Identification of causative pregnancy of gestational trophoblastic neoplasia diagnosed during pregnancy by short tandem repeat analysis.

•Gestational trophoblastic neoplasia can arise during the first trimester originating from trophoblasts of concurrent pregnancy.•Intraplacental choriocarcinoma can be developed from trophoblasts of a previous pregnancy.

Introduction

Gestational trophoblastic neoplasias (GTNs) are tumors that arise from trophoblasts such as invasive mole, choriocarcinoma, placental site trophoblastic tumor (PSTT) and epithelioid trophoblastic tumor (Lurain, 2011). Invasive mole is a pre-malignant disease which occurs in 10–20% cases of hydatidiform mole. Choriocarcinoma and PSTT are malignant tumors which arise from the trophoblasts of any kind of pregnancy, including hydatidiform mole. It has been shown that the causative pregnancy of GTNs is not necessarily the immediately antecedent pregnancy using DNA analysis (Fisher et al., 1995).

GTNs which are detected during pregnancy are very rare and most of them are intraplacental choriocarcinomas. Although the causative pregnancy of GTN during pregnancy may be a concurrent pregnancy or a previous pregnancy, as an extensive examination of the literature shows that there is only one study which demonstrates that the origin of intraplacental choriocarcinoma was a concurrent pregnancy (Kanehira et al., 2013). In this paper the patients presented with PSTT and intraplacental choriocarcinoma, which were diagnosed pathologically in the 11th and 38th gestational weeks, respectively. To identify the causative pregnancies, DNA analysis for 15 short tandem repeats (STRs) was performed using a commercially released kit.

Cases

Case 1 is a 37-year-old Japanese woman, gravida 8, para 6, who visited a local hospital because she found that she had become pregnant. Her last pregnancy ended in normal delivery at term and she had been having regular menstrual periods for two years. Pelvic ultrasonography detected a myoma (72 × 68 × 58 mm) and a normal sized fetus in a gestational sac. A hysterectomy was performed at the 11th week because the patient and her husband hoped for an artificial abortion and the myoma caused anemia. There was a 17 mm tumor in the myometrium next to the myoma macroscopically. Histological examination demonstrated that the tumor was composed of intermediate trophoblastic cells which invaded the myometrium in a sheet-like structure (Fig. 1B). The cells were positive for human placental lactogen (hPL) and weakly positive for hCG (Fig. 1C–D). Focal necrosis and extensive vascular invasion were seen in the myometrium. One mitotic figure per 10 high-power fields was observed. Diagnosis of PSTT was strongly considered. All villi and trophoblasts in anchoring villi were normal (Fig. 1A). Serum hCG and hPL levels were 101.4 mIU/ml and less than 0.07 μg/ml on the 14th postoperative day, and decreased with time after surgery. There has not been any clinical evidence of recurrence for over 18 months.

Fig. 1

Histological features of PSTT during pregnancy (A–D) and intraplacental choriocarcinoma (E–F). (A) Intermediate trophoblasts invading from anchoring villi into the myometrium were of normal appearance (H&E). (B) The tumor was composed of intermediate trophoblastic cells with atypia and myometrium invasion (H&E). Cells were positive for (C) hCG and (D) human placental lactogen in immunohistochemistry. (E) Choriocarcinoma in the placenta and (F) in the jejunum showed avillous trophoblastic proliferation and atypia (H&E). × 100 magnification, scale bar = 100 μm.

Case 2 concerns a 31-year-old Japanese woman, gravida 6, para 3, who became pregnant naturally. Her last pregnancy ended in spontaneous abortion and she had been having regular menstrual periods for a year. She had an ultrasound examination once or twice a month. Melena started at the 35th week which caused severe anemia. An emergency Cesarean section was performed at the 38th week and a tumor was found in the jejunum. Pathological examination showed that a small part of the placenta and the tumor of the jejunum consisted of malignant trophoblasts like syncytiotrophoblasts, cytotrophoblasts and intermediate trophoblasts. Necrosis and hemorrhage were found more in the jejunal tumor than in the placenta (Fig. 1E–F). These cells were positive for hCG strongly and the pathological diagnosis was made as choriocarcinoma. MRI, CT scans and colonoscopy showed that choriocarcinoma had spread to the brain, the lung, the liver, the ileum and the colon, as well as the myometrium. The patient had chemotherapy with MEA therapy (methotrexate, etoposide and actinomycin-D) and whole brain radiation. The patient achieved remission after seven cycles of chemotherapy, received four additional cycles for consolidation, and has been in remission for 18 months.

To identify the causative pregnancies of the two cases, STR analysis was performed. This study was approved by the ethics committee of Nagoya University Graduate School of Medicine. Informed consent was obtained from the patients and their partners. Genomic DNA was extracted from parental oral cells and microdissected tissue of villi and GTNs from paraffin sections. DNA was amplified with 15 STR markers and a gender-determination marker. The villous and PSTT allotypes showed a complete match in all 16 loci analyzed (S1). These results suggest that PSTT arose from the concurrent pregnancy. In case 2, the results of STR analysis in 14 loci were informative (Table 1). The results of the choriocarcinoma in the placenta and the jejunum were the same as the results of the villi in nine loci but different in five loci. These results suggest that choriocarcinoma originated not from the concurrent pregnancy, but from one of the previous pregnancies.

Table 1

STR analysis of choriocarcinoma and the normal villi.

	Maternal		Paternal		Villi		CC (P)		CC (J)		Origin
Marker	Allele 1	Allele 2	Allele 1	Allele 2	Allele 1	Allele 2	Allele 1	Allele 2	Allele 1	Allele 2
D8S1179	10	14	13	14	10	14	10	13	10	13	#
D21S11	28	30	30	30	30	30	30	30	30	30
D7S820	12	12	10	11	ND	ND	ND	ND	11	12	NE
CSF1PO	10	12	11	12	ND	ND	11*	11*	11*	11*	NE
D3S1358	15	17	14	15	15	17	15	17	15	17
TH01	6	6	6	9	6	9	6	9	9*	9*
D13S317	9	12	8	9	8	12	9	12	9	12	#
D16S539	9	11	9	13	9	9	9	11	9	11	#
D2S1338	19	23	20	24	23	24	20	23	20	23	#
D19S433	13	14.2	13	15.2	14.2	15.2	14.2	15.2	14.2	15.2
vWA	14	19	14	18	14	18	14	18	14	18
TPOX	11	11	8	11	8	11	8	11	8	11
D18S51	14	17	13	18	17	18	17*	17*	17	18
AMEL	x	x	x	y	x	x	x	x	x	x
D5S818	10	10	10	13	10	10	13*	13*	10	13	#
FGA	19	26	22	25	19	25	19	25	19	25

CC (P), choriocarcinoma in the placenta; CC (J), choriocarcinoma in the jejunum; ND, not detected; and NE, not estimated. DNA of choriocarcinoma has different alleles in five loci from villi (#). Choriocarcinoma DNA showed loss of heterozygosity in four loci (*).

Discussion

The PSTT case in this study was detected in the uterus during the first trimester with a normal fetus. An extensive literature search including MEDLINE (1984–2013) demonstrated that only three cases of PSTT were reported to be diagnosed pathologically during normal pregnancy (Table 2) (Hopkins et al., 1992; Su et al., 1999; Liszka et al., 2009). All three cases had hysterectomy or biopsy of the myometrium when Cesarean sections were performed in the third trimester. Another case of a four-month-old boy was reported who died of PSTT with metastasis to multiple organs (Monclair et al., 2002). His mother was diagnosed as PSTT a year after the delivery when a hysterectomy was performed; however, it was strongly suspected that PSTT had arisen in the uterus during pregnancy. If the PSTT case in this study had kept her pregnancy, she would have been diagnosed after bleeding or fetal distress during pregnancy. Even if she could have had a baby at term, she would have had vaginal bleeding postpartum.

Table 2

Reported cases of PSTT during pregnancy.

Case	Age	G/P	Clinical presentation	Treatment	Outcome	Reference
1	27	G2P2	Vaginal bleeding and cramping at 30 wk	TAH + BSO 3 wk pp Chemotherapy	DOD at 10 wk	Hopkins et al. (1992)
			CS and myometrial biopsy at 35 wk due to fetal distress
2	37	G3P2	Partial salpingectomy at 30 wk due to tubal bleeding	TAH + BSO + pOMT + tumor resection at CS	NED at 12 mo	Su et al. (1999)
			CS at 34 wk due to tubal malignant tumor
3	29	G0P0	CS at 39 wk due to twin pregnancy	Tumor resection at CS	NED at 30 mo	Liszka et al. (2009)
4	37	G8P6	TAH at 11 wk due to myoma and artificial abortion	TAH	NED at 18 mo	Present case

G, gravidity; P, parity; wk, week; CS, Cesarean section; TAH, total abdominal hysterectomy; BSO, bilateral salpingo-oophorectomy; pp, postpartum; DOD, dead of disease; pOMT, partial omentectomy; NED, no evidence of disease; and mo, month.

There are many case reports of intraplacental choriocarcinoma and 15–25% were found incidentally after uneventful pregnancies. There is only one reported case in the English literature whose genetic origin was examined (MEDLINE; 1946–2013; search terms: “intraplacental choriocarcinoma”, “DNA”, and “genetic”) (Kanehira et al., 2013), possibly because it has been believed that intraplacental choriocarcinoma arises naturally from trophoblasts of concurrent pregnancy. This paper is the first report to show genetically that choriocarcinoma arises during a pregnancy whose origin is a previous pregnancy. Thus, the choriocarcinoma arose in the uterus first and then metastasized to the other organs including the placenta.

Exactly which pregnancy was the origin of the intraplacental choriocarcinoma could not be confined because the DNA of the children could not be used due to ethical issues or the villous tissues of all abortions could not be used. There is also a possibility that a vanishing twin pregnancy could be the origin of the malignant trophoblasts that caused the intraplacental choriocarcinoma because other GTNs have occurred during twin pregnancies (Liszka et al., 2009). However, the patient visited an obstetrical clinic first in the fifth gestational week and this possibility may be very slight.

It is difficult to find GTNs during normal pregnancy because hCG cannot be used as a tumor marker during pregnancy. Image diagnosis, including ultrasound, CT and MRI, is not useful to diagnose GTN in the placenta. The cases in this study suggest that the possibility that GTNs arise at any time should be considered, even during pregnancy. Molecular genetic studies on the origin of GTNs can lead to a better understanding of the nature of the diseases and provide prognostic information for better management of patients.

Conflict of interest statement

The authors have no conflicts of interest to declare.