego)7 Academic Press; 2001. p. 99 - 116. [49] Nicklas JA, Buel E. Development of an Alu-based, QSY 7–labeled primer PCR method for quantification of human DNA in forensic samples. J Forensic Sci 2003;48:282 - 91. [50] Chew SH, Perlman EJ, Williams R, Kurman RJ, Ronnett BM. Morphology and DNA content analysis in the evaluation of first trimester placentas for partial hydatidiform mole (PHM). Hum Pathol 2000;31:914 - 24. [51] Sebire NJ, Fisher RA, Rees HC. Histopathological diagnosis of partial and complete hydatidiform mole in the first trimester of pregnancy. Pediatr Dev Pathol 2002;6:69 - 77. [52] Paradinas FJ. The diagnosis and prognosis of molar pregnancy: the experience of the National Referral Center in London. Int J Gynecol Obstet 1998;60:S57 - S64. [53] Matsuoka S, Edwards MC, Bai C, et al. Imprinting of the gene encoding a human cyclin-dependent kinase inhibitor, p57KIP2, on chromosome 11p15. Proc Natl Acad Sci 1996;93:3026 - 30. [54] Ariel I, Lustig O, Oyer CE, et al. Relaxation of imprinting in trophoblastic disease. Gynecol Oncol 1994;53:212 - 9.