Persistent gestational trophoblastic disease (GTD) includes recurrence of a hydatidiform mole as well as gestational trophoblastic neoplasms, with choriocarcinoma being the most notable of these. Persistent GTD is estimated to occur in 1 in 40,000 pregnancies. Approximately half of cases of persistent GTD arise from a previous molar pregnancy, and the risk is higher for those with a complete hydatidiform mole (15%–20%) compared with a partial hydatidiform mole (1%–5%) 1. Recurrence is uncommon: 3% of patients with persistent GTD recur, and most occur within the first year of surveillance 2. Still, recurrence has been reported many years after initial diagnosis—in 1 case even 11 yr later 3. This makes early confirmation of a normal gestation important in all pregnant women with a history of GTD. Studies have demonstrated the value of DNA genotyping for correctly diagnosing and subclassifying molar pregnancies and determining the relationship between metachronous GTD events 4–13.
We describe a case of genetically related choriocarcinoma developing 5 yr after an invasive complete hydatidiform mole and simulating a cornual ectopic pregnancy. This case was included in a recent series reporting genotyping of choriocarcinomas 14.
A 25-yr-old woman, gravida 2 para 0010 was sent to the Emergency Department from clinic with an abnormally rising human chorionic gonadotropin (hCG) level and concern for ectopic pregnancy. Two weeks before her presentation, her hCG level had been 2989 mIU/mL. Weekly hCG values over the next 2 wk were 3398 and 2961 mIU/mL. A pelvic ultrasound revealed a 3.4×3.4×3.8 cm echogenic hyper-vascular mass in the right lateral uterine fundus and an empty endometrial cavity (Fig. 1A). Magnetic resonance imaging confirmed no definite connection with the endometrium (Fig. 1B). Her hCG was 2547 mIU/mL 2 d later. Her past medical history was remarkable for a complete hydatidiform mole that had been diagnosed 5 yr earlier by dilation and curettage, confirmed as such by loss of p57 expression per immunohistochemical analysis (Figs. 2A, B) and genotyping demonstrating a purely androgenetic conception. The diagnostic dilation and curettage procedure had been complicated by uterine perforation. One month into surveillance, she developed vaginal bleeding and a rise in her hCG from 7519 to 55,092 mIU/mL. A repeat dilation and curettage demonstrated persistent complete hydatidiform mole with detached fragments of molar-type trophoblast but no evidence of choriocarcinoma (Figs. 2C, D). Additional imaging revealed a right lower lobe lung mass. Wedge resection of the lung lesion demonstrated molar villi consistent with invasive/metastatic complete hydatidiform mole (Fig. 2E). She was treated with weekly methotrexate; however, her hCG levels were noted to rise on this regimen and she was switched to biweekly actinomycin D. Her hCG was undetectable after 2 mo of treatment and she was followed monthly for 2 yr. There was no concern for recurrence until she presented with the cornual mass.
The patient underwent a mini-laparotomy, wedge resection of the right cornual mass and right salpingectomy to address the differential diagnosis of a cornual ectopic pregnancy versus GTD (either new or persistent). Intraoperative findings were notable for a 5 cm necrotic cornual mass, with normal fallopian tubes and ovaries bilaterally. Intraoperative frozen section analysis demonstrated a high-grade epithelioid neoplasm. Final pathologic evaluation was consistent with choriocarcinoma. The neoplasm was comprised predominantly of relatively uniform but cytologically malignant mononucleated trophoblastic cells with interspersed individual and clustered intimately associated multinucleated syncytiotrophoblastic cells (Figs. 2F, G). Immunohistochemical analysis demonstrated hCG expression in the tumor, with strong expression highlighting the syncytiotrophoblastic cells (Fig. 2H). Given the long interval since the prior hydatidiform mole, DNA genotyping was undertaken to determine the relationship between the original complete hydatidiform mole and the choriocarcinoma. Genotyping was performed using unstained 10-μm formalin-fixed paraffin-embedded tissue sections prepared using PCR precautions. PCR amplification of 15 short tandem repeat loci from 13 different chromosomes (chromosomes 2, 3, 4, 5, 7, 8, 11, 12, 13, 16, 18, 19, 21) and the amelogenin locus (for XY determination) was performed, with thermal cycling conditions and capillary electrophoresis carried out according to the manufacturer’s instructions (AmpFlSTR Identifiler kit; Applied Biosystems; Foster City, CA). Details of p57 immunohistochemical analysis and DNA genotyping are provided in previous publications 5,15,16. Genotyping demonstrated that the prior complete hydatidiform mole and the subsequent choriocarcinoma were purely androgenetic and identical, with multiple loci displaying obligate paternal alleles not present in the maternal sample (Fig. 3).
Her hCG level on postoperative day 1 was 29. Her World Health Organization (WHO) score was 11, and she was started on etoposide, methotrexate, actinomycin D, cyclophosphamide, vincristine (EMA-CO) therapy. After 1 cycle, her hCG was undetectable and she completed 3 subsequent cycles. She is now 24 mo into surveillance with no evidence of disease.
This case highlights several important aspects of the clinical evaluation, pathologic diagnosis, and follow-up of GTD. These include: (1) the potential risk associated with uterine perforation, (2) the need for a pathologic tissue diagnosis in certain settings, (3) the role of ancillary testing (DNA genotyping) in establishing the relationship between metachronous GTD lesions and informing the WHO risk determination, and (4) maintaining a high index of suspicion for persistent GTD despite a long period of remission.
There are limited data on any risk of recurrent GTD associated with diagnostic and therapeutic dilation and curettage procedures. Two case reports document intramural and cornual recurrences of complete hydatidiform moles at short time intervals (2–3 mo) following dilation and curettage procedures 17,18. However, we are not aware of any prior reports of delayed recurrence as a more severe form of persistent GTD in the setting of uterine perforation, as occurred in our case. Given that the choriocarcinoma was proven by DNA genotyping to be genetically related to the prior complete hydatidiform mole, we speculate that some residual molar trophoblastic tissue must have been implanted in the cornual region at the time of uterine perforation and remained dormant/quiescent until progression 5 yr later. While some hydatidiform moles are invasive and can recur/metastasize even in the absence of uterine perforation, clinicians should take great care to avoid uterine perforation at the time of dilation and curettage, especially in cases of suspected or known molar pregnancy, to avoid creating a nidus for persistent GTD.
In certain clinical situations, there may be a need to surgically excise tissue in the setting of an abnormally elevated hCG for further diagnosis when GTD is suspected but cannot be confirmed. While choriocarcinoma typically has high hCG levels (>100,000), this is not uniformly the case 2,14. In the present case, the patient’s hCG level was not markedly elevated (<3000 on most measurements) and imaging could not reliably delineate between an involuting cornual ectopic pregnancy versus persistent GTD. Pathologic evaluation of resected lesional tissue established a definitive diagnosis of choriocarcinoma and excluded a cornual ectopic pregnancy. Despite the predominance of mononucleate trophoblast in the tumor, the identification of intimately associated syncytiotrophoblastic cells and diffuse hCG expression confirmed a diagnosis of choriocarcinoma. We speculate that the lower than usual hCG levels were a reflection of the tumor morphology, which was characterized by a much greater number of mononucleated trophoblastic cells and relatively limited numbers of syncytiotrophoblastic cells. The provision of tumor tissue also enabled us to perform comparative DNA genotyping to determine the relationship between the choriocarcinoma and the prior complete hydatidiform mole.
DNA genotyping of metachronous GTD lesions played a crucial role in the risk assessment and treatment planning in this case. While the successful response of the initial complete hydatidiform mole and the long time interval might have favored the development of a new independent GTD, delayed development of choriocarcinoma related to a prior molar pregnancy has been reported 6–9. To determine proper therapy for GTD, patients are grouped into either low-risk or high-risk subgroups using the WHO Prognostic Scoring System 2. For patients with gestational choriocarcinoma, some of the factors used to determine the WHO prognostic score are the type of antecedent pregnancy, the time interval from the index pregnancy, and history of previously failed chemotherapy 19. Gestational choriocarcinoma related to a molar pregnancy is lower risk than that related to a nonmolar abortion or a term pregnancy and a shorter time interval since the index pregnancy is a more favorable factor. It is often assumed that the immediately antecedent pregnancy is the causative pregnancy of a choriocarcinoma that is believed to be gestational. However, many spontaneous abortions go unrecognized and studies have shown that the immediately antecedent or concurrent pregnancy is not always the causative pregnancy 6–12.
At initial diagnosis with a complete hydatidiform mole, the patient’s WHO score was 2, which qualified her for low-risk therapy. When she was diagnosed with choriocarcinoma 5 yr later, DNA genotyping of the specimens enabled us to determine that the antecedent pregnancy was the prior complete hydatidiform mole and thus, that the time interval was 5 yr rather than a more recent unrecognized nonmolar pregnancy. Consequently her WHO score was >7, warranting high-risk therapy with EMA-CO. Other studies have demonstrated the value of DNA analysis for establishing the relationship between GTD lesions 10,12,13.
Surveillance for patients with GTD is recommended for at least 1 yr after hCG normalization and completion of chemotherapy because of the 3% risk of recurrence. Beyond the first 12 mo of surveillance, the risk of recurrence is exceedingly low 2. However, there have been several reports in the literature of late persistent GTD 3,14. These examples and the current case demonstrate that clinicians should maintain a high index of suspicion for recurrence in patients with a history of GTD who present with an atypical pregnancy regardless of the time interval from initial diagnosis. Patients with a prior history should be followed closely by an obstetrician/gynecologist until a viable intrauterine pregnancy is confirmed. The placenta of the viable pregnancy should be sent to pathology upon delivery to confirm that there is no abnormal tissue.
In summary, we report a case of genetically related persistent GTD—specifically, choriocarcinoma confined to a cornual mass without endometrial cavity involvement—5 yr after initial diagnosis of an invasive complete hydatidiform mole. This report supports the importance of avoiding uterine perforation during dilation and curettage for diagnosis and treatment of GTD. It also highlights the importance of high suspicion of recurrence when patients with a history of GTD present with an atypical pregnancy, and also the benefit of molecular genotyping to differentiate new primary versus persistent GTD in order to better inform prognosis and guide therapy.
1. Ngan S, Seckl MJ. Gestational trophblastic neoplasia management: an update. Curr Opin Oncol 2007;19:486–91.
2. Lurain JR. Gestational trophoblastic disease
II: classification and management of gestational trophoblastic neoplasia. Am J Obst Gynecol 2011;204:11–7.
3. Couder F, Massardier J, You B, et al. Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone. Am J Obstet Gynecol 2016;215:81–7.
4. Hui P, Buza N, Murphy KM, et al. Hydatidiform moles: genetic basis and precision diagnosis. Annu Rev Pathol 2017;12:449–85.
5. McConnell TG, Murphy KM, Hafez M, et al. Diagnosis and subclassification of hydatidiform moles utilizing p57 immunohistochemistry and molecular genotyping
: validation and prospective analysis in routine and consultation practice settings with development of an algorithmic approach. Am J Surg Pathol 2009;33:805–17.
6. Fisher RA, Newlands ES, Jeffreys AJ, et al. Gestational and nongestational trophoblastic tumors distinguished by DNA analysis. Cancer 1992;69:839–45.
7. Suzuki T, Goto S, Nawa A, et al. Identification of the pregnancy responsible for gestational trophoblastic disease
by DNA analysis. Obstet Gynecol 1993;82:629–34.
8. Fisher RA, Soteriou BA, Meredith L, et al. Previous hydatidiform mole
identified as the causative pregnancy of choriocarcinoma
following birth of normal twins. Int J Gynecol Cancer 1995;5:64–70.
9. Kaneta Y, Yoshiyama R, Inagaki N, et al. Gestational choriocarcinoma
whose responsible pregnancy was a complete hydatidiform mole
identified by PCR analysis with new sequence tagged site primers. Jpn J Clin Oncol 1999;29:504–8.
10. Aranake-Chrisinger J, Huettner PC, Hagemann AR, et al. Use of short tandem repeat analysis in unusual presentations of trophoblastic tumors and their mimics. Hum Pathol 2016;52:92–100.
11. Zhao J, Xiang Y, Wan XR, et al. Molecular genetic analyses of choriocarcinoma
. Placenta 2009;30:816–20.
12. Yamamoto E, Niimi K, Shinjo K, et al. Identification of causative pregnancy of gestational trophoblastic neoplasia diagnosed during pregnancy by short tandem repeat analysis. Gynecol Oncol Case Rep 2014;9:3–6.
13. Fisher RA, Savage PM, MacDermott C, et al. The impact of molecular genetic diagnosis on the management of women with hCG-producing malignancies. Gynecol Oncol 2007;107:413–9.
14. Savage J, Adams E, Veras E, et al. Choriocarcinoma
in women: analysis of a case series with genotyping. Am J Surg Pathol 2017;41:1593–606.
15. Murphy KM, McConnell TG, Hafez MJ, et al. Molecular genotyping
of hydatidiform moles: analytic validation of a multiplex short tandem repeat (STR) assay. J Molec Diagn 2009;11:598–605.
16. Banet N, DeScipio C, Murphy KM, et al. Characteristics of hydatidiform moles: analysis of a prospective series with p57 immunohistochemistry and molecular genotyping
. Mod Pathol 2014;27:238–54.
17. Nair K, Al-Khawari H. Invasive mole of the uterus—a rare case diagnosed by ultrasound: a case report. Med Ultrason 2014;16:175–8.
18. Chen PL, Jhuang JY, Lin HH, et al. Successful treatment of gestational trophoblastic neoplasia in the uterine cornus with laparoscopic cornuostomy and postoperative methotrexate injection. Taiwan J Obstet Gynecol 2017;56:261–3.
19. Ngan HY, Seckl MJ, Berkowitz RS, et al. Update on the diagnosis and management of gestational trophoblastic disease
. Int J Gynaecol Obstet 2015;131(suppl 2):S123–6.