Pathology and molecular genetics of trophoblastic disease

Expectant Management of a Twin Pregnancy with Complete Hydatidiform Mole and Coexistent Normal Fetus

Twin pregnancies complicated by complete hydatidiform mole coexisting with a viable fetus are rare and may result in significant complications. We describe the expectant management and our surgical approach in a 27-year-old Rh-negative woman presenting with recurrent episodes of vaginal bleeding and a twin pregnancy consisting of a molar pregnancy coexisting with a normal fetus. Inpatient management was undertaken with close maternal and fetal monitoring until cesarean delivery of a healthy female infant and histopathologically confirmed complete hydatidiform molar pregnancy (karyotype 46XX) at 34 weeks with no evidence of malignancy.

Current issues in the histopathology of gestational trophoblastic tumors

Gestational trophoblastic neoplasia (GTN) encompasses several entities including complete (CHM) and partial (PHM) hydatidiform mole (HM) and the malignant gestational trophoblastic tumors (GTTs), choriocarcinoma (CC), and placental-site trophoblastic tumor (PSTT), including epithelioid trophoblastic tumor (ETT). To detect pGTN, postmolar surveillance by measurement of maternal human chorionic gonoadotropin (hCG) levels should be performed. With such a protocol, many cases of pGTN are identified early at a presymptomatic stage based on plateuing or rising hCG concentrations and subsequently treated successfully with chemotherapy. In such cases, histopathological confirmation of the precise nature of the pGTN usually is not available. However, GTT also may present clinically with primary or metastatic disease, either following and unrecognized HM or developing from a nonmolar gestation. Due to their distinctive clinical and histological features, malignant GTTs are generally clearly subdivided into CC and PSTT (including ETT). CC essentially represents malignant trophoblastic tumors with differentiation toward villous trophoblast, with extensive hematogenous spread and high hCG levels, which are highly chemoresponsive. However, PSTTs, represent malignant differentiation toward implantation-site type trophoblast, with lower hCG levels and less response to chemotherapy. Current issues regarding the clinical and histological features of CC and PSTT/ETT are discussed.

Differential effects of gonadotropin-releasing hormone I and II on the urokinase-type plasminogen activator/plasminogen activator inhibitor system in human decidual stromal cells in vitro

To date, the factors capable of regulating the coordinate expression of the urokinase-type plasminogen activator (uPA) and its endogenous inhibitor, plasminogen activator inhibitor (PAI-1), at the maternal-fetal interface remain poorly characterized. In these studies we examined the ability of the classical form of gonadotropin-releasing hormone (GnRH) I and the second, mammalian form of this hormone, GnRH II, to regulate uPA and PAI-1 mRNA and protein expression levels in cultures of stromal cells isolated from first trimester decidual tissues using quantitative competitive-PCR and ELISA, respectively. GnRH I and GnRH II increased uPA mRNA and protein expression levels in these primary cell cultures in a dose- and time-dependent manner. In contrast, GnRH I increased, whereas GnRH II decreased PAI-1 mRNA and protein expression levels in these cells. Cetrorelix, a GnRH receptor antagonist, inhibited the regulatory effects of GnRH I, but not GnRH II, on uPA and PAI-1 expression levels in these decidual stromal cell cultures. Taken together, these observations suggest that GnRH I and GnRH II differentially regulate the balance between uPA and PAI-1 expression levels in the human decidua, possibly via distinct receptor-mediated signaling pathways.

Does plasminogen activator inhibitor-1 (PAI-1) control trophoblast invasion? A study of fetal and maternal tissue in intrauterine, tubal and molar pregnancies

Urokinase plasminogen activator, its receptor and the inhibitor PAI-1 are believed to control proteolysis and remodelling of maternal tissue during trophoblast invasion. This system appears to be strictly regulated in normal intrauterine pregnancies whereas tubal and molar pregnancies seem to be characterized by an uncontrolled excessive placental invasion. This study evaluates subcellular PAI-1 by immunohistochemistry in the villous placenta, in the basal plate and placental bed, and in the decidual compartments of normal, tubal and molar pregnancies. PAI-1 was present in villous syncytiotrophoblasts and co-localized focally with fibrin-type fibrinoid on the surface of the chorionic villi. Basal plate and placental bed extravillous interstitial trophoblasts, as well as vascular trophoblasts, were also PAI-1 positive. In the decidua parietalis, PAI-1 was observed in the cytoplasm of the non-invaded decidual cells. In the decidua basalis comprising the basal plate, PAI-1 was seen to be membrane-associated or confined to the extracellular matrix (ECM) facing the invasive front of anchoring villi. The ECM of decidua capsularis and chorion laeve displayed the most pronounced PAI-1 expression towards the maternal interface. In contrast, the majority of placental bed decidual cells adjacent to the interstitial and vascular trophoblasts were PAI-1 negative. Only a few stromal cells distant from the implantation site were PAI-1 positive in the tubal pregnancies and decidualization was not present. Likewise, excessive decidual necrosis and fibrinoid deposition devoid of PAI-1 was a common finding in complete molar pregnancies. These results suggest that PAI-1 defines specific extravillous invasive trophoblasts within the maternal decidua. Moreover, maternal cellular lack of PAI-1 in tubal pregnancies and excessive decidual necrosis in molar pregnancies indicate an uncontrolled placental invasion. The present data indicate that trophoblast invasion is primarily regulated by signals from decidual cells.

Choriocarcinoma and partial hydatidiform moles

BACKGROUND

Partial hydatidiform moles (PMs) rarely require chemotherapy and have never previously been proven to transform into choriocarcinoma, the most malignant form of gestational trophoblastic disease (GTD). Consequently, some have questioned whether women with PMs need human chorionic gonadotropin (hCG) follow-up. Here, we investigate whether PMs can transform into choriocarcinomas.

METHODS

Patients with a PM who developed a subsequent choriocarcinoma were identified from our GTD database. The histology of both PM and ensuing choriocarcinoma was reviewed and flow cytometry used to verify the triploid status of the PMs. To determine whether the choriocarcinoma arose from the PM, DNA from the PM and choriocarcinoma in each patient was compared using microsatellite polymorphisms.

FINDINGS

Of the 3000 patients with PM, 15 required chemotherapy for persisting GTD. This was identified as choriocarcinoma in three cases. In one patient, the local pathologist could not differentiate between a PM or a hydropic abortion and neither central histological review nor hCG follow-up were obtained. This patient nearly died before the diagnosis of choriocarcinoma was made. Fortunately, the local pathologists correctly diagnosed PM in the two other patients who were then registered for hCG follow-up. Some months later, the hCG was rising and repeat uterine evacuation revealed choriocarcinoma. The PM was confirmed to be triploid in all three cases and genetic analysis showed that the subsequent choriocarcinomas contained identical single maternal and two paternal alleles at several independent loci.

INTERPRETATION

Our results show that PMs can transform into choriocarcinoma. All patients with suspected PM should be reviewed centrally and, if confirmed, need hCG follow-up.

Localization and significance of urokinase plasminogen activator and its receptor in placental tissue from intrauterine, ectopic and molar pregnancies

Urokinase plasminogen activator receptor (uPAR) is a membrane-anchored protein with urokinase plasminogen activator (uPA) as the ligand. This complex induces proteolysis and remodelling of maternal decidua during placental implantation. The presence of uPAR on trophoblasts is supposed to promote adhesion, migration and invasion. In cancer tissue, high levels of uPAR are correlated with a poor prognosis. This immunohistochemical study shows the localization of uPA and uPAR in a prospective design with stereological sampling of fetal and maternal tissue from normal, ectopic and hydatidiform molar (HM) pregnancies. Cytokeratin and Ki67 were used as markers for trophoblasts and proliferating cells. Membrane-bound uPAR was observed on villous non-proliferating intermediate trophoblasts (IT) within cell columns in intrauterine and ectopic pregnancies. The corresponding proliferating IT with cytological atypia sprouting from the chorionic villi in HM was uPAR-negative. uPA but not uPAR was observed in anchoring distal IT at the attachment-point to the basal plate. In the placental bed, extravillous interstitial trophoblasts were uPA-positive but uPAR-negative. The trophoblast giant cells were both uPA- and uPAR-negative. In relation to the maternal vessels, a focal distribution for uPA and uPAR was present in the endovascular and perivascular trophoblasts. The intraluminal trophoblasts overlying endothelial cells were uPAR-positive only. In maternal tissue from intrauterine and molar pregnancies, uPAR was seen in the decidual cells in a zone facing the anchoring villi and the fibrinoid lesions with embedded trophoblasts. In contrast, the stromal cells of the fallopian tube without a decidual reaction facing the implanted gestation were uPAR-negative. Non-invaded decidual, myometrial and muscular tissue of the pregnant uterus and fallopian tube was extensively positive for uPA whereas 'pseudodecidual' cells from the intrauterine evacuate in patients with an ectopic pregnancy only showed a focal and scanty reaction for uPA. When trophoblast invasion of the decidua was present, the decidual cells were uPA-negative. A semi-quantitative assessment of the receptor was estimated in villous IT within cell columns in normal and molar pregnancies but, in conclusion, quantitative evaluation of uPAR cannot be used to predict development of post-molar persistent trophoblastic disease (PTD).