Age and Lymph Nodes Examination May Be Related to the Outcome of Ovarian Clear Cell Carcinoma: A SEER Analysis

Aim To analyze and compare the demographics, treatment, and survival rates in patients with ovarian clear cell carcinoma (OCCC). Methods We conducted a population-based retrospective study examining the Surveillance, Epidemiology, and End Results Program from 1998 to 2016. Data of 4344 women with OCCC were compared, and survival was analyzed using the Kaplan–Meier method. Factors predictive of outcome were compared using the Cox proportional hazards model. There was no signicant difference in cause specic survival (CSS) regardless of chemotherapy in stage I and stage II OCCC. In women with stage III/IV OCCC, there was an increased mortality in women without chemotherapy CSS vs. Among women younger than 60 years old, the 5-year CSS of those underwent chemotherapy was worse than that of non-chemotherapy (86.4% Among these patients, chemotherapy had improved CSS (HR 95% and omitting lymph nodes examination had decreased CSS (HR 95% CI 1.230-2.256). In stage III/IV women who were 60 years or older, the 5-year CSS of those underwent chemotherapy was better than that of non-chemotherapy (32.60% vs. Among these patients, omitting chemotherapy (HR 95% CI and omitting nodes examination (HR 95% CI 1.371-2.130) had lower CSS.


Introduction
Ovarian cancer is the most lethal gynecological neoplasm in the world [1] . More than 90% of malignant ovarian tumors are of epithelial origin, designated epithelial ovarian cancer (EOC) [2] . Over the past few decades, there have been many novel advances in the treatment of ovarian cancer, including chemotherapy, immunotherapy, and targeted therapy. These treatments have substantially improved median survival; however, overall cure rates remain relatively unchanged [3] . Several factors are associated with ovarian cancer survival, including race, use of oral contraceptives, tubal ligation, menopausal hormone use, stage, size of residual tumor after debulking surgery, and histotype [4][5][6][7][8] .
Low-stage OCCC have a relatively good prognosis, however, high-stage OCCC have a poorer prognosis than stage-matched high-grade serous ovarian carcinomas. OCCC in advanced stage has poor prognosis due to its inherent resistance to standard carboplatin paclitaxel chemotherapy [10,11] .
Adjuvant chemotherapy in early-stage OCCC is commonly used and con icting data have been reported [9,12] . Whether early ovarian clear cell carcinoma needs chemotherapy is still controversial. We retrospectively analyzed the data of patients with OCCC obtained from the National Cancer Institute's Surveillance, epidemiology, and End Results (SEER) program. The aim of this study is to investigate the impact of chemotherapy on survival among OCCC patients in different stages and different ages.

Materials And Methods
Data source This is a population-based retrospective observational study examining data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program. The SEER database is a public, de-identi ed database of cancer statistics. It has a more than a four-decade history of operation, and currently covers approximately 35% of the U.S. population. In this study, all data were extracted from the SEER database, and all patients were anonymous. Informed consents were not available.

Study eligibility
Women with OCCC who were diagnosed from 1998 to 2016 were eligible for the analysis. Women who were not diagnosed by positive histology were excluded, including diagnosed by death certi cate, diagnosed at the time of autopsy and diagnosed by cytology without surgery. And women with a previous diagnosis of cancer were excluded from the study. Patients diagnosed outside of the study period were excluded from the study. The SEER Public Use Data include SEER incidence and population data grouped by age, race, year of diagnosis, and marital status in addition to providing information regarding stage of disease; tumor size, lymph node status; treatment with surgery and chemotherapy; and cause speci c survival for each registered patient. Data extraction SEER*Stat 8.3.8. (IMS Inc., Calverton, MD, USA) was used for the data abstraction. The histology code was 8310/3.Age at diagnosis was categorized as <60, ≥60. Marital status was categorized as: married, unmarried and unknown. Race was categorized as: white, black, Asian, and unknown. Tumor stage according to FIGO staging was categorize as: stage I, stage II, and stage III/IV. Tumor size in diameter was categorized as: <10cm, ≥10cm. Performance of lymphadenectomy, surgery, and chemotherapy was noted for each subject. Years of diagnosis were classi ed into 1998-2003, 2004-2010, and 2011-2016. The outcome variables included vital status and the time-to-event from the date of diagnosis until death, censoring, or last follow-up, as veri ed by the SEER program vital status determination.

Statistical Analysis
Median age of each group was calculated by independent samples t-test, and the rest variables were evaluated by Pearson chi-squared test, Fisher exact test. We used the Kaplan-Meier method to estimate survival curves in order to compare observed survival between women with different age and different stage OCCC. Survival curves were constructed to show cause speci c survival within the rst ve years of diagnosis for each tumor stage, although the hazard ratios (HR) and resulting P values were calculated using all available data through last date of follow-up at the end of 2017, not only the rst ve years after diagnosis.. Log-rank test with 95% con dence intervals (CIs) were applied to evaluate the outcomes. Cox proportional hazard model with 95% CIs was used to identify the predictors for cause speci c survival (CSS).We considered p < 0.05 as statistically signi cant. All analyses were performed using the SPSS Statics software, version 20.0 (IBM Corporation, NY, USA).

Patient selection
As shown in Figure 1, 4943 OCCC patients diagnosed from 1998 to 2016 were extracted from the SEER database. 559 patients were excluded from the nal analysis: 39 cases were not diagnosed based on positive histology, 513 cases because of history of previous malignancy, and 47 cases because of disease stage were unknown. The nal study group therefore consisted of 4344 women with OCCC, 2446 (56.31%) women were stage I, 550(12.66%) women were stage II, and 1348 (31.03%) women were stage III or stage IV. The demographic and clinical characteristics of the study population were shown in Table  1.
For patients with stage I disease, the 5-year CSS of patient with and without chemotherapy was 86.40% and 89.70% respectively (P = 0.104). Among patients with stage II disease, the 5-year CSS of patient with and without chemotherapy was 67.40% and 71.80% respectively (P = 0.345). In patients with stage III/IV disease, the 5-year CSS of patient with and without chemotherapy was 29.80% and 24.90% respectively (P<0.001). Five-year CSS for each stage with and without chemotherapy is shown in Table 2. Figure 2 displays Kaplan-Meier analyses of CSS for OCCC patients with and without chemotherapy across all stages.
To further analyze the relationship between age and chemotherapy outcome, we compared the survival time of patients age <60 and those age ≥60 with stage I and advanced stage III/IV. The demographic and clinical characteristics of stage I patients were shown in Table 3. It can be seen that the proportion of patients age <60 receiving chemotherapy was higher than that of age ≥60 in stage I (71.7% vs. 64.6%, p<0.001). The proportion of tumor size ≥10 cm in patients age ≥60 was higher than that in patients age <60 (55.0% vs. 45.3%, p<0.001). The demographic and clinical characteristics of stage III or IV patients were shown in Table 4.The proportion of patients age <60 receiving chemotherapy was higher than that of age ≥60 in stage III /IV (82.7% vs. 77.1%, p=0.012). The proportion of tumor size ≥10 cm in patients age ≥60 was lower than that in patients age <60 (43.4% vs. 51.7%, p=0.004).
For patients age <60 with stage I OCCC, the 5-year CSS of patient with and without chemotherapy was 86.40% and 97.50% respectively (P = 0.002). There was no bene t from chemotherapy in patients with stage I OCCC who were 60 years or older, and the 5-year CSS of these patients with and without chemotherapy was 86.40% and 84.60% respectively (P >0.05). Among patients age ≥60 with stage III/ IV disease, the 5-year CSS of patient with and without chemotherapy was 32.60% and 24.30% respectively (P<0.001). Meanwhile, there was no signi cant difference in the cause speci c survival of women age <60 with or without chemotherapy, and the 5-year CSS of these patients with and without chemotherapy was 28.10% vs. 26.40% respectively (P >0.05). Five-year CSS for each group with and without chemotherapy is shown in Table 5. Figure 3 displays Kaplan-Meier analyses of CSS for each group with and without chemotherapy.
We further analyzed the factors related to the prognosis of patients age<60 in stage I OCCC. The Cox proportional hazards model identi ed an independent association of tumor diameter ≥10cm, absence of lymph node dissection and chemotherapy with cause speci c mortality ( The Cox proportional hazards model identi ed an independent association of absence of lymph node dissection and omitting chemotherapy with cause speci c mortality (Table 7). Among patients age≥60 with stage III/IV, omitting lymph nodes examination had decreased CSS compared to receiving lymph nodes examination (HR 1.709; 95% CI 1.371-2.130). Omitting chemotherapy had lower cause spec c survival compared to receiving chemotherapy (HR 1.769; 95% CI 1.385-2.258).

Discussion
Ovarian cancer is a heterogeneous group of tumors. It is a distinct subtype of epithelial ovarian cancer that demonstrates a different clinical behavior from other histologic subtypes and is frequently associated with endometriosis [13] .Ovarian clear cell carcinoma is considered to be relative resistant to chemotherapy. The reported response rates to chemotherapy range from 22% to 56% in patients with OCCC [14] .
It remains controversial whether patients with OCCC may truly bene t from adjuvant chemotherapy, especially in patients with early-stage OCCC. Some studies suggested that chemotherapy has a role in improving disease-free survival. Shimizu D et al. [15] demonstrated that patients with stage I OCCC who received adjuvant chemotherapy had better disease-free survival than patients who did not. Bogani G et al. [16] observed that the administration of chemotherapy associate with an improvement in term of 5-year overall survival in patient with stage IC OCCC. Other investigations suggested that the administration of chemotherapy is not useful in OCCC. Lee H Y et al. [10] reported that in stage IA or IB patients, adjuvant chemotherapy was not related to longer relapse free survival. Takano M et al. [17] suggested that adjuvant chemotherapy had little impact on the survival of stage I OCCC patients.
At present, various clinical guidelines have different recommendations for the treatment of early-stage ovarian clear cell carcinoma. The National Comprehensive Cancer Network (NCCN) demonstrate that adjuvant chemotherapy is optional to women with stage IA, but recommended to women with stage IB/IC, since ovarian clear cell carcinoma is regarded as grade 3 tumor, and has high-risk characteristic for relapse [18] [19] . The European Society of Medical Oncology (ESMO) practice guidelines recommend the bene t of adjuvant chemotherapy is uncertain for patients with early-stage ovarian clear cell carcinoma and should be discussed on an individual patient basis [2] . The Gynecologic Cancer Intergroup (CGIC) suggests that observation could be acceptable for those with surgical stage IA disease given the observed excellent survival rates [7] .
In this study, we used a large sample of the SEER database to analyze the characteristics, prognosis, and factors related to the prognostic outcome of patients with ovarian clear cell carcinoma. We found that the administration of chemotherapy has not impact on stage I OCCC in term of 5-year cause speci c survival.
Moreover, after further strati cation by age, we found that in stage I OCCC patients younger than 60 years old, the cause speci c survival of those receiving chemotherapy was signi cantly lower than those omitting chemotherapy. Meanwhile, whether adopt chemotherapy or not has no signi cant difference on cause speci c survival in advanced OCCC. This may be related to the natural instinct of chemotherapy resistance of OCCC. After strati cation by age, we found that in patients age≥60 with advanced stage OCCC, omitting chemotherapy had lower cause spec c survival compared to receiving chemotherapy. It is the rst time that age has been found to be related to the prognosis of chemotherapy in ovarian clear cell carcinoma.
According to the current retrospective literature, the necessity of systemic lymph node dissection in patients with OCCC is still controversial. Suzuk et al. [20] demonstrated that signi cant differences were observed in progression-free survival and overall survival between patients optimally and nonoptimally staged with stages IA/IC1 OCCC, but no signi cant difference was found in those with stages IC2/IC3. Surgical staging category was the only independent prognostic factor for recurrence-free survival in stages IA/IC1 OCCC. Yamazaki et al. [21] illustrated that pelvic lymph node dissection and para-aortic lymph node dissection were signi cantly and independently related to longer disease-speci c survival.
The study from Hirose et al. [22] showed the tendency that patients who received systematic lymph node dissection occurred fewer lymphogenous recurrences. In a multicenter retrospective study, Takano et al. [23] found that completion of surgical staging procedures was not a prognostic factor for overall survival in OCCC.
The NCCN treatment guidelines recommend investigation of retroperitoneal lymph nodes in early-stage ovarian cancer, and dissection of clinically negative nodes is not required for patients with stage≥IIB ovarian cancer [19] . The Japan Society of Gynecologic Oncology (JSGO) guidelines demonstrate that for patients with stage I-IIA ovarian cancer, pelvic/para-aortic lymph node dissection (biopsy) is recommended in addition to bilateral salpingo-oophorectomy + total hysterectomy + omentectomy + peritoneal cytology + biopsies from sites in the abdominal cavity. For patients thought to have stage IIB or higher ovarian cancer, pelvic or para-aortic lymph node dissection is suggested to be not performed if no lymph node metastasis is clinically detected on imaging or by intraoperative palpation and visual inspection. When lymph node metastasis is clinically detected on diagnostic imaging or by intraoperative palpation and visual inspection, pelvic or para-aortic lymph node dissection or removal of swollen lymph nodes is recommended if complete resection can be achieved [24] . However, Magazzino et al [25] retrospectively assessed an Italian cohort of patients with clear cell ovarian cancer observed in the years 1991-2007 in 20 Italian centers, and found that disease-free survival was longer in patients undergoing lymphadenectomy at surgery, both in early stages and in stage III and IV diseases. The impact of lymphadenectomy was also evident on overall survival in patients with advanced-stage disease.In this study, we found that among patients age<60 in stage I and patients age≥60 with stage III/IV, omitting lymph nodes examination had decreased CSS compared to receiving lymph nodes examination. Therefore, the therapeutic signi cance of lymph node dissection for OCCC patients needs to be further clari ed.

Conclusion
The present investigation reviewed the current evidence on the role of adjuvant chemotherapy in OCCC. We found that chemotherapy has different effects in patients with OCCC at different stages and ages. Lymph nodes examination and age may be related to the outcome of ovarian clear cell carcinoma. One weakness of our study is its retrospective nature and the lack of preoperative assessment. Further studies are needed to clarify the role of chemotherapy in ovarian clear cell carcinoma.

Declarations Ethics approval and consent to participate
In this study, all data were extracted from the SEER database, and all patients were anonymous. Ethics approval was waived. Informed consents were not available.

Consent for publication
In this study, all data were extracted from the SEER database, and all patients were anonymous. Consent for publications were not available.

Availability of data and materials
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. They could be achieved upon reasonable request to the authors.    Kaplan-Meier survival curves for CSS for each group of OCCC strati ed by stage. Figure 3 Kaplan-Meier survival curves for CSS for each group of OCCC strati ed by stage and age.