Radical Hysterectomy Versus Simple Hysterectomy and Brachytherapy for Patients with Stage II Endometrial Cancer

Purpose To compare the survival outcome between radical hysterectomy and total hysterectomy with radiation therapy in patients with stage II endometrial cancer. Methods This is a retrospective cohort study. We identied 1349 patients diagnosed with stage II endometrial cancer from Jan 1, 1988 to Dec 31 2015 in the Surveillance, Epidemiology, and End Results. Patients were divided into two groups based on the primary treatment (total hysterectomy combined with brachytherapy or radical hysterectomy). All patients received external beam radiation therapy after the surgery. The primary outcome was the rate of 5-year-cause-specic survival and 5-year-overall survival. Results A total of 1349 patients were enrolled in the study, 117(7.35%) patients received radical hysterectomy and 460 patients who received total hysterectomy combined with vaginal brachytherapy were selected as control. All patients received external beam radiation therapy after the surgery. Overall, the median follow-up duration was 82.77±1.44months (95%CI: 79.94-85.61months). There was no difference in the baseline information between two groups, including ages, ethnicity, and rates of histologic subtypes. The 5-year overall survival was 62.31% among women who underwent radical hysterectomy which was lower than 78.48% among those who underwent total hysterectomy combined with vaginal brachytherapy (HR, 2.22; 95% CI, 1.52 to 3.24; P <0.001 by the log-rank test). Women who underwent radical hysterectomy also had shorter 5-year cause-specic survival (74.60 vs.85.38%; HR, 1.91; 95% CI, 1.13 to 3.23; P =0.01 by the log-rank test) than those who underwent total hysterectomy combined with vaginal brachytherapy. However, the negative outcomes were further validated in patients with high-risk endometrial cancer, not in patients with grade 1-2 low-risk endometrial cancer both on cause-specic survival and overall survival. In patients with grade 3 low-risk endometrial cancer, the tendency was only found with lower overall survival not cause-specic survival. Conclusions This study revealed that in patients’ stage II endometrial cancer, radical hysterectomy was associated with shorter overall survival and cause-specic survival than total hysterectomy combined with vaginal brachytherapy. The choice of different treatment modalities should base on the histology subtype of patients and further study based on molecular classication is needed.


Introduction
The incidence and mortality rates of endometrial cancer (EC) are increasing among women in developing countries and remain high in developed countries [1,2]. As a general principle, surgical staging is recommended in patients with localized EC which include total hysterectomy (TH) along with bilateral salpingo-oophorectomy (BSO), peritoneal cytology lavage, and dissection of lymph nodes if indicated [3][4]. For patients with suspected or gross cervical involvement, radical or modi ed radical hysterectomy may improve local control and survival when compared with TH [5,6]. For patients with stage II disease who have had a radical hysterectomy with negative surgical margins and no evidence of extrauterine disease, observation or o external beam radiation therapy (EBRT) are options. Alternatively, the patient may undergo EBRT and vaginal brachytherapy (BT) [5].
To date, there is no better choice due to limited comparative studies between the outcomes of two modalities. In developing countries, especially the radiation equipment not widely used, the radical hysterectomy may be a better choice. For stage II EC patients who received radical hysterectomy with a negative surgical margin, radiotherapy could be avoided in low and intermediate-risk women [7]. In previous systematic review, radical hysterectomy did not bring a signi cant survival bene t for patients with stage II EC from either overall survival (OS) or progress free survival. Radical hysterectomy showed a 27% survival bene t in earlier studies based on Federation International of Gynecology and Obstetrics (FIGO) 1988 staging, whereas showed increased no bene t in newly published studies based on FIGO 2009 staging [8]. However, all have no statistical signi cance. The reasons underlying this are still unclear. Surgeons may tend to perform radical hysterectomy in more severe cases with poorer prognostic factors. The advances in radiotherapy may lower the risk of pelvic or vaginal recurrence resulting from insu cient parametrial margins in TH.
This study aimed to evaluate the survival of different treatment models in patients with EC involved cervix. We conducted a retrospective cohort study of 577 patients to give a more complete picture of the real-world clinical outcomes to enhance the body of different treatment modalities.

Study design and patient selection
This study was a retrospective study involving data from patients' stage II EC, registered in the Surveillance, Epidemiology, and End Results (SEER) registry (Third Edition, SEER 18 registry database November 2019 submission) [9] from 1988 to 2015. This database covers approximately 27.8% of the U.S. population and is publicly available and de-identi ed. The data reported in this study represent the most recent follow-up (April 30, 2019) available in the SEER database.
Patients diagnosed between Jan 1, 1988, and Dec 31, 2015, with primary EC involving cervix treated with radical hysterectomy or TH combined with vaginal BT were eligible for participation. All patients received EBRT after the surgery. Patients who didn't undergo radiation therapy (RT) for any reason were excluded, and we also excluded patients if we could not determine whether they matched the inclusion criteria because of missing data (e.g., surgery information or postoperative radiotherapy).

Study Procedures and Data Collections
SEER*Stat 8.2.3 was used to extract the data and women ful lling the aforementioned enrollment criteria were offered participation in the study. Detailed demographic, oncological, and survival data were collected. We divided our cohort into two groups according to treatment modalities: TH (group A) combined with vaginal BT and radical hysterectomy (group B). Group A was considered having postoperative with vaginal BT after TH with Federation International of Gynecology and Obstetrics (FIGO) stage II EC, Group B contained patients with FIGO stage II EC and didn't receive vaginal BT, all patients received EBRT after the surgery. The cancer stage was reclassi ed into FIGO 2019, based on tumor size, tumor extension, and lymph node status recorded in the database.
Propensity score matching for each group was computed for each case determined by multivariable logistic regression analysis. Patient demographics, tumor characteristics, and treatment patterns were entered in the propensity score model. Four-to-one propensity score matching between Group A and Group B was performed through an automated algorithm with the propensity score difference cut off being 1%.

Outcome Measurements
Survival data, including 5-year-cause-speci c survival (CSS) and 5-year-OS (all-cause mortality), are collected through linkages with state mortality records and the National Death Index. CSS was de ned as the time interval between the initial diagnosis of uterine EC and the date of death resulting from this speci c disease. OS was de ned as the time interval between the initial uterine EC diagnosis and the date of death for any reason. Among women who died, causes of death were examined (uterine EC and other diseases) and grouped as previously described.
The primary outcome was to examine the 5-year-OS among women in the two groups, respectively. The secondary outcome was to examine the 5-year-CSS in the two groups, respectively.

Statistical analysis
Rank sum test or χ2 test was used to exam base-line characteristics: age at diagnosis, ethnicity, year at diagnosis, grade, surgery modality, postoperative radiation, chemotherapy. Cox regression was used to evaluate proportional hazard regression models, and the magnitude of statistical signi cance was expressed with a hazard ratio (HR) and 95% CI.
We employed Kaplan-Meier analysis to construct survival and cumulative risk curves, and statistical signi cance between the curves was compared with log-rank tests. Survival was also examined using Cox Covariates, entered in the nal model were patient demographics, tumor factors, and treatment patterns.
Statistical analyses were conducted using SPSS23. All P-values re ected 2-sided tests, and signi cance was set at < 0.05.

Study Population
A total of 1718 patients were screened and 1349 patients were enrolled in the study from 1988 to 2015 (Fig. 1). All patients were found with cervical stromal invasion and received cancer-directed hysterectomy and postoperative EBRT. 117(7.35%) patients received radical hysterectomy and 460 patients who received TH combined with vaginal BT were selected as control. The baseline characteristics of the patients are summarized in Table 1. The mean age of the patients in group A was comparable to the patients in group B (61.27 ± 12.34 vs.61.39 ± 11.80 years, p = 0.92). There were no signi cant differences between the two groups with respect to age(p = 0.96), ethnicity(P = 0.06), year at diagnosis(P = 0.06), grade(P = 0.63), chemotherapy(p = 0.51).  Figure 2 shows the results of Kaplan-Meier analyses of 5-year OS ( Fig. 2A) and 5-year-CSS (Fig. 2B) 7 . However, other studies found that the type of hysterectomy was not an independent prognostic factor for Disease-free survival and OS in patients with stage II EC. Besides, they reported that perioperative and late adverse events were more common in patients receiving radical hysterectomy [11][12][13]. Although surgical resection is the standard treatment for patients with early-stage EC, randomized studies have shown that adjuvant radiation treatment for patients with early-stage disease and high-risk factors reduced rates of local recurrence [14,15], such as age ≥ 60 yrs., deep myoinvasion, lymph vascular invasion (LVSI) positive, and high-grade histology decreased rates of local recurrence. In previous studies in patients with stage II EC [16,17], no difference was found between TH followed by both whole pelvic and vaginal BT or by radical hysterectomy alone. However, these results are biased by heterogeneity and small sample sizes. And the staging system from FIGO 2009 had a different de nition of Stage II EC.
In our study, we found that in the case of external irradiation in both groups, the survival period of TH plus vaginal BT was higher than that of the radical operation group. However, the negative outcomes were further validated in patients with high-risk EC, not in patients with grade 1-2 low-risk EC both on causespeci c survival and OS. Grade 3 low-risk EC was only found with lower OS, not CSS. The different results in grade 1-2 low-risk EC patients indicated that other reasons undying the death of this population. Now endometrial cancers were classi ed into four categories: POLE-mutated (POLE-mt), mismatch-repairde cient (MMR-d), p53-abnormal (p53abn), p53-wild-type (p53wt) [18] .
High risk endometrial cancer (uterine carcinosarcoma, uterine serous carcinoma, and clear cell carcinomas) patients have a higher P53-abn than endometrioid carcinoma, even grade 3 endometroid endometrial cancer patients [19][20][21]. MMR de ciency was also prevalence in high-risk endometrial cancer [22]. The majority of grade 1-2 endometrial cancers were classi ed into the POLE-mt and p53-wt group which parameters of aggressiveness of this type is low [18]. Molecular classi cation of grade 3 endometrial cancer reveals that these tumors are a mixture of molecular subtypes of endometrial carcinoma, rather than a homogeneous group [23] .
Prevalence of risk factors (such as grade 3, deep myometrial invasion, LVSI, and lymph node involvement) is highest in the p53-abn group. The POLE-mt group was the only group that showed a null prevalence of lymph node involvement (0%). Simple hysterectomy and bilateral salpingo-oophorectomy maybe an appropriate selection. POLE-mt did not increase sensitivity to radiotherapy nor chemotherapeutics in mouse-derived embryonic stem cells [24]. The prevalence of parameters of aggressiveness (LVSI-positive and deep myometrial invasion ) was higher in the MMR-d group than the POLE-mt group. The p53-abn type showed the worst prognosis of four molecular types and the POLE-mt group has the exceptionally favorable prognosis. [19]. Fewer recurrences and a lower risk for cancerspeci c death were found in patients with POLE-mt grade 3 tumors than other type grade 3 tumors [25] .The prognosis of the p53-wt group varies from good to moderate [26]. Our study was in concordance with the above results which underlying the great need for molecular-driven clinical trials to stratify the patients into a better therapy modality on type of hysterectomy and adjuvant RT.
Our study did not acquire any information on tumor recurrence or exact details, which could have helped to investigate differences in progression-free survival. Data from large-scale trials and prospective multicentered studies are needed because of the rarity of Stage II EC in a single center.

Conclusions
This study revealed that in patients stage II EC, radical hysterectomy was associated with shorter OS and CSS than TH combined with vaginal BT in patients with high-risk EC.

Declarations Authors Contribution
Dr. Wu proposed the concept and designed the study. Drs Wang and Ran contributed to the acquisition of data. Dr Wang performed the statistics. Dr Wang interpreted the data and wrote the manuscript at the help of Ran. Dr. Wu performed critical revision of the manuscript and addressed the comments from the journal.
Compliance with ethical standards Con ict of interest: All authors declare that they have no con ict of interest.
Ethics approval and consent to participate: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by the Ethic Committee (EC) from the Beijing Gynecology and Obstetrics Hospital. All data involving in this study come from the SEER registry.
Availability of data and material: The datasets generated and analyzed during the current study are available in the SEER database.
Funding: There is no funding support for this study.
Informed consent: Informed consent was obtained from all individual participants included in the study Figure 1 Patients selections and study design Cox proportional hazards model was used to determine the hazard ratio and 95% con dence interval.
Tick marks indicate censored data. Figure 3