This study was approved by the Institutional Review Board of Peking Union Medical College Hospital (approval number: S-K2102). Since the study was retrospective and fully anonymous, the requirement for written informed consent was waived.

We searched the electronic database and reviewed female patients with primary extragonadal malignant germ cell tumors (EMGCTs) who were treated in our hospital between 1983 and 2021. Patients with primary EMGCTs were identified during this period, in whom patients with EMGCTs outside the female genital tract were excluded, and patients with EMGCT in the female genital tract were included in the present study.

Clinical and demographic data, including age at onset, presenting complaints, clinical features, radiologic findings, tumor markers, pathology, treatment modality, and prognosis, were collected from the medical records and via telephone. Histological slides were reviewed by two independent gynecological pathologists from the Department of Pathology at our hospital. Information on survival outcome, menstruation, and fertility was collected on March 23, 2022, using both medical records and telephone follow-up.

Complete remission (CR) was defined by either a normal serum alpha-fetoprotein level (defined in our hospital <20 U/ml) or a negative pathology result. The patients achieving complete remission were under close follow-up at the outpatient department. A routine medical examination, including physical examination, serum alpha-fetoprotein level monitoring, and pelvic and abdominal ultrasound, was performed at each follow-up visit. The patients were followed up according to the following plan: monthly during the first year, every 2 months during the second year, every 3 months during the third and fourth years, and twice a year during the fifth year and thereafter.

Descriptive statistics was used to describe the patients’ characteristics. Continuous variables are presented as mean ± SD or median (range) and categorical data are presented as number (percentage). Statistical analysis was performed using SPSS 25.0 statistical software (SPSS Inc., Chicago, IL, USA).

A total of 209 patients with primary extragonadal malignant germ cell tumors were identified, 177 patients with such tumors outside the female genital tract were excluded and 32 (15.3%) with tumors in the female genital tract were included. The clinical features of these 32 patients are presented in Table 1. All of 32 patients were histopathologically diagnosed with YST. Except for one 29-year-old patient with a vaginal YST, the remaining 29 patients with vaginal YSTs were all under 4 years. The patient ages at diagnosis of endometrial and cervical YST were 2 and 31 years, respectively. Among 31 (31/32, 96.88%) patients with a complaint of AVB or mucosanguineous discharge, 11 patients had necrotic tumor tissue protruding from the vagina, and one patient had difficulty passing urine concurrently. Preoperative serum alpha-fetoprotein (AFP) levels were markedly elevated in all patients (range, 251–54,000 ng/ml). Except for one patient whose imaging studies were unavailable, the other 31 patients underwent at least one type of pelvic imaging on initial presentation, including pelvic ultrasound, computed tomography, magnetic resonance imaging, and/or positron emission tomography–computed tomography. Twenty patients (64.5%) underwent general imaging evaluation (chest-abdomen-pelvis). Only two patients had radiologic metastases at diagnosis: one patient had minor nodules in the lower lobe of the lungs on x-ray, and another patient had multiple tumor nodules located on the peritoneal surface on computed tomography.

Tissue biopsy was performed in all 32 cases, and all cases were pure YST (100%). Figure 1 shows the morphological features of YST in one case. Pathological specimens were obtained from exfoliated tissues in 10 patients (31.3%), tumor biopsy in 19 patients (59.4%), and tumor resection in 3 patients (9.7%). Because the anatomical structures of pediatric patients are generally small, the vaginoscopic approach for hysteroscopy and pediatric rhinoscopy were used to inspect and obtain biopsies through the magnification of endoscopic images. Regarding tumor biopsy, an ultrasound-guided biopsy of the pelvic mass was performed in the case of endometrial YST; multiple cervical biopsies were performed by colposcopy in the case of cervical YST; the biopsy specimens of vaginal YSTs were obtained by vaginohysteroscopy in nine cases, biopsy forceps in five cases, pediatric rhinoscopy in two cases, and a combination of vaginohysteroscopy and pediatric rhinoscopy in one case (as shown in Table 2 and Figure 2). Immunohistochemical staining served as a helpful adjunctive tool for the diagnosis of YST in 22 of the 32 cases. Table 2 and Figure 2 shows the immunohistochemical results of the tested cases.

Table 3 shows detailed therapeutic information for each case. Nine patients (28.1%) were initially admitted to local hospitals and were subsequently referred to our hospital. Twenty-three patients (71.9%) were exclusively diagnosed with and treated in our hospital. Two patients with vaginal YSTs underwent surgery at local hospitals prior to chemotherapy. One patient (No. 24) underwent radical surgery, including vaginal tumorectomy, pelvic lymphadenectomy, and internal iliac venous embolectomy, and was administered a cycle of nedaplatin, vindesine, etoposide, and bleomycin (NVEB) postoperatively. Another patient (No. 18) underwent vaginal tumorectomy via laparotomy, but she did not receive any postoperative chemotherapy in the local hospital.

All 32 patients underwent combination chemotherapy: 30 patients started chemotherapy following pathologic diagnoses, and two patients received chemotherapy after clinical diagnoses based on high AFP levels and imaging findings. Excluding two patients who underwent surgery in local hospitals, the remaining 30 patients initially underwent chemotherapy alone. With regard to chemotherapy regimens, 28 of the 32 patients (87.5%) were initially started on a regimen of BEP (bleomycin/etoposide/cisplatin) chemotherapy. Among them, 21 patients were treated with BEP alone; 7 patients subsequently changed from BEP to other regimens, including PVB (cisplatin/vincristine/bleomycin)/TIP (paclitaxel/ifosfamide/cisplatin)/PE (cisplatin/epirubicin)/PEV (cisplatin/etoposide/vincristine)/VAC (vincristine/actinomycin/cyclophosphamide). Usually, the chemotherapy protocol was changed from BEP to other regimens when patients reached the individualized lifetime maximal dose of bleomycin (250 mg/m²) or cannot tolerate bleomycin-related pulmonary toxicity (Table 3). Four of the 32 patients (12.5%) were started on a non-BEP regimen, including NVEB, JEB (carboplatin/etoposide/bleomycin), VPEB (vincristine/cisplatin/etoposide/bleomycin), and PVB. Three patients were transferred from local hospitals to our hospital. After routine evaluation in our hospital, chemotherapy was continued in these three cases, in which the BEP regimen was chosen with prior consideration, since the bleomycin dosage was within the individualized lifetime maximal dose (250 mg/m²). The patient who was started on a PVB regimen was exclusively treated in our hospital. Her chemotherapy protocol was subsequently changed from PVB to both PEV and VAC to prevent potential bleomycin lung toxicity.

Due to our inexperience with pediatric nursing, we worked in close collaboration with pediatricians and nurses in the management of children. To provide nutritional support and ensure the efficacy and safety of intravenous (IV) chemotherapy, IV infusion ports were implanted subcutaneously in 79.3% of the pediatric patients. In addition to providing nutritional support, IV infusion ports can prevent repeated vascular punctures, reduce the risks of drug extravasation, and protect blood vessels. To prevent cisplatin nephrotoxicity during chemotherapy, 1/3–1/4 tonic fluid (60–80 ml/kg, no more than 100 ml/kg) was administered intravenously at 10–15 drip/min.

In the current study, elevated preoperative serum AFP levels were found in all 32 patients. As a sensitive tumor marker for YSTs, serum AFP was measured every course to evaluate tumor response. Chemotherapy was considered satisfactory if the serum AFP level decreased by 10¹ every course. The normal serum AFP level was <20 U/ml. The median number of total cycles of chemotherapy was five cycles (range, 2–13 cycles), the median number before AFP dropped to normal values was three cycles (range, 1–8 cycles), and the median number after a normalized AFP was two cycles (range, 0–7 cycles). Repeat examinations under anesthesia were usually performed to evaluate tumor response after AFP levels returned to normal levels. A delayed decrease in AFP levels and any suspicious radiological findings were also reasons for repeat evaluation.

A total of 27 patients underwent repeat examinations under anesthesia, of whom 22 had normalized serum AFP levels and 5 had abnormal or near-normal levels. Vaginohysteroscopy was used in 20 cases, and pediatric rhinoscopy and laparoscopy were used separately in 4 cases, either alone or in combination with vaginohysteroscopy. Yellow, grayish-yellow, or yellowish-white tissue varying in size and shape was the most common intraoperative finding during repeat examinations and was found in 18 cases (66.7%) (Figure 3). During repeat examinations under anesthesia, 23 patients underwent biopsies and 4 patients underwent additional procedures due to highly suspected residual disease, including partial vaginectomy (n = 2), vaginal tumorectomy (n = 1), and total hysterectomy plus bilateral salpingectomy (THBS) (n = 1). With regard to the pathological results, no evidence of malignancy was found in 25 cases, a little necrotic and degenerated tumor tissue was found in one case, and persistent disease was confirmed in only one uncontrolled case.

In the current study, myelosuppression, chemotherapy-induced nausea, and vomiting were the most common side effects. Oral ulceration, anorexia, hair loss, and bleomycin-induced interstitial disease were the least common side effects observed. To improve efficacy and reduce chemotherapy-induced adverse effects, supportive care was provided to patients, such as blood transfusion, platelet transfusion, anti-inflammatory treatment, and granulocyte colony-stimulating factor (G-CSF) support. For pediatric patients, G-CSF was administered when the neutrophil count dropped below 0.5 × 10⁹/L with a dosage of 2 μg/kg. The volume of blood transfused for children was 10–15 ml/kg. The vast majority of patients tolerated the toxic effects of multidrug chemotherapy well, except for one patient (No. 11). Her serum AFP level decreased to a normal value after three cycles of BEP, but she developed a fourth degree of bone marrow suppression, severe nausea and vomiting, and pulmonary function decline. Unfortunately, she progressed to acute dyspnea and heart failure after the fourth cycle of BEP (her cumulative dose of bleomycin was 44 mg) and finally died of respiratory and circulatory failure.

During a median follow-up of 63 months (2.4 months to 19.8 years), four (12.5%) patients showed disease progression. Details of the patients with disease progression are shown in Table 4. Thirty-one (96.88%) patients achieved CR after the initial treatment, but one of them (No. 11) died of chemo-related severe adverse events. One patient (No. 6) had uncontrolled disease after the initial treatment and died of the disease. Two patients (No. 3 and 32) experienced recurrence, and one of them (No. 32) died of recurrent disease. The patient with a cervical YST underwent laparoscopic THBS because of radiologic metastasis during follow-up. However, the final pathological diagnosis revealed no evidence of malignancy. To date, three patients (3/32, 9.38%) died, and the remaining 29 patients (29/32, 90.63%) are alive without evidence of disease. In general, we achieved an excellent prognosis (Table 1).

In a total of 27 patients, including one childbearing patient and 26 children, the uterus and both ovaries were preserved. For these 27 patients, the median follow-up period was 67.6 months (2.4 months to 19.8 years). The childbearing patient recovered her menstruation three and a half months after completion of chemotherapy. Five of the 26 patients underwent menarche during the long-term follow-up. None of the 27 patients had conceived yet because of the lack of childbearing requirements.