Clinical Characteristics and Risk Factors Analysis for the Recurrence of Pelvic Endodermal Sinus Tumors

Background: This study investigated the clinicopathological characteristics and factors inuencing the recurrence of pelvic endodermal sinus tumor. Methods: Fifty-four cases were retrospectively analyzed from at the Zhejiang Cancer Hospital. Imaging and serological indicators were used to determine whether disease recurred, to evaluate progression-free survival, and to compare the inuence of related factors on disease recurrence. SPSS 19.0 software was used for statistical analysis. Statistical signicance was dened as p < 0.05. Results: The median age at initial treatment was 21 years (range, 11–52 years). Six patients had extragonadal endodermal sinus tumor, and four had histological features of endodermal sinus tumor combined with embryonal carcinoma. Thirty-nine patients underwent fertility-preserving surgery, 18 patients had a childbearing history, and eight patients had residual tumor after initial treatment. Twenty-six patients had a tumor diameter of more than 15 cm, and 30 patients had a serum α ‐ fetoprotein level greater than 10,000 ng/mL before initial management. The median follow-up time was 47.5 months (range, 14–212 months). During follow-up, 15 patients experience recurrence, with a recurrence rate of 27.8% and a 5-year PFS rate of 61.1%. In univariate analysis, the International Federation of Gynecology and Obstetrics stage (stage III-IV VS. I-II; HR= 10.054 p<0.001), residual tumor (yes VS. no for the rst surgery; HR=5.014 p=0.001), histological features (endodermal sinus tumor combined with embryonal carcinoma VS. endodermal sinus tumor; HR=4.130 p=0.018), and use of platinum-based chemotherapy (courses ≥ 3 VS. courses<3; HR= 0.188 p=0.004) were independent factors inuencing recurrence; age, childbearing history, tumor site, tumor size, and serum α-fetoprotein level before initial management did not affect recurrence. In multivariate analysis, only stage was an independent risk factor for progression-free survival(cid:0)stage III-IV VS. I-II; HR=6.923 p=0.019(cid:0). Conclusions: Stage is a prognostic factor for recurrence of pelvic endodermal sinus tumor. The rst surgery should remove the tumor as completely as possible, and initial treatment should require a sucient dose and full course of platinum-based chemotherapy, which may reduce the recurrence rate. Patients with endodermal sinus tumor and embryonal carcinoma may have increased susceptibility of recurrence. review were compared with the baseline data. Recurrence was dened as complete remission after initial treatment, increased tumor marker AFP, and/or imaging ndings indicating tumor progression 1 month after discontinuation of treatment. The time points of recurrence were recorded. Progression-free survival (PFS) was dened as the interval between the initial treatment and tumor re Progression-free survival recurrence. If the patient was lost the follow-up, PFS was dened as the disease-free survival time at the last follow-up.

The pelvic endodermal sinus tumor (PEST) is de ned as an EST that occurs in the pelvic cavity.Patients with PEST usually present with pelvic masses accompanied by abdominal pain, abdominal distension, vaginal bleeding, and similar symptoms.
PESTs are sensitive to chemotherapy and had a high mortality rate until the development of effective chemotherapy drugs. Since the 1980s, the use of platinum-based chemotherapy has greatly improved the prognosis of this disease [7].However,tumor recurrence still occurs frequently.
PESTs are extremely rare, and little information is known about their clinical features and treatment outcomes. Moreover, there is currently a lack of studies about PEST with large sample sizes, so there has been no accurate conclusion about the factors related to recurrence in patients with PEST. If we can determine the relevant factors that affect the clinical recurrence of PEST, treatment will vastly improve.By clarifying the appropriate treatments for PEST, we can more rationally use medical resources and avoid wasting unnecessary medical resources to treatthese patients.The purpose of this study was to investigate the clinical features of PEST and the factors affecting the recurrence of PEST. level before the initial treatment, surgical approach and method, and recurrence.

Follow-up protocol
After treatment, the patients were observed every 2-4 months in the rst 2 years, every half year in years 3-5, and every year after 5 years. The follow-up included assessment of the physical condition, physical examination, serum tumor markers, and imaging-related examinations. Patients wereobserved via outpatient visits or a telephone connection. The follow-up period continued until February 28, 2021, or the tumor recurrence date.

Evaluation indicator
After the initial treatment, data from theimaging examination and serological examination were collected as the baseline characteristics. Patient health was regularly reviewed by imaging and serological indexes. The imaging and serological indexes of each review were compared with the baseline data. Recurrence was de ned as complete remission after initial treatment, increased tumor marker AFP, and/or imaging ndings indicating tumor progression 1 month after discontinuation of treatment. The time points of recurrence were recorded. Progression-free survival (PFS) was de ned as the interval between the initial treatment and tumor re Progression-free survival recurrence. If the patient was lost the follow-up, PFS was de ned as the disease-free survival time at the last follow-up.

Results
The clinical characteristicsof 54 patients with PEST are summarized in Table 1. The median age at initialtreatment was 21 years (range, 11-52 years). Thirty-six patients (66.7%)had a childbearing history;the serum AFP level before initial managementwas ≥10,000 ng/mL in 30 patients (55.6%);six patients had extragonadal EST; and the tumor size was ≥15cm in 26 patients (48.1%).
Of the 12 patients who did not receivefertility-sparing surgery, ve patients underwent total hysterectomy with BSO;six patients underwent radial hysterectomy with USO; one patient underwent subtotal hysterectomy with USO; and 10 patients underwent lymphadenectomy. Omentectomy was performed in eight patients.
According to the pathological and stage criteria, the numbers of patients by stage were 26 in stage I, three in stage II, 22 in stage III, and three in stage IV.
When it came to chemotherapy, the median course of postoperative platinum-containing chemotherapy was four courses (range, 0-7 courses) in 54 patients, of which two patients did not receive adjuvant therapy. Fifty-two patients received platinum (cisplatin or carboplatin) chemotherapy. Among them, 46 patients received bleomycin+etoposide+cisplatin (BEP) chemotherapy;three patients received BEP and etoposide+cisplatin (EP) chemotherapy; one patient received BEP with cisplatin intraperitoneal perfusion (IP) chemotherapy; one patient received paclitaxel+carboplatinchemotherapy; and one patient received vindesine+carboplatin+ifosfamide, vindesine+pirarubicin+cisplatin, and cisplatin IP chemotherapy at different times ( Table 3).
Fifteen of 54 patients experienced recurrence and had an average recurrence time of 13.1 months. Of these 15, two patients had FIGO stage I,10 patients had FIGO stage III, and three patients had FIGO stage IV. Two patients who did not receive chemotherapy experienced recurrence after primary surgery, and one patient with recurrencehad only received platinum-containing chemotherapy twice after the initial treatment. The histological featuresof three patientswith recurrent PEST wereEST with EC. Six patients with recurrence had residual tumor after the rst surgery.
After univariate analysis, the signi cant risk factors for PFS in patients with PEST included the course of platinum-based chemotherapy, the FIGO stage, thehistological features (EST combined with EC), and the presence of residual tumor (P<0.05). Age,childbearing history, tumor site, fertility-preserving surgery, tumor size, and serum AFP level before initial management did not affect PFS (P>0.05; Table 4, Figs. 1-4).
Only FIGO stage was an independent risk factor for PFS after multivariate analysis (Table 5).

Discussion
EST is the second most common ovarian germ cell malignancy [8] and ischaracterized by varied and diverse histological patterns [9].Most ESTs synthesize AFP,so elevation of this tumor marker cansupport a diagnosis. Fifty percent of ESTs are mixed and contain other germ cell elements, including EC, teratoma, germinoma, and choriocarcinoma.In addition, Schiller-Duval body is a histopathological feature of EST [5].
Our studyincluded a large sample size from a singlecenter for the retrospective study of PEST. The study included 48 patients with ovarian EST and six with extragonadal EST-a rare representation in the existing literature.
With regard to risk factors for the EST recurrence, multiple researchers have attempted to determine accurate markers. Yang et al. [10][11]believed that the size of postoperative residual tumor affected the prognosis of ovarian EST. A smaller residual tumor re ected a better prognosis. Kawai et al. [12]reported that the 5-year survival rate of patients with residual tumorslarger than 2 cm (36%) was signi cantly lower than that of patients with residual tumorssmaller than 2 cm (82%; P<0.05). Nawa et al. [13] and Wang et al. [14]also reported that patients with small residual tumors after initial surgery had better prognoses. However, Nasioudis et al. [15] found no signi cant correlation between residual tumor and prognosis. Cicin et al. [16]reviewed the surgical data of 32 patients with ovarian EST: 10 patients underwent hysterectomy and BSO, 18 patients underwent USO, two patients underwent BSO, and two patients underwent cystectomy. The results showed no signi cant difference in e cacy between fertilitypreserving surgery and radical surgery (16). Zhao et al. [17]retrospectively analyzed the data of 130 patients with malignant ovarian germ cell tumor. The results showed no signi cant difference in overall survival (OS) and disease-free survival between patients with bilateral ovarian germ cell tumor and patients with unilateral ovarian germ cell tumor after 5 years. Mahdi et al. [18] studied the clinical data of 1,529 patients with bilateral germ cell tumors, spanning 19 years, from a United States cancer center. The results showed no signi cant difference in 5-year OS and PFS between patients with bilateral and unilateral ovarian germ cell tumors [18]. Frazier et al. [19], Neeyalavira et al. [20], Chen et al. [21], and Zhang et al. [22]believe that stage is an important factor affecting the prognosis (i.e., earlier stageis associated with better prognosis).
In the univariate analysis of our study, we found that residual tumor (hazard ratio In the literature, Gershenson et al. [23]reported the treatment of patients who had ovarian EST with BVP(bleomycin+vincristine+cisplatin) chemotherapy. The e cacy of treatment in stage I and II was 95%, and e cacy in stages III and IV was 80% and 60%, respectively. The e cacy of this regimen for patients with recurrence was 40%. Chen et al. [21] noted that a reasonable platinum-containing chemotherapy regimen can bene t patients with ovarian EST. In addition, reports by Nawa et al. [13],Cicin et al. [16], and Mitchell et al. [24]also identi ed a statistically signi cant difference in 5-year survival rates between patients with malignant germ cell tumor treated with platinum and those not treated with platinum.
In this study, the course of platinum-containing chemotherapy was a related factor that affected the recurrence of PEST (HR=0.004, 95% CI: 0.052-0.677, P=0.004). This result might be due to the sensitivity of PEST to platinum-based chemotherapy; if so, a su cient amount of platinum-based chemotherapy should be given after surgery.
Patients with pure ESThave shown better prognoses than patients with other pathological types [19]. Some studied have shown no signi cant difference in prognosis between pure ovarian EST and germ cell tumor with mixed EST components [16,25]. However, Chen et al. [21] and Zhang et al. [22] noted that the pathological type is not a factor that affects the prognosis of ovarian EST.
We found that patients in our study with EST+EC were more likely to experience relapse than patients with EST without EC were(HR=4.130, 95% CI: 1.156-14.761, P=0.018). This nding has been rarely reported in the domestic or international literature.
Our results might be related to the following reasons: First, the degree of malignancy of ESTwasmuch higher than that of other types of germ cell tumors, such as dysgerminoma. The biological behavior of tumors associated with PEST combined with ECmight be more malignant.Thus,it would be easier for patients whose histological features were EST combined with EC to experience relapse. Second,the chemosensitivity of EC is worse than that of other types of germ cell tumors, which leads to a higher risk of recurrence. However, this study still lacked su cient sample size, so additional researchis needed to verify this conclusion.
Somestudies have reported other factors related to ovarian EST.Wang et al. [14] and de la Motte Rouge et al. [26] reported that AFP level affected the recurrence of ovarian EST, whereas Elashry et al. [27] found that AFP did not affect the prognosis of ovarian EST. Solheim et al. [28] reported that the prognosis of patients older than age 50 years was signi cantly worse than that of patients who were younger than age 50 years. Bilici etal. [29] also considered age a factor in uencing prognosis, but Neeyalavira et al. [20] and Chen et al. [21] did not agree.
In this study, we found that age,childbearing history, and serum AFP level before initial management (≥10,000ng/mL) were not factors in uencing the recurrence of PEST.

Conclusion
We found that stage is a prognostic factor for recurrence of PEST. The rst surgery for PEST should remove the tumor as thoroughly as possible, and the initial treatment should include su cient doses of platinum-based chemotherapy, which may reduce the recurrence rate. Patients with histological features of EST combined with EC may be more likely to experience relapse.
Some shortcomings tothis study still exist. The incidence of PEST is relatively small, and the known cases are limited. To our knowledge, the number of patients with histological features of EST combined with EC is extremely scarce. Therefore, larger sample clinical studies are needed to con rm the conclusions of this study.
This study primarily explored the factors in uencing tumor progression in patients with PEST. Stage was a prognostic factor for recurrence of PEST, and the recurrence rate increased with increasing stage. The rst surgery for PEST should remove the tumor as completely as possible, and the initial treatment should require su cient doses and a full course of platinum-based chemotherapy to reduce the recurrence rate. In practice, patients with histological features of EST combined with EC may have an increased susceptibility to recurrence. However, because of the low incidence of PEST and the limited number of known cases, the conclusions need additional validation with large sample studies.

Declarations
Acknowledgements Not applicable Authors' contributions YLG, TZ, and YLZ conceived and designed the study. YLG, QQL, XC extracted and analyzed data. TZ and YLZ were the statisticians who con rmed the analysis of this study. YLG wrote the manuscript. XC and YLZ critically revised the manuscript. All authors reviewed and approved the nal manuscript.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
Approval to review the patients' medical records and pathological reports was obtained from the institutional review board of Zhejiang Cancer Hospital, Zhejiang, China. Written informed consent was obtained fromall patients.

Consent for publication
Written informed consent was obtained from the patient for publication and accompanying images. A copy of the written consent is available for review by the Editor of this journal.