A Case Series of Five Patients With Pure or Mixed Gestational Epithelioid Trophoblastic Tumors and a Literature Review on Mixed Tumors

Ka Yu Tse, MRCOG; Keith Wan Hang Chiu, FRCR; Karen Kar Loen Chan, FRCOG; Mandy Man Yee Chu, MRCOG; Siew Fei Ngu, MRCOG; Annie Nga Yin Cheung, FRCPath; Hextan Yuen Sheung Ngan, FRCOG; Philip Pun Ching Ip, FRCPath


Am J Clin Pathol. 2018;150(4):318-332. 

In This Article


ETT is a rare trophoblastic tumor with its origins in chorion laeve-type extravillous trophoblasts. In the first case series by Shih and Kurman,[3] 66.7% of patients with ETT had preceding term pregnancies. However, in the review by Zhang et al,[12] among 45 patients whose obstetrics history was available, 51.1% had undergone abortions, 24.4% had complete hydatidiform moles, 8.9% had invasive moles, 4.4% had experienced ectopic pregnancies, and only 8.9% had achieved term pregnancies. In our series, none of the ETTs were preceded by term pregnancies but by miscarriage (40%; 2/5), abortion (20%; 1/5), ectopic pregnancy (20%; 1/5), and molar pregnancy (10%; 1/5). Unlike the usual clinical presentations reported in the literature, only one of our patients had abnormal vaginal bleeding. The median time interval from the antecedent pregnancies in our series was 31.8 months, which was similar to that reported by Zhang et al,[12] which was 36 months (range, 0–300 months). Two (40%) of five of our patients received chemotherapy prior to surgery too, which was higher than the 14.3% to 20.9% in the literature.[10,12] ETT can recur late.[6] Zhang et al[12] reported that 19 (30.6%) of 62 patients relapsed, with a median relapse/metastasis time of 8 months (mean, 21.1 months; range, 1–145 months).[12] Mitotic count in pure ETT is usually about two per 10 hpfs (range, 0–9), and the mean Ki-67 index is 17.7% ± 4.5% (range, 10%-25%).[3,16] hCG is mildly elevated, where the median is 665 IU/L.[12] It is usually less than 2,500 U/L[17] and only rarely rises to more than 10,000 IU/L.[16]

It has been difficult to identify poor prognosticators in ETT due to the small sample size and short follow-up in previous case series. Davis et al[10] showed that an interval from the preceding pregnancy of more than 4 years might be associated with a poor prognosis. Of the three of their seven patients who had an interval of 9 to 22 years from their last pregnancies, two had metastatic disease on diagnosis and had static or progressive disease after chemotherapy. The remaining patient had a hysterectomy for localized disease, but she recurred and had progressive disease despite chemotherapy. Tumor size, tumor necrosis, and cytologic atypia appeared not to have a prognostic value.[6,18] Shih and Kurman[3] stated that the Ki-67 index was not related to the clinical behavior of ETT in their early series, and the interpretation might be difficult due to the similar morphology between intermediate trophoblasts, cytotrophoblasts, decidua, and smooth muscles.[19] The assessment might also have been obscured by the surrounding immune cells, such as natural killer cells and lymphocytes, which had a high Ki-67 index too. High mitotic activity of more than five MFs/10 hpfs has been shown to correlate with poor outcomes in PSTT.[20] However, its significance is unclear in ETT.[18,19,21] For example, the patient in the case report by Coulson et al[21] whose mitotic figure was up to 30 MFs/10 hpfs remained disease free for 15 months after hysterectomy. The patient in the series by Sung et al[18] whose mitotic count was 36 MFs/10 hpfs had stage I disease and was disease-free for 8 years after hysterectomy. It is noteworthy that the assessment and interpretation of mitotic activity and proliferative index could be problematic in tumors among patients who have received preoperative chemotherapy since these parameters are expected to be lower, and comparison with cases without treatment would be difficult. p53 overexpression has been implicated in the development of GTN.[22,23] Nagai et al[24] found that p53 immunostaining was positive in six (100%) of six patients with metastatic or recurrent PSTT, while it was positive in only one (16.7%) of six patients whose PSTT was confined in the uterus (P = .015), indicating there might be a role of p53 in the disease progression of PSTT, despite a small sample size. Both their patients with ETT and all three patients in Sung et al[18] also had positive p53 immunostaining, but no poor outcomes were observed. The value of a high serum β-hCG level is also controversial.[6,25] In fact, a high β-hCG level might raise the rare possibility of coexisting CC.[1] The use of chemotherapy alone without hysterectomy might be a risk factor since definitive treatment might be delayed. Zhang et al[12] demonstrated that the use of adjuvant chemotherapy after surgery increased the risk of recurrence compared with surgery alone (P = .005). However, only stage was an independently significant risk factor for overall survival after multivariate analysis (P = .015).

In our cohort, patient D had an interval of more than 4 years from her last pregnancy, extensive tumor necrosis, and vascular invasion, and she received nine cycles of chemotherapy before her hysterectomy. However, she did not experience a recurrence. In contrast, patient A, the only patient who had a recurrence, had no obvious risk factors apart from the presence of tumor necrosis. The recurrence in the ureter was exceptional, and this could not have been detected by PET-CT or intraoperative examination. Since occult distant metastasis could not be excluded, chemotherapy was offered, but the patient declined.

The median overall survival in Zhang et al[12] was 24 months (mean, 29.3 months; range, 6–145 months), and the FIGO stage was the only significant prognosticator for survival. Because patients with GTN have lifelong follow-up in our unit, we were able to demonstrate a long overall survival after a median follow-up of 102 months.

In addition to the three cases from our center,[14] there have been only 13 other mixed trophoblastic tumors reported in the literature Table 3 and Table 4.[3,5,14,18,26–35] The median age was 34 years (range, 15–60 years), and at least six were Asian. It is not known if mixed ETT is more common in Asia, but in general, gestational trophoblastic disease, which includes benign molar pregnancy to the malignant GTN, is more common in Asia than in Western countries. More than half of the patients had abnormal vaginal bleeding. However, none of them could be diagnosed correctly on presentation, which highlighted the diagnostic difficulty. The longest diameter of the uterine tumor for those who did not have preoperative chemotherapy ranged from 2 to 8 cm. Most tumors appeared not to be very mitotically active, and the highest mitotic count was 17 MFs/10 hpfs in a patient with mixed PSTT and CC,[31] and Ki-67 positivity was up to 50% in the CC component in another patient with mixed ETT and PSTT.[35] Tumor necrosis was common. Among 14 patients whose information about their preceding pregnancies was available, the median interval from the antecedent pregnancy was about 26.2 months (range, 14 weeks to 38 years). Six (42.9%) of 14 had a history of molar pregnancies or GTN, and three (21.4%) required systematic chemotherapy.

The median overall survival for the 16 patients was at least 37.5 months (range, 1–231 months).[3,5,14,18,26–35] Four had distant metastasis. Apart from one who defaulted[3] and one who died of disease,[31] the other two patients recurred and yet were alive for at least 1 year after the diagnosis.[26,27] Pathologic risk factors could not be determined for the two patients who recurred and the one patient who died due to the incompleteness of the pathologic data. Thirteen patients underwent a hysterectomy. Four did not receive chemotherapy, among whom one defaulted and the rest survived up to 84 months without recurrence. Some have suggested that patients with mixed ETT and CC might present in a similar way as those with CC.[32] However, this is not the same in our patients. CHAMOC and EMA-CO have been effective in treating high-risk or recurrent GTN and CC,[15,36] but our patients C and D did not respond well to these regimens, and their diseases could be controlled only after hysterectomy. These two patients did not receive any adjuvant chemotherapy after hysterectomy, and they have remained disease free for more than 10 years. For patient E, only two cycles of CHAMOC were given postoperatively, and she has not experienced any recurrence for more than 7 years. These patients had a disease course that resembled ETT more than CC. Conventionally, postmolar GTN is diagnosed by hCG surveillance, and patients who fulfill the diagnostic criteria can be readily treated by chemotherapy without a histologic diagnosis.[11] Because of such a high rate of previous molar pregnancy or GTN, as stated above, it is not surprising these patients might have been treated as having recurrent GTN without histologic confirmation since otherwise this might have potentially delayed treatment. However, our findings illustrate that for those who fail to respond to chemotherapy for suspected GTN, histologic evaluation is mandatory, and pathologists specializing in GTN should review the specimen so as not to miss the rare possibility of coexisting ETT or PSTT, which is typically resistant to chemotherapy. Ramondetta et al[26] suggested that for those CCs mixed with either PSTT or ETT, sensitivity to chemotherapy might be related to the ratio of syncytiotrophoblasts and cytotrophoblasts to intermediate trophoblasts. However, the proportion of each component in mixed trophoblastic tumors was not well documented in the literature, and this practice of reporting is not universal among pathologists. It may be worthwhile for the pathologists to report the percentage of different components in mixed GTN so more data can be collected to ascertain whether clinical behavior will follow the more prominent element.

The pathogenesis of mixed ETT is poorly understood. Shih and Kurman[16] proposed that CC, PSTT, and ETT might derive from the same trophoblastic stem cells (ie, cytotrophoblasts), where CC is the most primitive tumor comprising various proportions of neoplastic cytotrophoblasts, syncytiotrophoblasts, and intermediate trophoblasts.[25,37] PSTT and ETT are more differentiated, where the neoplastic cytotrophoblasts in PSTT differentiate into intermediate trophoblastic cells in an implantation site, and those in ETT differentiate into chorionic-type intermediate trophoblastic cells in the chorion laeve. "Atypical CC," which was later believed to be ETT, had been reported after chemotherapy for proven metastatic CC.[38–40] Using Shih and Kurman's hypothesis,[16] it was speculated that chemotherapy might allow some CC cells to differentiate into ETT that were less sensitive to chemotherapy because the cells were less primitive in contrast to CC. However, there is still not enough evidence to support that mixed ETT is a result of previous chemotherapy. In Davis et al,[10] two of their patients had elevated hCG and received chemotherapy for a working diagnosis of postmolar GTN before the diagnosis of ETT could be made. They suspected that these patients had mixed tumor with a CC component, and the latter component resolved after chemotherapy, leaving only ETT that could be found on the final histology.

Successful fertility-sparing management has been widely reported in CC.[41] It is estimated that the term live birth rate was higher than 70% after treatment for persistent trophoblastic disease or CC, and there was no increased risk of fetal abnormalities. There are very few case reports on fertility-sparing treatment in PSTT, and the success rate is about 67%. One of our patients (patient B) had fertility-sparing treatment and experienced no recurrence after 89 months of follow-up. Whether this approach is safe for ETT is unknown, and there was only one other report of fertility-sparing treatment.[32] Careful counseling is needed because failed conservative management has been reported, as in the case of PSTT.[42] Until more long-term safety data on fertility-sparing treatment are available, the standard treatment of ETT or mixed ETT is still hysterectomy.