Placental site trophoblastic tumor

Abstract

Placental site trophoblastic tumor (PSTT) is a rare neoplasm that rises from intermediate trophoblasts and commonly presents with low and variable concentration of HCG immunoactivity in serum, which can be difficult to differentiate from early stage choriocarcinoma/gestational trophoblastic neoplasm (GTN) or quiescent gestational trophoblastic disease. PSTT can occur after a normal pregnancy, spontaneous abortion, termination of pregnancy, ectopic or molar pregnancy. There is a wide clinical spectrum of presentation and behavior ranging from a benign condition to an aggressive disease with fatal outcome. Nontrophoblastic malignancies such as germ cell tumors or other tumors secreting low HCG must also be considered in the differential diagnosis. Because treatments for these conditions are different, a means of differentiating PSTT from other diagnoses is important. Surgery is the cornerstone of treatment. Chemotherapeutic regimen should be EMA/CO for first line chemotherapy; EMA/EP should be used in EMA/CO refractory cases. This article reviews the literatures on this rare but fatal disease.

Introduction

Placental site trophoblastic tumor is a rare form of gestational trophoblastic neoplasia composed predominantly of intermediate trophoblast. It constitutes about 1–2% of trophoblasic tumor (Ajithkumar et al. 2003). While Marchand in 1985 used the term of atypical choriocarcinoma, today’s terminology comes from Kurman who described 12 patients with a distinct form of trophoblastic disease (Marchand 1985; Kurman et al. 1976) and regarded it as an exaggerated form of reaction of the placental bed (Larsen et al. 1991).

Following the report by Twiggs et al. (1981) of a woman who died of this disease, Young et al. (1985) coined the name PSTT to describe the malignant potential of this tumor. Since then, approximately 150 cases have been reported in the literature (Guvendag et al. 2004).

The etiology, epidemiology and risk factors for the development of PSTT are not well understood (Ajithkumar et al. 2003). Recently, it has been shown that Human chorionic gonadotropin-free B-subunit is a reliable marker for diagnosis of placental site trophoblastic tumors (Cole et al. 2006).

Tumor genesis

Genetic studies show that PSTT can originate from either a normal conceptus or from hydatiform mole.

Karyotype of PSTT is diploid, but few cases of tetraploidy have been reported (Xue et al. 2002). Genetic analysis showed an XX genomic composition in 89% of PSTT. Most of PSTT were derived from the antecedent female conceptus and were likely to have possessed a functional paternal X chromosome. Sex chromosome analysis of patients with PSTT suggests a unique genetic basis for the development of PSTT that involves the paternal X chromosome (Hui et al. 2000).

Inactivation of dysregulation of P53 gene and up regulation of epidermal growth factor receptor (EGF-R) and MIB-1 (Ki67, a proliferation-associated antigen) are suggested to have a role in the tumor genesis and propagation of PSTT (Muller-Hocker et al. 1997).

Pathology

PSTTs are also growing malignant tumors of intermediate trophoblast that form a nodule in the endometrium and myometrium (Begent 1995). The gross appearance is gray or yellow with a circumscribed border and little hemorrhage (Ajithkumar et al. 2003). PSTT is composed almost exclusively of intermediate trophoblasts that proliferate to form a tumor mass. These cells constitute a transitional form between cytotrophoblasts and syncytiotrophoblasts. Their physiologic role is to form placental bed, which is why they are also called extravillus trophoblasts (Arato et al. 2000). The intermediate trophoblasts are variable in appearance but tend to form sheets or nests of cells that invade in uterus by splaying apart individual myometrial fibers. The tumor cells can also extensively invade blood vessels, reminiscent of their role in normal implantation. The intermediate trophoblasts range from completely bland and monotonous cells to enlarged, hyper chromatic cells with marked nuclear atypical. The cytoplasm is generally ample and can be clear, amphophilic or eosinophilic. Some tumor cells may be spindled. Most tumor cells are mononuclear but occasionally they coalescence into multinucleated giant cells. The mitotic rate is variable from less than 1 to greater than 30 mitoses per 10 HPF (Moore-Maxwell and Robbery 2004). Syncytiotroohoblasts can sometimes be identified, but only sparingly (Anderson and Russell 2002).

The occasional observation of fibrinoid deposition is a typical feature of this tumor (Ishi et al. 1984). The endometrium adjacent to the neoplastic cells may show a pseudodecidual change (Fox et al. 1995). With immunohistochemical staining, 50–100% of intermediate trophoblasts show positivity for human placental lactogen (HPL) and fewer than 10% of tumor cells stain for human chorionic gonadotropin (HCG)(Kurman et al. 1984, 1991).

This can be valuable in diagnosis and in discriminating PSTT from carcinoma or sarcoma. More PSTT cells stain with specific monoclonal antibodies for b core segment and for HPL than for intact HCG (Eskstein et al. 1982).

Pregnancy-associated major basic protein (pMBP), a marker of intermediate trophoblast, is useful in distinguishing PSTT from other forms of throphoblastic tumor immunohistochemically. It presents in 78% of cases (Rhoton-Vlasak et al. 1998).

Recently, Shin and Karman have described a double staining technique using MIB-1 antibody to determine the Ki67 proliferative index in Mel-CAM (also known as CD 146 and MUC 18) defined intermediate trophoblast (Shih and Kurman 1998). The proliferation rate of PSTT as measured by Ki67 labeling is approximately 14%, which helps to distinguish it from exaggerated placental site reaction that demonstrates no proliferation.

Inhibin staining of tissue samples is useful in distinguishing PSTT from uterine sarcoma and carcinoma (Guvendag Guven et al. 2004). Immunohistochemical staining for P53 reveals intense nuclear labeling (Iwamoto et al. 2003).

Clinical features and diagnosis

PSTT shows widely variable characteristics, from a comparatively benign type confined within the uterus to a highly malignant type causing widespread metastasis.

PSTT presents with metastases in about 10% of cases and develop in an additional 10% of patients during follow up (Disaia and Creasman 2002).

Symptom may present weeks to years after the preceding pregnancy. The first symptoms of the disease are usually seen after delivery or abortion. The dominating signs of the clinical picture are vaginal bleeding (79%) and enlargement of the uterus (Fukunaga and Ushigome 1993). Vaginal bleeding or amenorrhea is the usual mode of presentation. Nephritic syndrome or hematuria appears to be related to glomerular membrane deposition of immunoglobulin and fibrin may occur in 10% of patients. It resolves after removal of the tumor (Young et al. 1985).

Virilization due to raised testosterone levels, secondary to an ovarian response to gonadotropin derived from the tumor, is also reported (Nagelberg and Rosen 1985).

PSTT is staged according to the FIGO staging of gestational trophoblastic disease. The WHO scoring system for GTD couldn’t be used for PSTT (Goldstein et al. 1998).

Sepsis, erythrocytosis and pneumothorax, anorexic syndrome may be the presenting symptoms (Bower et al. 1996).

Differential diagnosis

Differential diagnosis includes choriocarcinoma/gestational trophoblastic neoplasm (GTN) or quiescent gestational trophoblastic disease (quiescent GTD), exaggerated placental site, placental site nodule and epithelioid type of leiomyosarcoma. In PSTT, mean Ki 67 staining is 14% of total specimen area, when compared with 69% of total specimen area in choriocarcinoma. However, there can be overlap area in staining mitotic index which will be useful in differential diagnosis in that situation (Guvendag Guven et al. 2004). In choriocarcinoma, BHCG titer is high but in PSTT it is low.

Nontrophoblastic malignancies such as germ cell tumors or other tumors secreting low HCG must also be considered in the differential diagnosis. Because treatments for these conditions are different, a means of differentiating PSTT from other diagnosis is important. Laurence and associates reported that HCG-free B-subunit (cutoff values of >35 and >80%) discriminates PSTT and nontrophoblastic malignancies from other GTN possibilities, with extreme sensitivity. The statistical analysis also supports the use of HCG-free B-subunit (%) to differentiate PSTT from nontrophoblastic malignancy. Overall, hCG-free B-subunit (%) accurately differentiates PSTT from all other GTN and from nontrophoblastic malignancy option (Cole et al. 2006).

Placental site nodule is a focal, usually superficial, well-circumscribed nodule with a pauci cellular center surrounded by radiating intermediate trophoblasts enmeshed in fibrin. This residue may cause menorrhagia after delivery or abortion, and may cicatrize.

The other benign anomaly is the so-called exaggerated placental site reaction (Arato et al. 2000). By its size and its morphology, it may be similar to PSTT, can be the cause of bleeding disorders but can be treated with endometrial ablation. Ki67 labeling is negative. On the other hand, PSTT is a real tumor of the intermediate trophoblast cells.

Epithelioid trophoblastic tumor is composed of chorionic type intermediate trophoblast, behavior is similar to PSTT. It is generally benign, metastasis and death is 25 and 10%. Another differential diagnosis is epitheliod type of leiomyosarcoma.

Investigation

The variable and often low level of HCG detected in these tumors reflects the lack of syncytiotrophoblast. In bulky PSTT, the level of B-HCG is in the range of only 102–103 IU/It (Bagshawe and Rustin 1995). Hence, the level of HCG is not an accurate indication of tumor burden and cannot be used as a reliable tumor marker.

Unlike choriocarcinoma, the level of serum HCG in PSTT correlates neither with tumor burden, nor with the malignant behavior. Thus, BHCG appear to have no predictive value and disease may still progress even if levels are not raised (Felmate et al. 2001). The range of serum BHCG concentration at diagnosis in 79% is below 1,000 IU/It and in 58% lower than 500 IU/It (Papadopoulos et al. 2002).

In a study with the aim of diagnosis of persistent trophoblastic disease (PTD), we have reported, the normal B-HCG regression curve might be useful for quicker detection of PTD than the plateau or rise of level (Behtash et al. 2004a).

Imaging

Ultrasonography shows an echogenic mass that can involve endometrium and myometrium. Both solid intramural masses and cystic lesions are described. PSTT divided into two types, a hypervascular type and a relatively hypovascular type.

In the hypervascular type, massive bleeding following dilatation and curettage has been reported. When prominent vascularity of the tumor is indicated by imaging finding, dilatation and curettage should be avoided. On the other hand, uterine conservative surgery may be possible in patients with the localized hypovascular type. The main role of imaging diagnosis of PSTT is to clarify the vascularity of the tumor (Sumi et al. 1999).

CT scan may also delineate uterine mass. In MRI, PSTT presented as myometrial masses that are isodense to healthy myometrium (Sumi et al. 1999).

To perform an open uterine surgery with an aim at preservative fertility, it is important to precisely determine the size and location of the tumor and execute an effective resection procedure. In these cases, the operation was aided by combined use of MRI, PET scan, sonohystrography, high-resolution digital hysteroscopy (Tsji et al. 2002).

Clinical behavior and prognostic factors

The majority of PSTT behave in a benign fashion; approximately 10–15% are clinically malignant. It is usually confined to the uterus at the time of initial diagnosis. PSTT can metastasize and is usually resistant to chemotherapy. The most common site of metastasis is the lung, but metastasis to the scalp, brain, stomach, liver, spleen, bowel, pancreas, kidney, and adjacent pelvic organs also has been reported. The most common site of metastasis is the lung and vagina (Guvendag Guven et al. 2004).

PSTT occurring simultaneously in mother and infant have been reported (Tsji et al. 2002). Prediction of metastatic potential is difficult. Mitotic counts have not proved to be reliable because tumors with low mitotic counts can do metastasize. In addition, mitotic counts in the curettage specimen may differ from those in the hysterectomy specimen and from those in the resected metastases.

Differences between the mitotic counts of the endometrial curettage and the corresponding hysterectomy specimens were also noticed in some patients. Uterine tumor was intermixed with abundant smooth muscle bundles; whereas metastatic tumor was mostly composed of tumor cells. Although, the ratio of mitotic tumor cells to all tumor cells in these sections was similar, the smooth muscle bundle diluted the mitotic count over a fixed observation area. Hence, the mitotic count per 10 HPF in a tumor is also inversely related to the proportion of stromal cells n that tumor and does not completely reflect the proliferation activity of the tumor (Monclair et al. 2002).

How et al. found that prognosis is poor if tumor specimens show a mitotic count of more than five mitoses per 10 HPF(Chang et al. 1999).

Chang reported FIGO stage is the most important prognostic factor in PSTT. Patients with disease confined to genital tract (FIGO I and II) had 93.5% survival, which was higher than that (33.3%) of those with disease extended outside genital tract (FIGO III and IV) (P<0.0001) (Monclair et al. 2002). Extra uterine spread of the disease appears to be the most useful prognostic factor for progression.

Other reported poor prognostic factors are higher gravity and term deliveries with female fetuses (Armes 1995). Length of time from antecedent pregnancy of two or more years appears to be a major prognostic variable in PSTT (Lathrop et al. 1988).

According to multivariate analysis, the risk for unfavorable behavior of the disease increases considerably with the length of this interval (Newslands et al. 2000). Diagnosis less than 2 years from the antecedent pregnancy, and the disease localized to the uterus are associated with better outcome (Papadopoulos et al. 2002; Newslands et al. 2000).

In the report of charring cross-hospital, all seven deaths had lung metastases and presented more than 4 years since the last pregnancy (Papadopoulos et al. 2002).

In Chang’s review, it is noticed that patients with PSTT extended outside the uterus at presentation were on an average of 3 years older in age and had longer intervals between their antecedent pregnancy and the diagnosis of PSTT when compared to those whose disease were confined to the uterus. They also had a higher incidence of term delivery as their antecedent pregnancy (Monclair et al. 2002).

Vascular space involvement, endomyometrial invasion are adverse prognostic factors. The occurrence of brain metastasis is also a poor prognostic factor. PSTT previously complicated by hidatiform mole has a poor prognosis and high risk of brain metastasis. The WHO scoring system for GTD did not correlate with this outcome (Armes 1995).

Treatment

Surgery remains the cornerstone of therapy. With primary hysterectomy being the optimal therapy once the diagnosis of PSTT is established (Bower et al. 1996). The ovaries may be preserved if they appear macroscopically normal. Role of pelvic and paraaortic lymphadenectomy in early stage disease is unclear, but isolated metastasis in lymph node is reported. Consequently, surgery is the mainstay of treatment of nonmetastatic PSTT (Behtash et al. 2005). However, in locally advanced and metastatic disease, all extra tumors should be resected if technically feasible.

Only in cases in which preservation of fertility is desired, conservative surgery (D&C or local resection) could be justified (Hopkins et al. 1985). Also we reported, successful pregnancy after localized resection of perforated uterus in choriocarcinoma (Behtash et al. 2006) and Complete remission of an unusual location of metastatic gestational trophoblastic neoplasia (Behtash et al. 2004b).

In most series more than one treatment modalities have been used. Local uterine resection may be considered if a patient is resistant to chemotherapy and wants to retain fertility. When local resection is considered, ultrasound, MRI, and/or PET scan may identify the site of resistant tumor (Behtash et al. 2005). Conservative surgery recommended when the lesion is confined to the uterus, mitotic count is low, there is no uterine enlargement and close follow-up (including serial HCG estimation with periodic pelvic examination and ultrasonography)is possible(Armes 1995; Kurman et al. 1984). However, in the absence of reliable prognostic indicators, conservative therapy in the form of curettage alone should be used with caution (Ajithkumar et al. 2003).

PSTT is resistant to chemotherapeutic agents commonly used for chorionic tumor, such as Methotrexate and Actinomycin D, but combination chemotherapy, such as EP/EM or EMA/CO utilizing Etoposide, Cisplatin, Methotraxate, Actinomycin D, Cyclophosphamide and Vincristine, has been reported to be highly effective (Tsji et al. 2002; Behtash et al. 2004b).

The best choice of chemotherapy for PSTT is uncertain, clinical impression is that EMA/EP is more potent than EMA/co schedule of chemotherapy (Fisher et al. 1992). The complete response rate for metastatic PSTT managed with EMA/CO chemotherapy is only 28% (Swisher and Drescher 1998).

Chemotherapeutic regimen should be EMA/CO for first line chemotherapy. EMA/EP should be used in EMA/CO refractory cases Taxol is added to the regimen for Cisplatin-refractory cases (Taxol/Cis-Taxol/Etoposide regimen) (Lathrop et al. 1988).

The most recent data from the Charing Cross Hospital and other centers suggest the EMA/EP is the most effective treatment for metastatic or recurrent PSTT (Newlands et al. 1998). Alternative second-line treatments for recurrent or progressive PSTT are BEP (Bleomycin, Etoposide, Cisplatin) and VIP (Etoposide, Ifosfamide, Cisplatin) protocols (Ajithkumar et al. 2003).

Chemotherapy should be used in patients with advanced PSTT and may be considered in patients with FIGO stage I disease with length of time from antecedent pregnancy>2 years or high mitotic count per 19 HPF in histopathology (Hoesketra et al. 2004).

Radiation is reported to be useful in the setting localized and isolated recurrence or setting palliative therapy (Armes 1995). Radiotherapy has been used in six cases with varying success. In two reported cases, pelvic external beam radiotherapy contributed to remission of the disease. Irradiation combined with surgery and chemotherapy might be more helpful in cases with pelvic residual disease (Lathrop et al. 1988). However, use of radiation should be based on individual basis and no generalized conclusions can be drawn regarding its use.

In cases of unrespectable progressive PSTT, salvage chemotherapy is based on the initial treatment. Patients who had initial treatment with EMA/CO regimen can be salvaged with EMA-EP (Kandall et al. 2000).

Conclusion

PSTT is a well recognized, but uncommon form of gestational trophoblastic disease with a highly variable clinical behavior. Most of the tumors in the uterus are well circumscribed and localized. Hysterectomy seems to be sufficient treatment of patients whose disease is limited to the uterus. Patients with extensive or metastatic disease are candidates for cytoreductive surgery and chemotherapy, although the clinical outcome is largely unpredictable.