Pure nongestational ovarian choriocarcinoma: A scoping review

Choriocarcinoma of the ovary is a rare and highly malignant germ cell tumor. There are three ways in which an ovarian choriocarcinoma can arise: as a primary gestational choriocarcinoma that results from an ectopic ovarian pregnancy, as a metastatic choriocarcinoma that arises from a gestational choriocarcinoma, and as a germ cell tumor with differentiation into trophoblastic structures. Ovarian choriocarcinomas are therefore classified as gestational or nongestational. Recently, DNA polymorphism analysis has allowed investigators to determine the etiology of choriocarcinoma (gestational versus nongestational). Herein, in this scoping review, we detail the classification of, and clinical aspects of, pure ovarian choriocarcinoma. Introduction Choriocarcinoma is a very rare and highly malignant germ cell tumor that accounts for <1% of all malignant germ cell tumors [1]. There are three ways in which an ovarian choriocarcinoma can arise: as a primary gestational choriocarcinoma that results from an ectopic ovarian pregnancy, as a metastatic choriocarcinoma that arises from a gestational choriocarcinoma from another primary site in the female genital tract, and as a germ cell tumor with differentiation into trophoblastic structures [2]. Choriocarcinoma is classified as gestational or nongestational. Gestational choriocarcinoma is a form of gestational trophoblastic disease, which arises from a partial mole, a complete mole, or a normal pregnancy. Gestational choriocarcinoma is estimated to occur in about 2 to 7 pregnancies per 100,000 in the United States [3]. Nongestational choriocarcinoma does not arise from a pregnancy event and is an extremely rare occurrence. The incidence of primary ovarian nongestational choriocarcinoma is estimated to be 1 in 369,000,000 [4]. Differentiating gestational and nongestational choriocarcinoma can be difficult, as their clinical presentation and pathology can be identical. Traditionally, a definitive diagnosis of nongestational choriocarcinoma has been restricted to prepubescent females in whom the possibility of a pregnancy event can be eliminated with certainty. Choriocarcinoma of the ovary can be pure choriocarcinoma or, more commonly, mixed with other germ cell components [1]. When examined histologically, if other germ cell components are present mixed with choriocarcinoma, a diagnosis of a nongestational origin can be made. However, if no other germ cell component is present, differentiating a nongestational from gestational origin is impossible using histologic means alone. Recently, however, DNA polymorphism analysis has allowed investigators to determine the etiology of choriocarcinoma via analysis of the patient’s DNA, her partner’s DNA, and the DNA of the choriocarcinoma. By examining the DNA composition of the tumor and comparing it to the maternal and paternal DNA, the etiology of the Correspondence to: Joshua P. Kesterson, Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Penn State Hershey Medical Center, 500 University Drive, Hershey, PA 17036, USA, Tel: (717) 531-8144, Fax: (717) 531-0007, E-mail: jkesterson@hmc.psu.edu Received: February 03, 2017; Accepted: February 26, 2017; Published: February 28, 2017 Garrett A (2017) Pure nongestational ovarian choriocarcinoma: A scoping review Integr Cancer Sci Therap, 2017 doi: 10.15761/ICST.1000229 Volume 4(1): 2-8 search string: ((("choriocarcinoma"[MeSH Major Topic] OR choriocarcinoma [Title/Abstract]) AND (nongestational[Title/ Abstract] OR non-gestational[Title/Abstract])) AND ((ovary[Title/ Abstract] OR ovarian[Title/Abstract]) OR gonad[Title/Abstract])) AND english[Language] which resulted in 66 titles. Additionally, we searched Web of Science using All Databases with the following search string: (ts=(choriocarcinoma AND (nongestational OR nongestational)) OR ti=(choriocarcinoma AND (nongestational OR nongestational))) AND LANGUAGE: (English), which yielded 57 titles for a sum total of 123 titles. The final search in PubMed and Web of Science was performed on January 6, 2016. Duplicates (n=43) were excluded. We included case reports and retrospective chart reviews. Review articles and book chapters were excluded (n=7). Review articles and those articles included in the analysis were cross-referenced to identify additional studies missed in initial database search (n=6). Included articles were limited to those dealing with pure non-gestational choriocarcinoma of the ovary. Criteria used to define a non-gestational origin of the tumor were cases occurring in prepubertal girls or a cases establishing a diagnosis of non-gestational origin via DNA analysis. All other cases were excluded from the results and discussion. Some literature described young G0 females without mention of their history of sexual activity. In these cases, if the patient was <20 years of age, these cases were included (unless nongestational origin was confirmed upon DNA analysis). Because Jacobs et al published an exhaustive literature review on pure choriocarcinoma of the ovary of cases occurring prior to 1981; we included only studies published after 1980. Those articles not meeting this criteria (n=49) were excluded, see Table 4. Additionally, nine articles were inaccessible to us despite multiple requests for full-text articles, and so these were excluded [6-14]. The total number of articles included in this review is 21 (Figure 1). Results Classification Since 1980, a total of 22 possible cases of pure nongestational ovarian choriocarcinoma have been published in the English language. Of these, nine occurred in premenarchal females [12-18], 11 occurred in postmenarchal females but were confirmed by DNA analysis [18,1925], and eight were considered possible cases of pure NGOC according to the strict criteria described above in the methods section [18,26-32] (Tables 1-3).

choriocarcinoma can be determined. If the DNA of the tumor exactly matches the maternal DNA, the tumor is of non-gestational origin. However, if the tumor contains any alleles matching paternal DNA, the tumor is of gestational origin [5].
In 1982, Jacobs et al. published a comprehensive literature review on pure ovarian choriocarcinoma [6]. The study classified all published cases of pure ovarian choriocarcinoma at the time into three categories: gestational ovarian choriocarcinoma, pure nongestational ovarian choriocarcinoma, and choriocarcinoma of uncertain etiology. The pure nongestational ovarian choriocarcinoma category was assigned to all pure ovarian choriocarcinoma cases occurring in prepubertal females, and the uncertain etiology was assigned to all pure ovarian choriocarcinoma cases in postpubertal women who were said to be sexually abstinent or virginal. Since the publication of this review, several other literature reviews on this topic have been published; however, limitations exist, including lack of definitive categorization strategies. As a result, little is known about the incidence, clinical course, most effective treatment regimen, as well as outcomes for pure ovarian nongestational choriocarcinoma. The purpose of this scoping review is to provide a strict definition of pure ovarian choriocarcinoma that allows appropriate classification of nongestational and gestational origin in order to better understand this disease process.

Introduction
Choriocarcinoma is a very rare and highly malignant germ cell tumor that accounts for <1% of all malignant germ cell tumors [1]. There are three ways in which an ovarian choriocarcinoma can arise: as a primary gestational choriocarcinoma that results from an ectopic ovarian pregnancy, as a metastatic choriocarcinoma that arises from a gestational choriocarcinoma from another primary site in the female genital tract, and as a germ cell tumor with differentiation into trophoblastic structures [2].
Choriocarcinoma is classified as gestational or nongestational. Gestational choriocarcinoma is a form of gestational trophoblastic disease, which arises from a partial mole, a complete mole, or a normal pregnancy. Gestational choriocarcinoma is estimated to occur in about 2 to 7 pregnancies per 100,000 in the United States [3]. Nongestational choriocarcinoma does not arise from a pregnancy event and is an extremely rare occurrence. The incidence of primary ovarian nongestational choriocarcinoma is estimated to be 1 in 369,000,000 [4]. Differentiating gestational and nongestational choriocarcinoma can be difficult, as their clinical presentation and pathology can be identical. Traditionally, a definitive diagnosis of nongestational choriocarcinoma has been restricted to prepubescent females in whom the possibility of a pregnancy event can be eliminated with certainty.
Choriocarcinoma of the ovary can be pure choriocarcinoma or, more commonly, mixed with other germ cell components [1]. When examined histologically, if other germ cell components are present mixed with choriocarcinoma, a diagnosis of a nongestational origin can be made. However, if no other germ cell component is present, differentiating a nongestational from gestational origin is impossible using histologic means alone.  [Language] which resulted in 66 titles. Additionally, we searched Web of Science using All Databases with the following search string: (ts=(choriocarcinoma AND (nongestational OR nongestational)) OR ti=(choriocarcinoma AND (nongestational OR nongestational))) AND LANGUAGE: (English), which yielded 57 titles for a sum total of 123 titles. The final search in PubMed and Web of Science was performed on January 6, 2016. Duplicates (n=43) were excluded.
We included case reports and retrospective chart reviews. Review articles and book chapters were excluded (n=7). Review articles and those articles included in the analysis were cross-referenced to identify additional studies missed in initial database search (n=6). Included articles were limited to those dealing with pure non-gestational choriocarcinoma of the ovary.
Criteria used to define a non-gestational origin of the tumor were cases occurring in prepubertal girls or a cases establishing a diagnosis of non-gestational origin via DNA analysis. All other cases were excluded from the results and discussion. Some literature described young G0 females without mention of their history of sexual activity. In these cases, if the patient was <20 years of age, these cases were included (unless nongestational origin was confirmed upon DNA analysis). Because Jacobs et al published an exhaustive literature review on pure choriocarcinoma of the ovary of cases occurring prior to 1981; we included only studies published after 1980. Those articles not meeting this criteria (n=49) were excluded, see Table 4. Additionally, nine articles were inaccessible to us despite multiple requests for full-text articles, and so these were excluded [6][7][8][9][10][11][12][13][14]. The total number of articles included in this review is 21 ( Figure 1).

Age
In the premenarchal group, the average age at diagnosis was 13.6 years. One study did not include the age of the patient at presentation [20]. Another case occurred in a 39 year-old female with a history of gonadal dysgenesis and primary amenorrhea [19].
In the DNA confirmed group, the average age at diagnosis was 24.5 years. One study that reported on three of these cases did not include the age at diagnosis [21].
In the possible cases group, the average age at diagnosis was 14.0 years.

FIGO Stage
In the premenarchal group, the FIGO stage of eight of the cases did not state the FIGO stage. For the one case that did, the FIGO stage was reported as IC [16].
For the DNA confirmed group, the FIGO stage of five cases was not reported. The remaining cases had FIGO stages of IA [26], IIIC [23], and IV [25].
For the possible cases group, the FIGO stage of five cases was not reported. The remaining cases had FIGO stages of IA [31], I [32], and III [35].

Surgery performed
In the premenarchal group (n=9), surgical therapy was reported for five of the cases but not stated for four cases. Of the five cases which reported surgical therapy, two had a unilateral oophorectomy performed [15], one had a unilateral salpingo-oopherectomy performed [17], one had a unilateral salpingo-oopherectomy and partial omentectomy performed [16], and one had a bilateral salpingooopherectomy, hysterectomy, omentectomy, and thoracoscopy and wedge resection for pleural lesions [19].
For the DNA confirmed group (n=11), surgical therapy was reported for all but five of the cases. Of the 11 cases which reported surgical therapy, one had a total abdominal hysterectomy, bilateral salpingo-oopherecotmy, and omentectomy performed two months after completion of chemotherapy [22]; one had an initial removal of the ovarian mass and two rounds of chemotherapy which were followed by followed by a total abdominal hysterectomy, bilateral salpingooopherecotmy, omental resection, pelvic lymph node dissection, and appendectomy [23]; one had a total abdominal hysterectomy, bilateral salpingo-oopherecotmy, omentectomy, and pelvic lymph node dissection performed [24]; two had a left salpingo-oopherectomy and partial omentectomy performed [25,26]; and one had a right salpingooopherectomy performed [27].

Chemotherapy
In the premenarchal group, one case did not state whether chemotherapy was used [21]. Two cases were treated with surgical therapy alone [15]. Two cases were treated with PVB (cisplatin, bleomycin, and vinblastine) therapy [16,18]. One case was treated with vincristine, methotrexate, leukovorin, bleomycin, Adriamycin, and cyclophosphamide [14]. One cases was treated with multiple rounds of various chemotherapeutic agents, see Table 1 [19]. One case was treated with "radiotherapy and chemotherapy," but the study did not describe any further details [20]. One case was treated with "three drug chemotherapy," but the study did not expound upon what these three drugs were [20].
For the DNA group, five cases did not include information of chemotherapy treatment. The BEP chemotherapy regimen was used in two cases for four and five cycles, respectively [22] and [23]. One patient was treated with the MAC regimen for four cycles [16]. One patient was treated with the EMA regimen for four cycles [27]; and one patient was treated with one course of EMA followed by 7 cycles of just etoposide and actinomycin due to methotrexate-toxicity [26].
For the possible cases group, chemotherapy treatment was not reported in two cases. Chemotherapy was not given in one case [     The BEP regiment was used in two cases for three and six cycles, respectively [30,31]. The BEP regimen was also used on another patient for four cycles, followed by high-dose CTx with carboplatin, etoposide, ifosphamide with subsequent bone marrow transplantation [35]. Finally, two patients were treated methotrexate, actinomycin, and an alkylating agent (MAC) regimen for 4 cycles [32] and 5 cycles [34].

Survival
In the premenarchal group, survival outcomes were not reported for three cases [16,17,21]. One patient was reported to be alive and well at ten-year follow-up [15], one patient had no evidence of disease at six-month follow-up [18], one had no evidence of disease at 17-month follow-up [19], and one had no evidence of disease at one year followup [20]. One patient died from immediate postoperative complications [15], and one patient was reported dead of disease [20].  [35], and 62 months [30]. In one case, the patient arrested during initial imaging and all resuscitation efforts were unsuccessful [33] ( Tables 1-3).

Discussion
This study highlights several features of pure nongestational ovarian choriocarcinoma. First, this study emphasizes the rarity of pure nongestational ovarian choriocarcinoma. Nine cases of premenarchal pure nongestational ovarian choriocarcinoma and eight cases of possible pure nongestational choriocarcinoma have been reported since 1980. Just twelve cases of pure ovarian nongestational ovarian choriocarcinoma have been confirmed by DNA polymorphism analysis. Because of the rarity of this disease process, stricter diagnostic criteria should be used in order to correctly categorize nongestational origin from gestational origin, as unless confirmed by DNA analysis or the disease occurs in a patient who is premenarchal, one cannot with absolute certainty classify an ovarian choriocarcinoma. Recently, DNA analysis has been used to successfully determine nongestational versus gestational origin of ovarian choriocarcinoma. This technology will allow for appropriate classification of this disease process which will ultimately lead to improved therapeutic strategies as more information is learned about pure nongestational ovarian choriocarcinoma.
Secondary to the rarity of this disease process, no standard therapy has been established. Treatment is often extrapolated from treatment strategies for gestational choriocarcinoma and germ cell tumors, thereby leading to significant heterogeneity in treatment strategies for pure nongestional ovarian choriocarcinoma. In reviewing the clinical outcomes of those with pure ovarian suspected choriocarcinoma (confirmed and suspected), it is difficult to make definitive treatment recommendations secondary to heterogeneity in, and inconsistent reported of, relevant clinical factors, including disease classification, patient age, stage, surgery, adjuvant therapy and outcomes, combined with the rarity of this particular entity. For those reported cases in which outcomes were reported (n=16), at a follow-up ranging from one month to ten years, 12 were reported as NED, a majority (n=11) were treated with adjuvant chemotherapy (Tables 1-3). Commonly used adjuvant combinational treatment regimens include BEP and EMA.
Herein, in this scoping review, we have detailed the classification of, and clinical aspects of, pure ovarian choriocarcinoma. Secondary to its rarity and variability in reporting, conclusive recommendations regarding ideal therapy is lacking. Going forward, definitive categorization, via DNA polymorphism analysis, and creation of an international tumor registry is warranted for rare disease entities such as pure ovarian choriocarcinoma, in order to facilitate better comprehension of its etiology and standardization of therapy with optimization of outcomes.