Article Text

Adjuvant treatment and outcomes for patients with stage IIIA grade 1 endometrioid endometrial cancer
1. Mary Katherine Montes de Oca1,
2. Benjamin B Albright2,
3. Angeles Alvarez Secord2,
4. Laura J Havrilesky2 and
5. Haley A Moss2
1. 1Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, North Carolina, USA
2. 2Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, North Carolina, USA
1. Correspondence to Dr Mary Katherine Montes de Oca, Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, NC 27710-4699, USA; mm765{at}duke.edu

## Abstract

Objective The role and type of adjuvant therapy for patients with International Federation of Gynecology and Obstetrics (FIGO) stage IIIA grade 1 endometrioid endometrial adenocarcinoma are controversial. This retrospective cohort study aimed to determine associations between adjuvant therapy use and survival among patients with stage IIIA grade 1 endometrial cancer.

Methods Patients who underwent primary surgery for stage IIIA (FIGO 2009 staging) grade 1 endometrial cancer between January 2004 and December 2016 were identified in the National Cancer Database. Demographics and receipt of adjuvant therapy were compared. Overall survival was analyzed using Kaplan–Meier curves, log-rank test, and multivariable Cox proportional hazard models.

Conclusions The use of adjuvant chemotherapy for the treatment of stage IIIA grade 1 endometrial cancer increased over time and was associated with improved overall survival compared with radiation alone or chemoradiation. Patients with lymphovascular space invasion may benefit from combination therapy.

• endometrial neoplasms

## Data availability statement

Data are available upon reasonable request. Data are available upon reasonable request. The data from our study were extracted from the National Cancer Database repository with permission. The data are available upon request to Mary Katherine Montes de Oca at mm765@duke.edu.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

### HIGHLIGHTS

• Chemotherapy is associated with improved survival for patients with stage IIIA grade 1 disease.

• Adding radiation to chemotherapy is not associated with survival benefit over chemotherapy alone.

• Lymphovascular space invasion may be predictive of worse survival.

## INTRODUCTION

Endometrial cancer is the fourth most common cancer in women in the USA.1 While most patients present with early stage disease that is generally curative, those with advanced disease have a poor prognosis and optimal adjuvant therapy remains controversial.2 3 Treatment options include chemotherapy, radiotherapy, or chemotherapy in combination with radiation. Trastuzumab may be added to chemotherapy for those with HER2neu overexpressing serous endometrial cancers.4

A major limitation of randomized trials on adjuvant treatment of advanced endometrial cancer is the heterogeneity of the population studied with varying stage, grade, histology, and molecular sub-type. Staging of endometrial cancer was redefined in 2009, with the change that positive cytology alone is no longer classified as stage IIIA disease. While cytology alone does not affect International Federation of Gynecology and Obstetrics (FIGO) staging, the National Comprehensive Cancer Network still recommends routine cytology collection and result reporting as positive cytology is an established adverse risk factor.5 Among patients with advanced stage disease, those with FIGO stage IIIA grade 1 endometrioid endometrial cancer have favorable outcomes compared with those with nodal metastasis or intra-abdominal disease.6 7 Survival rates differ based on histologic sub-type and extent of disease among patients with stage IIIA endometrial cancer, and those with low-grade endometrioid carcinomas may require a unique treatment approach.7 8 There is a paucity of data regarding management of stage IIIA grade 1 endometrioid cancers. In two recent pivotal trials, stage IIIA disease accounted for just 12.6% and 20.5% of patients enrolled, limiting generalizability of study results for patients in this under-represented sub-group.9 10

In several randomized trials comparing adjuvant radiotherapy with chemotherapy in patients with advanced stage endometrial cancer, chemotherapy was associated with improved progression-free survival and reduced risk of death.11–13 A sub-group evaluation of those with stage IIIA disease in the GOG 122 randomized trial found that chemotherapy use was associated with improved survival.12 The efficacy of combined adjuvant therapy versus monotherapy has been investigated in prospective clinical trials, including patients with stage IIIA disease, with mixed results. Most trials found no difference in overall survival with chemoradiation compared with monotherapy, with conflicting results regarding progression-free survival.10 14 15 In PORTEC-3 specifically, failure-free survival was superior in the chemoradiation group compared with the radiation group, but there was no difference in overall survival between groups. Conversely, in the post-hoc survival analysis of PORTEC-3 assessing survival after recurrence and updated survival, the use of chemoradiation significantly improved overall and failure-free survival compared with radiation alone, with the greatest benefit in women with stage III disease.9 16

Although a prospective clinical trial on patients with stage IIIA grade 1 endometrial cancer would be ideal, it is unlikely to be feasible as this is a relatively rare sub-group of disease. Therefore, the purpose of this study was to examine treatment and outcomes for patients with stage IIIA grade 1 endometrial cancer using the National Cancer Database.

## METHODS

### Data Source

The National Cancer Database registry is a joint project of the American Cancer Society and Commission on Cancer of the American College of Surgeons. It contains nationwide data on oncology outcomes for over 1500 accredited programs. The database includes de-identified information on patient demographics, disease characteristics, grade, stage, treatment, and survival outcomes for nearly 70% of all newly diagnosed malignancies in the USA annually.17

### Study Sample

A retrospective cohort study was performed from January 2004 to December 2016 on the National Cancer Database Uterus Corpus and Uterus, not otherwise specified files. We selected patients with endometrioid adenocarcinoma histology and stage IIIA (FIGO 2009 staging) disease. Site-specific coding fields (FIGO stage, pathologic tumor node metastasis stage, and tumor extension) were used to define disease stage. We excluded cases which would have been classified as stage IIIA by positive cytology alone, based on FIGO 2009 staging guidelines, patients undergoing non-standard surgical treatment, neoadjuvant/intra-operative treatment, and patients with in situ disease or lacking tissue confirmation of disease. The National Cancer Database does not code a field for adnexal tumor extension. In our effort to avoid including cases that would have been labeled as stage IIIA by positive cytology alone in older staging schemes, it is possible that some cases with adnexal involvement were excluded if they were coded as positive cytology alone. Lastly, we limited the study cohort to patients with grade 1 disease and excluded those with missing data on receipt of adjuvant treatment. Inclusion and exclusion criteria are summarized in Figure 1.

Figure 1

Identification of study cohort. CT, computed tomography; LVSI, lymphovascular space invasion; RT, radiation.

### Study Data

Adjuvant treatment modality was categorized as none, chemotherapy alone, radiation alone, or chemoradiation (chemotherapy and vaginal brachytherapy and/or external beam radiation). Demographic information including patient age (<50, 50–69, ≥70), race (white, black, Asian/Pacific, other/unknown), ethnicity (non-Hispanic/unknown vs Hispanic), insurance status (private, Medicare, Medicaid, uninsured), Charlson–Deyo comorbidity score (0, 1, 2, ≥3), income (<$38 000,$38 000–$47 999,$48 000–$62 999, >$63 000), facility type (academic vs non-academic) and geographic region (Northeast, South, Midwest/Central, West) were collected. Clinical information included stage (IIIA), grade (1), tumor size (<2 cm, ≥2 cm, unknown), year of diagnosis (2004–2009, 2010–2016), lymph node dissection status (yes, no), and lymphovascular space invasion status (+/−/unknown). The primary outcome was overall survival in months.

### Statistical Analysis

Demographics were compared by adjuvant treatment cohort using a t-test for continuous variables and χ2 test for categorical dichotomous variables. We used Kaplan–Meier curves, log-rank test, and multivariable Cox proportional hazard models adjusting for age, comorbidity, facility type, tumor size, year of diagnosis, lymph node dissection status, and lymphovascular space invasion status to estimate overall survival based on adjuvant treatment modality. Of note, lymphovascular space invasion status was not reported prior to 2010, and tumor size was missing for approximately 25% of observations. We included unknown categories for these variables in the multivariable model to avoid excluding a significant portion of observations. We included a separate model for known lymphovascular space invasion status.

We followed STROBE guidelines for observational research. A p value <0.05 was considered significant, and all analyses were conducted with STATA V.15.1 (Statacorp, College Station, Texas, USA). This study was granted exemption by the Duke Investigational Review Board.

## RESULTS

Table 1

Patient characteristics

Compared with non-Medicare patients, those with Medicare were less likely to receive chemotherapy or chemoradiation but more likely to receive radiation alone. These differences were non-significant when compared with non-Medicare patients ≥65 years of age. There were no differences in receipt of adjuvant therapy based on income, high school graduate rate, rurality, or facility type. Those in the southern region were less likely to receive chemotherapy (51.1% vs 61.8%; p=0.001).

In more recent years there was a significant lower utilization of no adjuvant treatment (26.9% in 2004–2009 vs 19.1% in 2010–2016; p=0.002) and higher utilization of chemotherapy (45.4% in 2004–2009 vs 69.2% in 2010–2016; p<0.001). There was no difference in the rate of radiation use (50.0% in 2004–2009 vs 53.8% in 2010–2016; p=0.21). There was no difference in vaginal brachytherapy or external beam radiation use. Patients with lymphovascular space invasion were more likely to receive chemoradiation than those without invasion (52.5% vs 37.6%; p=0.007).

The 5-year overall survival rate was 83.0% (95% CI 80.1% to 85.6%) (Table 2). Overall survival was significantly longer with chemotherapy versus no chemotherapy (90.3% vs 74.0%, log-rank p<0.001), and chemoradiation versus no chemoradiation (89.7% vs 79.8%, log-rank p<0.001), but there was no difference in survival with receipt of radiation (Figure 2).

Figure 2

(A) Kaplan–Meier curves for overall survival by adjuvant treatment among women with stage IIIA, grade 1 endometrioid endometrial carcinoma in the National Cancer Database, 2004–2015. (B) Kaplan–Meier curves for overall survival by receipt of chemotherapy among women with stage IIIA, grade 1 endometrioid endometrial carcinoma in the National Cancer Database, 2004–2015.

Table 2

Log-rank survival data

In the primary multivariable Cox model, treatment with chemotherapy was associated with significantly improved overall survival compared with no adjuvant therapy (HR 0.49 (95% CI 0.31 to 0.79); p=0.003; Table 3). In contrast, there was no survival benefit from radiation alone versus no treatment, and no benefit of adding radiation to chemotherapy compared with chemotherapy alone. We assessed survival differences by radiation type independently but found no significant results. Among control variables, older age, increasing comorbidity count, and lymphovascular space invasion status were significant negative predictors of survival. Race, ethnicity, income, education level, and rurality were not associated with survival differences, and inclusion or exclusion in the multivariable model did not impact HR estimates for the characteristics included in the presented model.

Table 3

Multivariable logistic regression

### Supplemental material

Lymphovascular space invasion was a significant risk factor in the primary model (HR 2.09; p=0.006), but has only been reported since 2010. Therefore, we considered models stratified by lymphovascular space invasion status. Those without lymphovascular space invasion who received chemotherapy had a significantly better adjusted overall survival compared with no treatment (HR 0.27 (95% CI 0.09 to 0.80); p=0.019). Those with lymphovascular invasion (n=124/507) had improved unadjusted overall survival with chemoradiation compared with monotherapy by log-rank test (5-year overall survival 91.2% vs 76.7% for chemotherapy alone and 27.3% for radiation alone; p<0.001). However, after adjusting for age and comorbidity in a Cox regression model, there was no difference in overall survival with chemoradiation (HR 0.25 (95% CI 0.05 to 1.41); p=0.12) among this sub-set of patients.

## DISCUSSION

### Summary of Main Results

In this retrospective cohort analysis of 1120 women with stage IIIA grade 1 endometrioid adenocarcinoma from the National Cancer Database, treatment with chemotherapy was associated with significantly improved overall survival compared with no adjuvant therapy. There was no significant survival benefit from radiation alone versus no treatment, and no benefit of combined chemoradiation compared with chemotherapy alone. Our findings are similar to GOG 258 results and align with current National Comprehensive Cancer Network guidelines recommending systemic therapy with or without external beam radiation or vaginal brachytherapy for patients with stage III endometrial cancer.5 Older age, increasing comorbidity count, and lymphovascular space invasion were statistically significant negative predictors of survival.

### Results in the Context of Published Literature

Our results support the sub-group findings of prospective randomized clinical trials that have shown a survival benefit to adjuvant chemotherapy. For example, in a post-hoc subgroup analysis of GOG 122, which demonstrated an overall survival advantage for chemotherapy over radiation, patients with stage IIIA endometrioid disease were more likely to have improved overall survival with chemotherapy alone compared with those with more advanced stage disease.12 While the Japanese Gynecologic and Oncology Group trial found no significant survival difference with chemotherapy versus radiation, a sub-group analysis that included stage IIIA disease found significantly improved survival with chemotherapy compared with pelvic radiation.13

Several prior studies have used the National Cancer Database to evaluate management patterns and associations with survival in patients with stage III endometrial cancers.23–26 Four studies found that adjuvant chemoradiation was associated with improved survival compared with monotherapy. Among these, Lester-Coll et al concluded this association was specifically in patients with stage IIIA disease, while those with grade 1 disease benefited from chemotherapy alone.24 Syeda et al presented a sub-group analysis of women with stage IIIA tumors and found similarly decreased mortality with chemoradiation compared with chemotherapy.25 Our study is unique from these prior studies as we focus our analysis on a specific stage, grade, and histologic sub-type of endometrial cancer together, as well as examining the use of chemotherapy, radiation, and chemoradiation.

### Implications for Practice and Future Research

Further studies are necessary to determine which patients benefit from combined therapy. Our findings support consideration for systemic therapy, consistent with the National Comprehensive Cancer Network guidelines. While this study includes patients in the USA, further studies are needed to determine if our findings are reproducible internationally. Prospective studies are needed to determine if patients with lymphovascular space invasion benefit from additional therapy. Future trials are evaluating the clinical role of integrated clinicopathologic and molecular risk profiles to direct adjuvant treatment in patients with endometrial cancer.

## Conclusions

In patients in the National Cancer Database with stage IIIA grade 1 endometrioid endometrial cancer, chemotherapy alone was associated with overall survival benefit. However, future studies are needed to identify which adjuvant treatment modality may improve survival in this specific patient population.

## Data availability statement

Data are available upon reasonable request. Data are available upon reasonable request. The data from our study were extracted from the National Cancer Database repository with permission. The data are available upon request to Mary Katherine Montes de Oca at mm765@duke.edu.

• ## Supplementary Data

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.