Sage Journals: Discover world-class research

Abstract

Introduction

Low-grade endometrial stromal sarcoma (LG-ESS) is a rare malignant mesenchymal neoplasm for which there is no consensus regarding the role of radiotherapy in treatment. This study aimed to evaluate the prognostic significance of external beam radiotherapy (EBRT) using data from the Surveillance, Epidemiology, and End Results (SEER) database.

Methods

This retrospective study evaluated the role of EBRT in 1254 patients with LG-ESS using SEER data (2000-2021). Propensity score matching (PSM) was applied to compare outcomes between EBRT and non-EBRT groups. Cox and competing risk models assessed overall survival (OS) and cancer-specific survival (CSS).

Results

EBRT showed no survival benefit in either unmatched or matched cohorts. Post-PSM, OS (HR = 1.21, 95% CI 0.61-2.39) and CSS (HR = 1.75, 95% CI 0.69-4.43) remained unaffected by EBRT (P > 0.05). Lymphadenectomy and bilateral salpingectomy and oophorectomy (BSO) also demonstrated no significant associations with survival outcomes. Key prognostic factors included older age, larger tumor size, advanced stage, and chemotherapy use, all linked to poorer OS and CSS.

Conclusion

EBRT, lymphadenectomy, and BSO do not offer significant survival benefits for patients with LG-ESS. Prognosis was independently influenced by age, tumor size, stage, and chemotherapy use (associated with poorer outcomes). These findings support a more individualized, risk-adapted approach to LG-ESS management and highlight the need for prospective studies to define the optimal role of comprehensive treatment strategies.

Plain Language Summary

What is this study about?This study looked at whether radiation therapy, lymph node surgery, or ovary removal improves survival for patients with a rare type of uterine cancer called low-grade endometrial stromal sarcoma (LG-ESS). Because this cancer is uncommon, there’s limited evidence to guide treatment decisions. What did the researchers do?Using a U.S. national cancer database (2000–2021), the researchers analyzed data from 1,254 LG-ESS patients. They compared survival outcomes between those who received radiation therapy, lymph node surgery, or ovary removal and those who did not. Advanced statistical methods were used to balance differences between patient groups. What did they find? Radiation therapy, lymph node surgery, and ovary removal did not improve survival for LG-ESS patients. Older age, larger tumor size, advanced cancer stage, and chemotherapy use (linked to worse survival) were stronger predictors of outcomes. What does this mean for patients?The results suggest that treatment for LG-ESS should focus on the cancer’s specific features (like size and stage) rather than routinely using radiation therapy. Avoiding unnecessary radiation could reduce side effects and improve quality of life. However, the reasons behind chemotherapy’s link to poorer survival need further study. Why is this important?This is the largest study ever conducted on LG-ESS, providing much-needed evidence to help doctors and patients choose personalized treatments. More research is needed to develop better guidelines for managing this rare cancer.

Keywords

low-grade endometrial stromal sarcoma external beam radiotherapy surveillance epidemiology and end results (SEER) database lymphadenectomy bilateral salpingectomy and oophorectomy

Introduction

Uterine sarcoma (US), a rare and aggressive malignancy of the uterine corpus, accounts for approximately 3% to 7% of all uterine neoplasms.^1,2 The 2020 WHO classification system delineates 4 subtypes of US: endometrial stromal nodules (ESN), low-grade endometrial stromal sarcoma (LG-ESS), high-grade endometrial stromal sarcoma (HG-ESS), and undifferentiated uterine sarcoma (UUS).³ LG-ESS is the second most prevalent uterine sarcoma and a rare malignancy of mesenchymal origin.⁴ LG-ESS typically manifests with abnormal uterine bleeding, pelvic pain, or dyspareunia.⁵ Additional clinical manifestations may be caused by uterine masses, adnexal extension, or distant metastases to lymph nodes or lungs.⁶

Currently, treatment decisions for LG-ESS are primarily guided by disease stage and hormone receptor status. In general, surgery remains the cornerstone of treatment, typically involving total hysterectomy with or without bilateral salpingectomy and oophorectomy (BSO).^7,8 Other therapeutic options include radiotherapy, chemotherapy, and hormonal therapy. Nevertheless, the efficacy of various therapeutic modalities, including radiotherapy, ovarian preservation, and lymph node dissection, remains inconclusive.^9,10 The role of lymph node dissection and the prognostic implications of lymph node metastasis are yet to be fully elucidated.^11,12 The utility of BSO in young, premenopausal patients remains a subject of ongoing debate. Due to the rarity of LG-ESS, data on the efficacy of radiotherapy (RT) in prospective randomized controlled trials remain limited.¹³ Current guidelines advocate for radiotherapy, despite inconsistent findings reported in prior studies of LG-ESS. A United States-based study involving 3650 endometrial stromal sarcoma (ESS) patients reported a significant recurrence-free survival benefit of pelvic adjuvant radiotherapy for localized disease. This benefit was evident not only in the overall cohort but also in specific subgroups of ESS patients.¹⁴ Conversely, another large study involving 1010 ESS patients found no significant impact of pelvic radiotherapy on overall survival.¹⁵ Moreover, the role of radiotherapy in unresectable or medically inoperable LG-ESS is rarely addressed in current literature.¹⁶ Thus, it is imperative to determine the efficacy of radiation therapy in LG-ESS patients and to identify risk factors for adverse outcomes and optimal therapeutic strategies.

In this context, we performed a retrospective analysis utilizing the Surveillance, Epidemiology, and End Results (SEER) database of the National Cancer Institute. In this study, we identified LG-ESS patients and associated variables, evaluated the impact of external beam radiotherapy (EBRT) on survival outcomes using propensity score matching, and applied Cox proportional hazards regression and competing risks analysis to assess prognostic factors. The primary endpoints of the study included overall survival (OS) and cancer-specific survival (CSS). Furthermore, the study investigated the impact of lymph node dissection on survival outcomes in LG-ESS patients and assessed the necessity of BSO.

Patients and Methods

Study Population and definitions

Data were extracted from the SEER database using SEER∗STAT software (version 8.4.4). The SEER database is publicly accessible via its official website. This study analyzed patients diagnosed with LG-ESS from 2000 to 2021. LG-ESS was defined as grade I or II tumors, characterized as well-differentiated or moderately differentiated.^17,18 Inclusion criteria included: (1) age >18 years; (2) pathological confirmation of corpus uteri/uterus cancer, classified under ICD-O-3 codes 8930/3, 8931/3, or 8935/3^15,19; (3) diagnosis of a single primary malignant neoplasm; and (4) complete follow-up data, including causes of death. Exclusion criteria included: (1) incomplete data on tumor stage (based on FIGO staging criteria), surgical procedure, or lymphadenectomy; and (2) death or follow-up duration of less than 1 month after diagnosis. The SEER database is publicly available and de-identified, so ethics committee approval was not required. Written informed consent was not required for this retrospective study.

Cohort Definition and Variable Recode

The variables obtained from the SEER database encompassed age at diagnosis, race and ethnicity, histologic subtype, year of diagnosis, marital status (at diagnosis), lymph node status, tumor size, stage at diagnosis, pathologic grade, surgery types, radiation recode, chemotherapy, survival months, vital status recode, cause of death, and cause-specific death classification. Surgical procedures were classified as hysterectomy with bilateral salpingectomy and oophorectomy (Hys + BSO); hysterectomy without bilateral salpingectomy and oophorectomy (Hys-BSO); hysterectomy, not otherwise specified (Hys NOS); and local excision, exenteration, or other forms of surgery (L/E/OTH). Tumor staging was recoded in accordance with the 2009 FIGO criteria: localized stage corresponded to FIGO stage I; regional stage encompassed FIGO stages II and III; and distant stage aligned with FIGO stage IV.¹⁹ The reporting of this study conforms to the RECORD guidelines.²⁰

Statistical Analysis

The primary endpoints of this study were OS and CSS. OS was defined as the time from diagnosis to death from any cause, or to the date of the last follow-up in November 2022. CSS was defined as the interval between the diagnosis of LG-ESS and death attributable to that malignancy.

Data are presented as the median (range) or n (%). Fisher’s exact test was employed for categorical variables to compare clinicopathological characteristics between groups, whereas the two-sample t test or Mann-Whitney U-test was used for continuous variables, as appropriate. Systematic differences in baseline covariates and cohort selection bias are major methodological challenges of retrospective studies.²¹ Propensity score matching (PSM) was performed with a tolerance of 0.05 to balance baseline characteristics between the external beam radiotherapy (EBRT) and no radiotherapy (NO RT) groups. OS was assessed using Kaplan-Meier survival curves and the Cox log-rank test. Multivariate survival analyses were performed using the Cox proportional hazards regression model. To identify risk factors for CSS, accounting for non-cancer deaths as competing risks, both univariate and multivariate analyses were conducted using competing risks regression (Fine and Gray method). Univariate and multivariate analyses were conducted for all variables, and those with a P-value <0.05 in both analyses were identified as independent risk factors. Statistical significance was set at a P-value <0.05. Data were analyzed using the R programming language and software environment (https://www.r-project.org/).

Results

Demographic and Clinical Characteristics

From 2000 to 2021, a total of 1254 patients with LG-ESS were identified from the SEER database. Of these, 10.12% underwent EBRT. The demographic and clinical characteristics of these patients with LG-ESS are summarized in Table 1. Approximately one-third of the cohort (411/1254) underwent lymphadenectomy, with 7.84% (33/421) demonstrating positive lymph node metastasis. Hys + BSO was the predominant surgical approach in 70.49% (884/1254) of patients with LG-ESS. Before PSM, demographic and clinical characteristics differed significantly between the No RT and EBRT groups. These included follow-up duration, ethnicity, year of diagnosis, tumor stage, lymph node involvement, chemotherapy use, and lymphadenectomy. A study cohort of 330 patients was identified through a 1:2 PSM approach. Following PSM, the cohorts were comparable in demographic and clinical characteristics (P > 0.05), minimizing potential confounders in subsequent analyses.

Table 1.

Demographic and clinical characteristics of patients with LG‐ESS

Characteristic	Before propensity score-matched (N=1254)			After propensity score-matched (N=330)
Characteristic	No RT, No. (%)	EBRT, No. (%)	P value	No RT, No. (%)	EBRT, No. (%)	P value
Total	1127	127		220	110
Follow-up,
median(range), mo	98 (2-263)	137(4-262)	0.012	108 (2-263)	140 (4-262)	0.075
Age at diagnosis, yr			0.861			0.105
<=50	719 (63.80%)	83 (65.35%)		121 (55.00%)	72 (65.45%)
>50, <=70	344 (30.52%)	36 (28.35%)		86 (39.09%)	30 (27.27%)
>70	64 (5.68%)	8 (6.30%)		13 (5.91%)	8 (7.27%)
Marital status			0.525			0.745
Single/unmarried	225 (19.96%)	26 (20.47%)		41 (18.64%)	19 (17.27%)
Married	667 (59.18%)	80 (62.99%)		136 (61.82%)	72 (65.45%)
Divorced/separated	124 (11.00%)	13 (10.24%)		22 (10.00%)	12 (10.91%)
Widowed	55 (4.88%)	6 (4.72%)		14 (6.36%)	6 (5.45%)
Unknown	56 (4.97%)	2 (1.57%)		7 (3.18%)	1 (0.91%)
Ethnicity			0.023			0.06
Black	133 (11.80%)	26 (20.47%)		26 (11.82%)	18 (16.36%)
White	854 (75.78%)	82 (64.57%)		136 (61.82%)	77 (70.00%)
Other	128 (11.36%)	18 (14.17%)		52 (23.64%)	14 (12.73%)
Unknown	12 (1.06%)	1 (0.79%)		6 (2.73%)	1 (0.91%)
Year of diagnosis			<0.001			0.072
<=2004	226 (20.05%)	36 (28.35%)		53 (24.09%)	29 (26.36%)
>2004, <=2009	231 (20.50%)	41 (32.28%)		43 (19.55%)	34 (30.91%)
>2009, <=2015	340 (30.17%)	29 (22.83%)		73 (33.18%)	26 (23.64%)
>2015	330 (29.28%)	21 (16.54%)		51 (23.18%)	21 (19.09%)
Tumor size			0.463			0.405
< 5 cm	318 (28.22%)	30 (23.62%)		39 (17.73%)	28 (25.45%)
5‐10 cm	345 (30.61%)	36 (28.35%)		73 (33.18%)	34 (30.91%)
≥10 cm	184 (16.33%)	26 (20.47%)		49 (22.27%)	20 (18.18%)
Not specified	280 (24.84%)	35 (27.56%)		59 (26.82%)	28 (25.45%)
Tumor stage*			<0.001			0.331
Localized	833 (73.91%)	54 (42.52%)		89 (40.45%)	53 (48.18%)
Regional	178 (15.79%)	47 (37.01%)		83 (37.73%)	39 (35.45%)
Distant	116 (10.29%)	26 (20.47%)		48 (21.82%)	18 (16.36%)
Grade			0.169			0.176
Well differentiated	409 (36.29%)	52 (40.94%)		73 (33.18%)	48 (43.64%)
Moderately differentiated	532 (47.20%)	62 (48.82%)		117 (53.18%)	50 (45.45%)
Unknown	186 (16.50%)	13 (10.24%)		30 (13.64%)	12 (10.91%)
Lymph node metastasis			<0.001			0.114
No	337 (29.90%)	51 (40.16%)		79 (35.91%)	43 (39.09%)
Yes	22 (1.95%)	11 (8.66%)		6 (2.73%)	9 (8.18%)
No nodes were examined	760 (67.44%)	64 (50.39%)		133 (60.45%)	57 (51.82%)
Unknown	8 (0.71%)	1 (0.79%)		2 (0.91%)	1 (0.91%)
Surgery			0.115			0.246
No surgery	35 (3.11%)	4 (3.15%)		13 (5.91%)	3 (2.73%)
Hys+BSO	793 (70.36%)	91 (71.65%)		145 (65.91%)	77 (70.00%)
Hys-BSO	124 (11.00%)	11 (8.66%)		19 (8.64%)	11 (10.00%)
Hys NOS	52 (4.61%)	12 (9.45%)		14 (6.36%)	11 (10.00%)
L/E/OTH	123 (10.91%)	9 (7.09%)		29 (13.18%)	8 (7.27%)
Chemotherapy			<0.001			0.769
No	1094 (97.07%)	110 (86.61%)		204 (92.73%)	101 (91.82%)
Yes	33 (2.93%)	17 (13.39%)		16 (7.27%)	9 (8.18%)
Lymphadenectomy			<0.001			0.095
No	776 (68.86%)	67 (52.76%)		139 (63.18%)	59 (53.64%)
Yes	351 (31.14%)	60 (47.24%)		81 (36.82%)	51 (46.36%)

LG‐ESS, low‐grade endometrial stromal sarcoma; RT, radiotherapy; EBRT, external beam radiotherapy; Hys + BSO, hysterectomy with bilateral salpingectomy and ovariectomy oophorectomy; Hys – BSO, hysterectomy without bilateral salpingectomy and ovariectomy oophorectomy; Hys NOS, hysterectomy, not specified; L/E/OTH, local surgery, exenteration, and other surgery types.

*The International Federation of Gynecology and Obstetrics (FIGO) criteria‐based tumor staging was used to determine tumor stage.

Prognostic Factors for CSS and OS

Competing risks regression analysis (Fine and Gray method) identified age, tumor size, tumor stage, chemotherapy use, and ethnicity as independent prognostic factors for CSS in patients with LG-ESS (Table 2). Patients older than 50 years had worse CSS, with those over 70 exhibiting the highest risk (hazard ratio [HR] = 3.95, P < 0.001). Tumor size between 5 and 10 cm was significantly associated with poorer CSS (HR = 2.89, P = 0.004), while tumors ≥10 cm showed a trend toward worse survival (HR = 2.14, P = 0.054). Both regional (HR = 2.76, P = 0.001) and distant (HR = 6.23, P < 0.001) stages were significantly correlated with reduced CSS. Chemotherapy use was strongly linked to worse CSS (HR = 8.47, P < 0.001). Compared to Black patients, White patients had a lower risk of cancer-specific death (HR = 0.50, P = 0.012). Patients in the ‘Other’ ethnic group showed a non-significant trend toward improved survival (HR = 0.63, P = 0.292). In contrast, EBRT did not improve CSS (HR = 1.25, 95% confidence interval [CI]: 0.69-2.27, P = 0.459).

Table 2.

Multivariate analyses on variables for predicting CSS and OS in LG‐ESS patients

Characteristic	N=1254 (%)	Cancer-specific survival		Overall survival
Characteristic	N=1254 (%)	Hazard Ratio (95% CI)	P value	Hazard Ratio (95% CI)	P value
Age at diagnosis, yr
<=50	802 (63.96%)	1 (reference)		1 (reference)
>50, <=70	380 (30.30%)	2.03 (1.20, 3.44)	0.016	2.77 (1.85, 4.15)	<0.001
>70	72 (5.74%)	3.95 (1.79, 8.71)	0.000	7.91 (4.62, 13.54)	<0.001
Marital status
Single/unmarried	251 (20.02%)	1 (reference)		1 (reference)
Married	747 (59.57%)	0.82 (0.48, 1.41)	0.485	0.74 (0.48, 1.14)	0.172
Divorced/separated	137 (10.93%)	0.37 (0.11, 1.28)	0.118	0.97 (0.52, 1.81)	0.936
Widowed	61 (4.86%)	2.17 (0.95, 4.94)	0.071	2.04 (1.13, 3.69)	0.018
Unknown	58 (4.63%)	0.96 (0.32, 2.89)	0.923	0.84 (0.36, 1.98)	0.695
Ethnicity
Black	159 (12.68%)	1 (reference)		1 (reference)
White	936 (74.64%)	0.50 (0.29, 0.86)	0.012	0.53 (0.35, 0.80)	0.003
Other	146 (11.64%)	0.63 (0.28, 1.46)	0.292	0.46 (0.22, 0.93)	0.030
Unknown	13 (1.04%)	NA	NA	NA	NA
Year of diagnosis
<=2004	262 (20.89%)	1 (reference)		1 (reference)
>2004, <=2009	272 (21.69%)	0.93 (0.50, 1.74)	0.800	0.96 (0.62, 1.49)	0.864
>2009, <=2015	369 (29.43%)	1.30 (0.71, 2.37)	0.398	1.09 (0.69, 1.72)	0.719
>2015	351 (27.99%)	0.58 (0.24, 1.41)	0.230	0.53 (0.26, 1.09)	0.084
Tumor size
< 5 cm	348 (27.75%)	1 (reference)		1 (reference)
5‐10 cm	381 (30.38%)	2.89 (1.39, 6.02)	0.004	2.15 (1.29, 3.60)	0.004
≥10 cm	210 (16.75%)	2.14 (0.99, 4.63)	0.054	1.57 (0.89, 2.78)	0.117
Not specified	315 (25.12%)	1.32 (0.59, 2.93)	0.461	1.49 (0.88, 2.54)	0.142
Tumor stage*
Localized	887 (70.73%)	1 (reference)		1 (reference)
Regional	225 (17.94%)	2.76 (1.51, 5.05)	0.001	2.07 (1.35, 3.18)	0.001
Distant	142 (11.32%)	6.23 (3.37, 11.51)	<0.001	3.66 (2.24, 5.98)	<0.001
Grade
Well differentiated	461 (36.76%)	1 (reference)		1 (reference)
Moderately differentiated	594 (47.37%)	1.40 (0.85, 2.31)	0.186	1.09 (0.76, 1.57)	0.641
Unknown	199 (15.87%)	1.36 (0.62, 3.00)	0.430	1.64 (0.95, 2.81)	0.075
Lymph node metastasis
No	388 (30.94%)	1 (reference)		1 (reference)
Yes	33 (2.63%)	2.28 (0.88, 5.88)	0.084	1.39 (0.61, 3.17)	0.436
No nodes were examined	824 (65.71%)	1.13 (0.24, 5.43)	0.885	0.95 (0.31, 2.87)	0.925
Unknown	9 (0.72%)	NA	NA	1.32 (0.20, 8.46)	0.772
Surgery
No surgery	39 (3.11%)	1 (reference)		1 (reference)
Hys+BSO	884 (70.49%)	0.71 (0.30, 1.67)	0.478	1.14 (0.55, 2.35)	0.722
Hys-BSO	135 (10.77%)	0.95 (0.25, 3.58)	0.924	1.26 (0.45, 3.52)	0.659
Hys NOS	64 (5.10%)	0.53 (0.16, 1.78)	0.297	0.52 (0.18, 1.46)	0.212
L/E/OTH	132 (10.53%)	0.69 (0.23, 2.04)	0.486	1.03 (0.42, 2.51)	0.951
EBRT
No	1127 (89.87%)	1 (reference)		1 (reference)
Yes	127 (10.13%)	1.25 (0.69, 2.27)	0.459	1.22 (0.76, 1.95)	0.414
Chemotherapy
No	1204 (96.01%)	1 (reference)		1 (reference)
Yes	50 (3.99%)	8.47 (4.81, 14.94)	<0.001	5.89 (3.57, 9.72)	<0.001
Lymphadenectomy
No	843 (67.22%)	1 (reference)		1 (reference)
Yes	411 (32.78%)	0.79 (0.17, 3.77)	0.750	0.80 (0.26, 2.44)	0.695

LG‐ESS, low‐grade endometrial stromal sarcoma; EBRT, external beam radiotherapy; Hys + BSO, hysterectomy with bilateral salpingectomy and ovariectomy oophorectomy; Hys – BSO, hysterectomy without bilateral salpingectomy and ovariectomy oophorectomy; Hys NOS, hysterectomy, not specified; L/E/OTH, local surgery, exenteration, and other surgery types.

*The International Federation of Gynecology and Obstetrics (FIGO) criteria‐based tumor staging was used to determine tumor stage.

Multivariate Cox regression analysis revealed that age, marital status, ethnicity, tumor size, tumor stage, and chemotherapy use were independent prognostic factors for OS in LG-ESS patients (Table 2). Older age was significantly associated with poorer OS. Patients over 50 years had worse survival, with those older than 70 years at the highest risk (HR = 7.91, P < 0.001). Widowed status was associated with worse OS (HR = 2.04, P = 0.018), while marital status (married, divorced, or single) did not significantly impact survival. White patients had a significantly lower risk of overall mortality compared to Black patients (HR = 0.53, P = 0.003), and patients of other ethnicities also demonstrated improved survival (HR = 0.46, P = 0.030). Tumors between 5 and 10 cm were associated with worse OS (HR = 2.15, P = 0.004), while tumors ≥10 cm did not reach statistical significance (HR = 1.57, P = 0.117). Both regional (HR = 2.07, P = 0.001) and distant (HR = 3.66, P < 0.001) stages were significantly associated with poorer OS. Chemotherapy use was strongly linked to worse OS (HR = 5.89, P < 0.001). Similarly, EBRT did not confer an improvement in OS (HR = 1.22, 95% CI: 0.76-1.95, P = 0.414).

Prognostic Factors for CSS and OS after PSM

To address potential confounders, PSM was used. As indicated by the multivariate analyses (Table 3), EBRT did not improve CSS or OS in patients with LG-ESS. Specifically, EBRT was not associated with a statistically significant improvement in CSS (HR = 1.75, 95% CI: 0.69-4.43, P = 0.228) or OS (HR = 1.21, 95% CI: 0.61-2.39, P = 0.580), suggesting that its use does not confer a survival advantage in this cohort after adjustment for confounding factors. In the PSM cohort (n = 330), patients who received EBRT did not exhibit significantly higher 10-year CSS (89.95% vs 90.83%, P = 0.793; Figure 1A) or 10-year OS (88.61% vs 87.96%, P = 0.840; Figure 1B).

Table 3.

Multivariate analysis of CSS and OS variables after predicting PSM in LG-ESS patients

Characteristic	N=330	Cancer-specific survival		Overall survival
Characteristic	N=330	Hazard Ratio (95% CI)	P Value	Hazard Ratio (95% CI)	P value
Age at diagnosis, yr
<=50	193 (58.48%)	1 (reference)		1 (reference)
>50, <=70	116 (35.15%)	2.62 (0.86, 7.92)	0.096	3.50 (1.56, 7.88)	0.002
>70	21 (6.36%)	14.60 (3.40, 62.66)	<0.001	8.79 (2.84, 27.19)	0.000
Marital status
Single/unmarried	60 (18.18%)	1 (reference)		1 (reference)
Married	208 (63.03%)	1.09 (0.39, 3.04)	0.860	0.96 (0.44, 2.08)	0.921
Divorced/separated	34 (10.30%)	NA	NA	0.24 (0.05, 1.19)	0.080
Widowed	20 (6.06%)	1.55 (0.34, 7.02)	0.559	2.04 (0.67, 6.22)	0.210
Unknown	8 (2.42%)	4.91 (0.31, 77.87)	0.221	1.43 (0.13, 15.89)	0.771
Ethnicity
Black	44 (13.33%)	1 (reference)		1 (reference)
White	213 (64.55%)	0.39 (0.13, 1.17)	0.094	0.43 (0.19, 0.99)	0.048
Other	66 (20.00%)	0.07 (0.01, 0.75)	0.027	0.15 (0.04, 0.63)	0.010
Unknown	7 (2.12%)	NA	NA	NA	NA
Year of diagnosis
<=2004	82 (24.85%)	1 (reference)		1 (reference)
>2004, <=2009	77 (23.33%)	0.92 (0.21, 4.01)	0.895	0.97 (0.40, 2.32)	0.939
>2009, <=2015	99 (30.00%)	1.18 (0.31, 4.45)	0.817	0.98 (0.36, 2.64)	0.967
>2015	72 (21.82%)	0.80 (0.13, 4.99)	0.782	0.59 (0.15, 2.35)	0.457
Tumor size
< 5 cm	67 (20.30%)	1 (reference)		1 (reference)
5‐10 cm	107 (32.42%)	6.70 (1.27, 35.22)	0.024	2.55 (0.89, 7.30)	0.082
≥10 cm	69 (20.91%)	8.18 (1.52, 43.93)	0.012	3.30 (1.15, 9.45)	0.026
Not specified	87 (26.36%)	2.16 (0.36, 12.79)	0.394	1.64 (0.60, 4.48)	0.338
Tumor stage*
Localized	142 (43.03%)	1 (reference)		1 (reference)
Regional	122 (36.97%)	11.94 (2.65, 53.88)	0.000	2.91 (1.26, 6.71)	0.012
Distant	66 (20.00%)	12.88 (2.83, 58.49)	0.000	3.10 (1.15, 8.32)	0.025
Grade
Well differentiated	121 (36.67%)	1 (reference)		1 (reference)
Moderately differentiated	167 (50.61%)	0.73 (0.30, 1.79)	0.478	0.70 (0.36, 1.36)	0.287
Unknown	42 (12.73%)	1.15 (0.28, 4.77)	0.854	1.73 (0.67, 4.44)	0.255
Lymph node metastasis
No	122 (36.97%)	1 (reference)		1 (reference)
Yes	15 (4.55%)	2.73 (0.51, 14.58)	0.231	1.39 (0.34, 5.76)	0.649
No nodes were examined	190 (57.58%)	1.18 (0.18, 7.87)	0.850	1.12 (0.34, 3.63)	0.853
Unknown	3 (0.91%)	NA	NA	NA	NA
Surgery
No surgery	16 (4.85%)	1 (reference)		1 (reference)
Hys+BSO	222 (67.27%)	0.64 (0.12, 3.40)	0.586	1.06 (0.28, 4.08)	0.930
Hys-BSO	30 (9.09%)	5.99 (0.56, 64.55)	0.123	2.39 (0.38, 15.21)	0.355
Hys NOS	25 (7.58%)	0.63 (0.09, 4.34)	0.600	0.66 (0.13, 3.34)	0.619
L/E/OTH	37 (11.21%)	0.41 (0.05, 3.58)	0.436	0.35 (0.05, 2.31)	0.277
EBRT
No	220 (66.67%)	1 (reference)		1 (reference)
Yes	110 (33.33%)	1.75 (0.69, 4.43)	0.228	1.21 (0.61, 2.39)	0.580
Chemotherapy
No	305 (92.42%)	1 (reference)		1 (reference)
Yes	25 (7.58%)	14.90 (4.24, 52.37)	<0.001	4.73 (1.78, 12.57)	0.002
Lymphadenectomy
No	198 (60.00%)	1 (reference)		1 (reference)
Yes	132 (40.00%)	0.51 (0.07, 3.83)	0.486	0.76 (0.22, 2.58)	0.656

*The International Federation of Gynecology and Obstetrics (FIGO) criteria‐based tumor staging was used to determine tumor stage.

Figure 1.

Survival Curves according to EBRT Following Propensity Score Matching: (A) CSS Curves; (B) OS Curves

Subgroup Analysis

Subgroup analyses for both CSS (Figure 2A) and OS (Figure 2B) consistently demonstrated that EBRT was not associated with a statistically significant survival benefit across any of the predefined LG-ESS subgroups. Regardless of patient stratification by age, stage, tumor size, or surgical intervention, no subgroup showed a clear advantage for EBRT in terms of either CSS or OS. These findings suggest that the role of EBRT may be limited in improving survival outcomes for LG-ESS.

Figure 2.

Forest Plot of Subgroup Analysis for Cancer-specific Survival (A) and Overall Survival (B). Hys + BSO, Hysterectomy With Bilateral Salpingectomy and Ovariectomy Oophorectomy; Hys-BSO, Hysterectomy Without Bilateral Salpingectomy and Ovariectomy Oophorectomy; Hys NOS, Hysterectomy not Specified; EBRT, External Beam Radiotherapy; RT, Radiotherapy

Lymphadenectomy

In a multivariate analysis evaluating CSS and OS, the impact of lymphadenectomy on patient outcomes was assessed (Table 2). For CSS, no statistically significant difference was observed between patients who underwent lymphadenectomy and those who did not (HR = 0.79, 95% CI: 0.17-3.77, P = 0.750). Similarly, for OS, the HR was 0.80 (95% CI: 0.26-2.44), with a P-value of 0.695, further suggesting that lymphadenectomy does not significantly influence survival outcomes. In the unmatched cohort, Kaplan–Meier survival analysis confirmed these findings, with no improvement in CSS (Figure 3A, P = 0.75) or OS (Figure 3B, P = 0.54) among patients undergoing lymphadenectomy.

Figure 3.

Survival Curves according to Lymphadenectomy before Propensity Score Matching: (A) CSS Curves; (B) OS Curves

Hysterectomy with or without Bilateral Salpingectomy and Oophorectomy

To evaluate the prognostic significance of BSO, we compared two cohorts from the overall study population: the Hysterectomy with BSO cohort (Hys + BSO) and the Hysterectomy without BSO cohort (Hys-BSO). The demographic and clinical characteristics of these cohorts, both before and after PSM, are detailed in Supplemental Table S1. Prior to matching, the Hys + BSO cohort included 884 patients, while the Hys-BSO cohort comprised 135 patients. After matching, the sample sizes were reduced to 240 and 120 patients, respectively, resulting in balanced cohorts with respect to baseline characteristics.

Multivariable analysis, presented in Supplemental Table S2 and S3, was performed to assess the impact of BSO on CSS and OS. Before matching, the hazard ratios for CSS and OS between the Hys + BSO and Hys-BSO cohorts were 0.71 (95% CI: 0.26-1.95, P = 0.469) and 0.83 (95% CI: 0.41-1.71, P = 0.620), respectively, indicating no significant survival benefit associated with BSO. After matching, these findings remained consistent, with hazard ratios for CSS and OS of 0.81 (95% CI: 0.22-2.91, P = 0.737) and 1.05 (95% CI: 0.43-2.57, P = 0.917), respectively. In addition, Kaplan–Meier survival curves demonstrated no statistically significant improvement in CSS (Figuer 4A, P = 0.94) or OS (Figure 4B, P = 0.37) between patients with and without BSO in the matched cohort. In summary, both the PSM analysis and multivariable analyses suggest that the addition of BSO to hysterectomy does not significantly improve either CSS or OS in patients with LG-ESS.

Figure 4.

Survival Curves according to Ovariectomy Oophorectomy Following Propensity Score Matching: (A) CSS Curves; (B) OS Curves. Hys + BSO, Hysterectomy With Bilateral Salpingectomy and Ovariectomy Oophorectomy; Hys-BSO, Hysterectomy Without Bilateral Salpingectomy and Ovariectomy Oophorectomy

Discussion

LG-ESS is an uncommon and heterogeneous uterine malignancy, accounting for a substantial proportion of endometrial stromal sarcomas.²² Due to its rarity, much of the existing knowledge regarding LG-ESS has been extrapolated from studies on uterine sarcomas as a whole, resulting in limited data specifically addressing this distinct subtype.^23,24 Total hysterectomy with BSO constitutes the primary treatment for LG-ESS. However, the utility of other interventions such as radiotherapy, lymphadenectomy, and ovarian preservation remains a topic of active debate in clinical practice.

In this study, we evaluated the prognostic impact of EBRT in patients with LG-ESS using data from the SEER database. To mitigate the influence of confounding variables, we employed PSM to balance baseline characteristics between patients treated with EBRT and NO RT. Our analysis revealed that EBRT was not associated with significant improvements in CSS or OS. After adjustment for confounders, the HR for CSS and OS were 1.25 (95% CI, 0.69-2.27) and 1.22 (95% CI, 0.76-1.95), respectively. These findings align with prior large-scale studies, which similarly reported no survival benefit from radiotherapy in patients with ESS.^15,25

The role of radiotherapy in the management of LG-ESS remains controversial. While some studies have suggested that radiotherapy, particularly pelvic radiotherapy with or without brachytherapy, offers benefits in local disease control, evidence for its impact on survival is inconsistent. In a large epidemiologic study of 3650 patients conducted in the United States, pelvic adjuvant radiotherapy was associated with significantly improved local recurrence-free survival, including among a subset of patients with ESS (n = 312).¹⁴ Similarly, a SEER analysis of 2677 uterine sarcoma cases reported a survival benefit of adjuvant radiotherapy in women with stage II and stage III/IV disease.²⁶ However, these studies did not stratify outcomes by tumor grade. It has been postulated that radiotherapy may be more effective in HG-ESS than in LG-ESS, where its utility appears limited.²⁷ A retrospective study of 152 patients with stage I–II LG-ESS found that adjuvant radiotherapy significantly reduced pelvic recurrence in patients with stage IB–IIB disease and improved survival in subgroup analyses.²⁸ Conversely, other studies have reported no significant survival benefit of radiotherapy in LG-ESS, consistent with findings by Barney et al,¹⁵ who analyzed 1010 ESS patients from the SEER database and observed no survival advantage from adjuvant radiotherapy across any FIGO stage or tumor grade. Kim et al. similarly noted no impact of radiotherapy on survival in LG-ESS.²⁵ Another study further demonstrated a lack of efficacy of adjuvant radiation therapy in patients with stage I–II ESS.²⁹ Thus, although radiotherapy may improve local control in LG-ESS, its role in enhancing CSS and OS remains uncertain. Furthermore, its potential long-term adverse effects, including secondary malignancies and tissue fibrosis, must be weighed against the typically favorable prognosis of patients with low-grade disease.

Our analysis found no significant survival benefit associated with lymphadenectomy in patients with LG-ESS. Only 7.84% of patients in our cohort exhibited lymph node metastasis, and multivariate analysis demonstrated that lymphadenectomy had no impact on CSS or OS. These findings are consistent with previous studies, which similarly failed to show a survival advantage from lymphadenectomy in LG-ESS patients.^9,30-32 For instance, Zhang et al. reported that patients who underwent lymphadenectomy did not experience a statistically significant improvement in progression-free survival or OS, even after applying PSM.³² While some authors have suggested that lymphadenectomy may improve staging accuracy and potentially benefit high-risk patients, its clinical utility in LG-ESS remains uncertain. The relatively low incidence of lymph node metastasis in LG-ESS may partially explain the limited impact of lymphadenectomy on survival outcomes. Therefore, the decision to perform lymphadenectomy should be individualized, considering each patient’s clinical characteristics and risk factors.

In terms of BSO, our study found that the inclusion of BSO in the surgical treatment of LG-ESS did not confer a significant survival advantage. This finding is consistent with several studies which have questioned the necessity of BSO in patients with LG-ESS, particularly in premenopausal women.¹¹ Li et al¹¹ reported similar results, showing that ovarian preservation had no impact on recurrence or overall survival. Given the generally favorable prognosis of LG-ESS, particularly in patients with early-stage disease, ovarian preservation may be a reasonable option for young women, with careful consideration of the tumor’s hormone receptor status, histology, and the patient’s fertility desires.³³ Thus, BSO may not be required for all patients with LG-ESS, particularly those in whom fertility preservation is desired.

Age, tumor size, and tumor stage were identified as independent prognostic factors for both CSS and OS in our study, consistent with the existing literature. Advanced age and larger tumor size are well-established negative prognostic indicators in a variety of cancers, including ESS.^19,27,34 Similarly, tumor stage was a significant determinant of survival in our cohort, with regional and distant stages associated with worse outcomes, a finding corroborated by previous research on ESS.³⁰ The impact of chemotherapy on survival, as observed in our study, aligns with previous studies suggesting that chemotherapy may be associated with poorer survival in patients with LG-ESS. This is likely due to the inherent aggressiveness of the disease in cases requiring chemotherapy.

Our study benefits from a large sample size and the use of a well-established database with extensive follow-up data. The application of PSM helped mitigate the effects of confounding variables, ensuring that our comparisons between the EBRT and NO RT groups were more robust. However, there are some limitations inherent in the retrospective nature of the study. The SEER database lacks information on certain factors, such as residual disease status and surgical margins, which may influence survival outcomes. Moreover, the absence of detailed information on radiotherapy techniques—including dose, radiation fields, and modality—poses a challenge to interpreting treatment efficacy, especially considering the long study period (2000-2021) during which radiotherapy practices have significantly evolved. The database also does not capture treatment-related adverse effects (e.g., radiation toxicities) or detailed systemic therapy data, such as the use and type of hormonal therapy. Finally, while our study is one of the largest to examine the role of radiotherapy in LG-ESS, prospective, randomized controlled trials are needed to definitively establish the efficacy of radiotherapy and other treatments in this rare malignancy.

Conclusion

Our analysis of the SEER database suggests that EBRT, lymphadenectomy, and BSO do not significantly improve survival outcomes in patients with LG-ESS. The prognosis of these patients is more strongly influenced by clinical factors such as age, tumor size, and stage. Given the lack of survival benefit associated with radiotherapy and the potential risks of radiation therapy, further research is needed to refine treatment strategies for LG-ESS, particularly to determine which patients may benefit from radiation therapy. For now, the primary treatment remains surgical, with careful consideration of fertility preservation in younger women and individualized management for those with advanced disease.

Supplemental Material

Supplemental Material - Prognostic Role of Radiotherapy in Low-Grade Endometrial Stromal Sarcoma: A SEER-Based Study

Supplemental Material for Prognostic Role of Radiotherapy in Low-Grade Endometrial Stromal Sarcoma: A SEER-Based Study by Huimin Jian, Jinju Guo, Wenxin Zhao, Wei Liu, Yuan Xiang, and Xia Wang in Cancer Control.

Supplemental Material

Supplemental Material - Prognostic Role of Radiotherapy in Low-Grade Endometrial Stromal Sarcoma: A SEER-Based Study

Supplemental Material

Supplemental Material - Prognostic Role of Radiotherapy in Low-Grade Endometrial Stromal Sarcoma: A SEER-Based Study

Footnotes

ORCID iD

Xia Wang

Ethical Considerations

This study relied on publicly available data from the SEER database. Ethics approval was waived as this study analyzed de-identified, publicly available data from the SEER database. Informed consent was not required from SEER registered cases, and the authors obtained Limited-Use Data Agreements from SEER.

Authors’ Contributions

HMJ and XW conducted the statistical analysis and made significant contributions to the drafting and writing of the manuscript. WL, WXZ, JJG and YX played an integral role by offering crucial insights and actively engaging in the revision process of the manuscript. XW served as the principal investigator, overseeing project coordination and management. All authors have thoroughly reviewed and approved the final version of the manuscript.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data files utilized in this study were directly obtained from the SEER website ().

Disclamier

We express our profound gratitude to the SEER*Stat team for their invaluable contribution to this study through the provision of patient data.

Supplemental Material

Supplemental material for this article is available online.

Appendix

References

Prat

Mbatani

. Uterine sarcomas. Int J Gynaecol Obstet. 2015;131(S2):S105-S110. doi:10.1016/j.ijgo.2015.06.006

Hosh

Antar

Nazzal

Warda

Gibreel

Refky

. Uterine sarcoma: analysis of 13,089 cases based on surveillance, Epidemiology, and end results database. Int J Gynecol Cancer. 2016;26(6):1098-1104. doi:10.1097/igc.0000000000000720

Akaev

Yeoh

Rahimi

. Update on endometrial stromal tumours of the uterus. Diagnostics. 2021;11(3):429. doi:10.3390/diagnostics11030429

Tropé

Abeler

Kristensen

. Diagnosis and treatment of sarcoma of the uterus. A review. Acta Oncologica. 2012;51(6):694-705. doi:10.3109/0284186x.2012.689111

Smith

Jansen

Miller

, et al. Primary characteristics and outcomes of newly diagnosed low-grade endometrial stromal sarcoma. Int J Gynecol Cancer. 2022;32(7):882-890. doi:10.1136/ijgc-2022-003383

dos Santos

Garg

Diaz

, et al. Incidence of lymph node and adnexal metastasis in endometrial stromal sarcoma. Gynecol Oncol. 2011;121(2):319-322. doi:10.1016/j.ygyno.2010.12.363

Denschlag

Ackermann

Battista

, et al. Sarcoma of the uterus. Guideline of the DGGG, OEGGG and SGGG (S2k-level, AWMF registry No. 015/074, april 2021). Geburtshilfe Frauenheilkd. 2022;82(12):1337-1367. doi:10.1055/a-1897-5124

Prat

. FIGO staging for uterine sarcomas. Int J Gynaecol Obstet. 2009;104(3):177-178. doi:10.1016/j.ijgo.2008.12.008

Bai

Yang

Cao

, et al. Ovary and uterus-sparing procedures for low-grade endometrial stromal sarcoma: a retrospective study of 153 cases. Gynecol Oncol. 2014;132(3):654-660. doi:10.1016/j.ygyno.2013.12.032

10.

Leath

Huh

Hyde

, et al. A multi-institutional review of outcomes of endometrial stromal sarcoma. Gynecol Oncol. 2007;105(3):630-634. doi:10.1016/j.ygyno.2007.01.031

11.

Giuntoli

Drake

, et al. Ovarian preservation in stage I low-grade endometrial stromal sarcomas. Obstet Gynecol. 2005;106(6):1304-1308. doi:10.1097/01.AOG.0000185511.91694.1e

12.

Gadducci

Cosio

Romanini

Genazzani

. The management of patients with uterine sarcoma: a debated clinical challenge. Crit Rev Oncol Hematol. 2008;65(2):129-142. doi:10.1016/j.critrevonc.2007.06.011

13.

Benson

Miah

. Uterine sarcoma – current perspectives. Int J Womens Health. 2017;9:597-606.

14.

Sampath

Schultheiss

Ryu

Wong

JYC

. The role of adjuvant radiation in uterine sarcomas. Int J Radiat Oncol Biol Phys. 2010;76(3):728-734. doi:10.1016/j.ijrobp.2009.02.077

15.

Barney

Tward

Skidmore

Gaffney

. Does radiotherapy or lymphadenectomy improve survival in endometrial stromal sarcoma? Int J Gynecol Cancer. 2009;19(7):1232-1238. doi:10.1111/IGC.0b013e3181b33c9a

16.

Weitmann

Knocke

Kucera

Pötter

. Radiation therapy in the treatment of endometrial stromal sarcoma. International Journal of Radiation Oncology*Biology*Physics. 2001;49(3):739-748. doi:10.1016/s0360-3016(00)01369-9

17.

Seagle

Shilpi

Buchanan

Goodman

Shahabi

. Low-grade and high-grade endometrial stromal sarcoma: a National Cancer Database study. Gynecol Oncol. 2017;146(2):254-262. doi:10.1016/j.ygyno.2017.05.036

18.

Garg

Shah

Toy

Bryant

Kumar

Morris

. Stage IA vs. IB endometrial stromal sarcoma: does the new staging system predict survival? Gynecol Oncol. 2010;118(1):8-13. doi:10.1016/j.ygyno.2010.04.008

19.

Zhang

, et al. A nomogram for predicting overall survival in patients with low-grade endometrial stromal sarcoma: a population-based analysis. Cancer Commun. 2020;40(7):301-312. doi:10.1002/cac2.12067

20.

Benchimol

Smeeth

Guttmann

, et al. The REporting of Studies Conducted Using Observational Routinely-Collected Health Data (RECORD) Statement:1549-1676. Electronic.

21.

Wang

Chen

, et al. Effect of Marital Status on the Survival of Patients With Adenocarcinoma of the Esophagogastric Junction: A Population-Based, Propensity-Matched Study. Cancer Control. 2021;28(Electronic):1526-2359. doi:10.1177/10732748211066309

22.

Thiel

Halmen

. Low-grade endometrial stromal sarcoma - a review. Oncol Res Treat. 2018;41(11):687-692. doi:10.1159/000494225

23.

L-Y

Zhang

H-T

J-S

X-G

S-K

. Treatment options in stage I endometrial stromal sarcoma: a retrospective analysis of 53 cases. Gynecol Oncol. 2008;108(2):306-311. doi:10.1016/j.ygyno.2007.10.023

24.

Rizzo

Pantaleo

Saponara

Nannini

. Current status of the adjuvant therapy in uterine sarcoma: a literature review. World J Clin Cases. 2019;7(14):1753-1763. doi:10.12998/wjcc.v7.i14.1753

25.

Kim

Lee

Choi

, et al. Low-grade endometrial stromal sarcoma: a single center's experience with 22 cases. Int J Gynecol Cancer. 2008;18(5):1084-1089. doi:10.1111/j.1525-1438.2007.01159.x

26.

Brooks

Zhan

Cote

Baquet

. Surveillance, Epidemiology, and End Results analysis of 2677 cases of uterine sarcoma 1989–1999. Gynecol Oncol. 2004;93(1):204-208. doi:10.1016/j.ygyno.2003.12.029

27.

Khan

Soomar

Asghari

Ahmed

Moosajee

. Prognostic factors, oncological treatment and outcomes of uterine sarcoma: 10 years’ clinical experience from a tertiary care center in Pakistan. BMC Cancer. 2023;23(1):510. doi:10.1186/s12885-023-11000-3

28.

Wang

Sun

Miao

Hou

Zhang

. Adjuvant radiotherapy improved survival in stage I to II low-grade endometrial stromal sarcoma: a retrospective study of 152 cases. Front Oncol. 2021;10:608152. doi:10.3389/fonc.2020.608152

29.

Harter

El-Khalfaoui

Heitz

du Bois

. Operative and conservative treatment of uterine sarcomas. Geburtshilfe Frauenheilkd. 2014;74(03):267-270. doi:10.1055/s-0034-1368204

30.

Chan

Kawar

Shin

, et al. Endometrial stromal sarcoma: a population-based analysis. Br J Cancer. 2008;99(8):1210-1215. doi:10.1038/sj.bjc.6604527

31.

Shah

Bryant

Kumar

Ali-Fehmi

Malone

Morris

. Lymphadenectomy and ovarian preservation in low-grade endometrial stromal sarcoma. Obstet Gynecol. 2008;112(5):1102-1108. doi:10.1097/AOG.0b013e31818aa89a

32.

Zhang

Wang

, et al. Long‐term impact of lymphadenectomies in patients with low‐grade, early‐stage uterine endometrial stroma sarcoma. J Obstet Gynaecol Res. 2020;46(4):654-662. doi:10.1111/jog.14198

33.

Roy

Musa

Taylor

Huang

. Uterine sarcomas: how to navigate an ever-growing list of subtypes. Am Soc Clin Oncol Educ Book. 2022;42(42):910-919. doi:10.1200/edbk_350955

34.

Ayhan

Toptas

, et al. Low-grade endometrial stromal sarcoma: a Turkish uterine sarcoma group study analyzing prognostic factors and disease outcomes. Gynecol Oncol. 2021;160(3):674-680. doi:10.1016/j.ygyno.2020.12.017

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.15 MB

0.10 MB

0.13 MB