Survival Analysis of Radiation Therapy in Ovarian Cancer: A SEER Database Analysis

Results A total of 20031 ovarian cancer patients were included, with 291 (1.45%) patients who received radiotherapy. The median overall survival (OS) in patients who received radiotherapy was shorter than which in patients without radiotherapy (23 vs. 75 months, P < 0.001). The Elderly, nonepithelial pathology, advanced American Joint Committee on Cancer (AJCC) stage, elevated level of CA125, and receiving radiotherapy were risk predictors to survival in both multivariable analyses before and after PSM. Among 11872 patients with III/IV stage, the radiotherapy group also showed a significantly worse prognosis (median OS: 19 vs. 44 months in patients without radiotherapy, P < 0.001). Consistent results were observed in stratification analyses on pathology and stage among patients with III/IV stage. Conclusions For patients with ovarian cancer, radiotherapy was associated with a poor prognosis regardless of pathology or stage. Considering this is a retrospective study, future studies concerning radiotherapy combination with other new agents in ovarian cancer are needed.


Introduction
Ovarian cancer (OC), the third most common cause of death in gynecologic cancer, ranked eighth for both cancer incidence and mortality among females in 2018, with 295,414 new cancer cases and 184,799 cancer deaths in 185 countries [1]. e prognoses of OC remain diverse according to pathological type, stage at the first diagnosis, and response to treatment strategy. Epithelial ovarian cancer (EOC) including five major subtypes of high-grade serous, low-grade serous, mucinous, endometrioid, and clear cell ovarian cancer accounts for around 90% of OC. Nonepithelial ovarian cancer mainly consisted of germ cell, sex-cord stromal cancers, and ovarian sarcoma has low morbidity and mortality relative to EOC. Early stage (stage I-II) of OC is highly curable with a 5-year survival rate of 60-80%. Whereas 60-75% of OC firstly diagnosed at stage III/IV, the proportion even up to 80% in serous ovarian cancer. e 5year survival rate of III/IV stage OC patients is only 19-41% [2,3].
Debulking surgery and platinum-based chemotherapy are the main treatments for OC, even for recurrent and advanced OC [4][5][6][7]. Over the past decade, targeted therapy, such as antiangiogenic agents and poly (ADP-ribose) polymerase-1 (PARP) inhibitors [8,9], has shown effects in some patients with OC. Immunotherapy is being tried in OC patients, although no significant advantage was observed till now [10]. As a traditional treatment, radiotherapy (RT) is at an inferior place in the treatment of OC due to the dose limit from adjacent normal tissues and is given only for palliative care in most situations. Nowadays, radiation techniques have been improved to enable directed conformal therapy delivery to the local lesion. us the position of RT in OC may be reconsidered [11]. However, current studies for RT on OC were limited, and the sample sizes were small [12][13][14]. Aimed to explore the survival impact of radiotherapy in OC, we analyzed the real-world data from the Surveillance, Epidemiology, and End Results (SEER) Database.

Database and Population Selection.
e National Cancer Institute's SEER program collects information about cancer patients in the United States, covering about 34.6% of the population, and is provided as a public service. e SEER * Stat software (version 8.3.6) was used to screen eligible patients for analysis. Women who were older than 18 years old with histologically confirmed primary ovarian cancer (ICD-O-3, C569) and had active follow-up record were included preliminarily. Patients who were diagnosed by autopsy or death certificate, had multiple tumors, or without the identified status of radiotherapy were excluded. All included patients were diagnosed from 2010 to 2015.

Variables.
Variables involved in the analysis included were age, race, insurance status, marital status, lateral, tumor size, histological subtype, tumor grade, SEER combined summary stage, tumor stage (according to the 6 th TNM classification of American Joint Committee on Cancer [AJCC] staging system), treatments, and tumor marker (CA125). Age was categorized into groups of ≤40, 40-60, and >60 according to perimenopausal age. Treatments included surgery, chemotherapy, and radiotherapy. e primary endpoint was overall survival (OS) and the second endpoint was cause specific survival (CSS). e OS was defined as the time interval from first diagnosed as OC to death due to any cause. e CSS was defined as the time interval from diagnosis of OC to OC-related death.

Statistical Analyses.
Categorical variables were described as counts (percentage) and compared using the chisquare test or Fisher's exact test. Continuous variables were described as means ± standard deviations (SD) or median (interquartile range) and compared by student's t-test or Kruskal-Wallis test. Univariate and multivariate Cox regression analyses were employed to identify independent predictors associated with survival. Survival comparisons were made using Kaplan-Meier analysis and log-rank tests. A 1 : 1 propensity score matching (PSM) analysis was employed to balance baseline variables for further analyses. A two-side P value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS 22.0 and figures were made in GraphPad Prism 6. Patients and characteristics after PSM were also shown in Table 1. Most variables were balanced, except pathology, SEER combined summary stage, AJCC stage, chemotherapy, and CA125.
Univariable (Table 2) and multivariable cox regression were performed to assess risk factors for OS before and after PSM. In the multivariate analysis before PSM, the age, race, marital status, pathology, tumor size, tumor grade, AJCC stage, and level of CA125 were all associated with survival. Besides, no matter RT (R) alone or in combination with surgery (S) or/and chemotherapy (C) impaired survival (R to non-R: HR � 2.25, P < 0.001; R + S to non-R: HR � 3.06, P < 0.001; R + C to non-R: HR � 2.39, P < 0.001; R + S + C to non-R: HR � 1.59, P < 0.001) ( Figure 2). After PSM, age, insurance, pathology, AJCC stage, and level of CA125 were considered as prognostic factors in multivariate survival analysis. As for therapy, only radiotherapy in combination with surgery showed significantly unfavorable effects on overall survival (Supplementary Figure 1).

Radiotherapy in III/IV Stage Ovarian Cancer.
According to baseline characteristics of patients, radiotherapy was common in III/IV stage ovarian cancer. erefore, we further analyzed OC patients with III/IV stage. Among 11872 patients with III/IV OC, 215 patients received radiotherapy (Supplementary Table 1  Stratification analyses were carried out to control confounders. In stratification analysis according to pathological type, 9612 (80.96%) patients were EOC, with the median OS of 45 and 27 months (P < 0.001) in non-RT and RT groups, respectively (Figure 4(a)). e median OS of non-EOC was only 35 and 12 months (P < 0.001) in two groups, respectively (Figure 4(b)). In stratification analysis according to AJCC stage, 7962 (67.07%) patients were first diagnosed as III stage. e median OS of the III stage was 50 and 32 months (P < 0.001) in non-RT and RT groups, respectively (Figure 4(c)). For 3910 (32.93%) patients with stage IV, the median OS was reduced to 34 and 14 months (P < 0.001) in two groups, respectively (Figure 4(d)).
e CSS in the stratification analysis was similar to OS (Supplementary Figure 2).

Discussion
Historically, conventional whole abdominal radiotherapy (WART) was used in ovarian cancer to reduce recurrence [15], while it has been proved to bring several severe acute and late toxic effects [16][17][18] and has almost been abandoned nowadays. According to National Comprehensive Cancer Network (NCCN) clinical practice guidelines for ovarian cancer, radiation therapy is only recommended for limited disease in malignant sex-cord stromal tumors and is considered as a palliative localized strategy for patients with recurrent disease. Given the advance of techniques of radiation therapy, researches on radiation therapy for OC are still ongoing.
e OVAR-IMRT-01 study, a phase I study, evaluated the clinical feasibility and acceptable toxicity of WART using intensity modulated radiotherapy (IMRT) as consolidation therapy in advanced OC [12]. Following that, the multicenter, single-arm, phase II study, OVAR-IMRT-02 showed that more than 70% of 20 included patients could tolerate intensity modulated WART and a mean decrease of global health status score was 18.1 points [13]. Chang made a prospective phase II trial on involved-field radiotherapy (IFRT) in 30 patients with recurrent EOC. e results showed an overall response rate (ORR) of 85.7%, median progress-free survival (PFS) of 7 months, and a 3year OS rate of 55.8%, revealing the effectiveness of IFRT for local control in EOC [19]. Kim retrospectively analyzed 61 recurrent EOC patients who were treated with IFRT. ey found that the 2-year in-field control, PFS, and OS rates were 42.7%, 24.2%, and 78.9%, respectively. e elevated CA125 level was considered to be related to a worse OS [20]. Chundury gave IMRT to 33 recurrent chemorefractory OC, finding that 2-year local control and OS rates were 82% and 63%, respectively, with limited radiation related toxicity [21]. Although positive results on RT are shown in the researches above, the sample size was small and all of those were single-arm studies without control groups. erefore, phase III randomized controlled clinical trials and big sample size real-world studies are needed to provide more powerful evidence.  Journal of Oncology In this real-world research that included 20031 OC patients, we found that the median OS of the RT group was 23 months, and 1-year, 2-year, and 3-year OS rates were 66.3%, 49.4%, and 40.7%, respectively, which were slightly worse than reported previously [14,19]. e difference could be explained by different sample sizes, radiation fields, dose and modality, and so on. Non-RT as a control group in this study had better survival, with a median OS of 75 months and a 5-year OS rate of 56.1%. On the whole, radiation therapy was commonly given to ovarian cancer patients with nonepithelial type, distant stage and AJCC III/IV stage, so at first, we blamed the poor survival to worse pathological type and more advanced stage in RT group. However, after subgroup and stratification analysis, similar results were obtained. For III/IV OC, median OS was 44 versus 19 months and 3-year OS rates were 56.9% versus 32.5% in non-RT and RT, respectively. For EOC, median OS was 45 versus 27 months in non-RT and RT group, respectively.
Multivariable survival analysis also supported the result above. Except for old age, poor pathology, advanced AJCC stage, and high level of CA125, radiotherapy also was a risk factor for poor survival. Especially, RT combination with surgery may result in shorter survival in III/IV OC patients, though the number of patients in the analysis was only 9 in this study. us, we thought that radiotherapy could not  bring survival benefits for ovarian cancer compared to patients without radiotherapy. Radiotherapy, as well as its combination with surgery or chemotherapy, in this study, did not bring survival benefits to ovarian cancer patients, but some preclinical studies showed that PARP inhibitors might be sensitizing agents for radiotherapy; others also agreed that radiotherapy could promote tumor immunity cycle in OC, which implied the potential synergy of radiotherapy with targeted therapy or immunotherapy in OC [22][23][24].
Our study with a big sample size and real-world data provided some information for clinical practice on radiotherapy in OC. However, there were still several limitations. Firstly, the proportion of ovarian cancer patients who received RT in this study was small, only 1.45% (291/20031).
us some underlying confounding factors could not be ignored. e types of surgery in this study varied from local tumor destruction to cytoreductive surgery, and regimens of chemotherapy were unknowable. Both chemotherapy and surgery can   Journal of Oncology influence survival, especially debulking surgery, which is acknowledged as improving the survival of ovarian cancer. Although we made PSM analysis to balance baseline variables as many as possible, the rates of surgery and chemotherapy were not balanced at the same time. So, surgery and chemotherapy were likely to be confounding factors to impaired survival in RT. Also, the difference in the surgeons who conducted the surgery (gynecological oncologist, general obstetrics and gynecology physician or general surgeon) could affect the scope of surgery, the status of residual lesions, and even survival. Secondly, we did not explore the effect of sequence between radiotherapy and surgery on survival because only a small number of patients received surgery after or during radiotherapy. Due to the lack of detail information of field, dose and modality of radiotherapy, and the information of adverse events, we were not able to compare the effect of different radiation patterns on survival and evaluated the safety of RT in OC.

Conclusions
Radiotherapy was often given to ovarian cancer patients with nonepithelial pathology and advanced stage, while it was associated with poor prognosis compared to patients without radiotherapy. e impact of radiotherapy in combination with other new agents in ovarian cancer is exploring. Large randomized controlled clinical trials are needed.
Data Availability e data used and/or analyzed in this study are available in the Surveillance, Epidemiology, and End Results (SEER) Database of the National Cancer Institute (http://seer. cancer.gov).

Consent
e SEER Dataset was a public-use dataset, of which the informed consent was waived.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this article.

Authors' Contributions
LSY designed the study, collected the data, conducted statistical analyses, interpreted data, drafted and revised the manuscript; QL, YW, and HHH collected and recorded the data, interpreted results, and revised the manuscript; HYG contributions to the conception and design of the study, interpreted data, and revised and reviewed the manuscript; HGR interpreted data and revised and reviewed the manuscript; QBS and TL contributions to the conception and design the study, interpreted data, and revised and reviewed the manuscript; all authors read and approved the final manuscript.