Uterine Cancer Incidence Trends and Five-Year Relative Survival by Race/Ethnicity and Histology Among Women Under 50 Years

Abstract

Background:

Uterine cancers diagnosed before age 50 years are increasing in the U.S., but changes in clinical characteristics and survival over time across racial/ethnic groups have not been previously described.

Objectives:

To investigate age-adjusted, hysterectomy corrected incidence rates and trends, and five-year relative survival rates of uterine cancer in women aged <50 years, overall and stratified by race/ethnicity and histology.

Study Design:

We included microscopically confirmed uterine cancer cases (diagnosed 2000–2019) in women aged 20–49 years from the Surveillance, Epidemiology, and End Results Program (SEER 22). Age-adjusted incidence and 5-year relative survival rates, and 95% confidence intervals were computed using SEER*Stat and compared across time periods (2000–2009 and 2010–2019). Incidence rates were adjusted for hysterectomy prevalence using Behavioral Risk Factor Surveillance System data, and trends were computed using the Joinpoint regression program.

Results:

We included 57,128 uterine cancer cases. The incidence of uterine cancer increased from 10.1 per 100,000 in 2000–2009 to 12.0 per 100,000 in 2010–2019, increasing at an annual rate of 1.7%/year for the entire period. Rising trends were more pronounced among women <40 years (3.0%/year and 3.3%/year in 20–29 and 30–39 years, respectively) than in those 40–49 years (1.3%/year), and among underrepresented racial/ethnic groups (Hispanic 2.8%/year, Non-Hispanic, [NH]-Black 2.7%, NH-Asian/Pacific Islander [PI] 2.1%) than in NH-White (0.9%/year). Recent (2010–2019) incidence rates were highest for endometrioid (9.6 per 100,000), followed by sarcomas (1.2), and non-endometrioid subtypes (0.9). Rates increased significantly for endometrioid subtypes at 1.9%/year from 2000–2019. Recent endometrioid and non-endometrioid rates were highest in NH-Native American/Alaska Native [NA/AN] (15.2 and 1.4 per 100,000), followed by Hispanic (10.9 and 1.0), NH-Asian/PI (10.2 and 0.9), NH-White (9.4 and 0.8), and lowest in NH-Black women (6.4 and 0.8). Sarcoma rates were highest in NH-Black women (1.8 per 100,000). The five-year relative survival remained unchanged over time for women with endometrioid (from 93.4% in 2000–2009 to 93.9% in 2010–2019, p≥0.05) and non-endometrioid subtypes (from 73.2% to 73.2%, p≥0.05) but decreased for women with sarcoma from 69.8% (2000–2009) to 66.4% (2010–2019, p<0.05).

Conclusions:

Uterine cancer incidence rates in women <50 years have increased from 2000 to 2019 while survival has remained relatively unchanged. Incidence trends can be primarily attributed to increasing rates of cancers with endometrioid histology, with the greatest increases observed among NH-Black, Hispanic, and NH-Asian/PI. Sarcomas, while much rarer, were the second most common type of uterine cancer among women <50 years and have poor prognosis and apparent decreasing survival over time. Rising rates of uterine cancer and the distinct epidemiologic patterns among women <50 years highlight the need for effective prevention and early detection strategies for uterine cancer in this age group.

Introduction

Uterine corpus cancer, hereafter referred to as uterine cancer, is the most common gynecologic cancer diagnosed in the United States (U.S.), with an estimated 67,880 new cases and 13,250 deaths estimated occurring in the year 2024.¹ Uterine cancer incidence and mortality rates have been increasing over the past two decades, primarily attributed to rising rates of non-endometrioid subtypes which are less common, but much more clinically aggressive than tumors with endometrioid histology.^2,3 These subtypes are increasing among all women aged 30–79 years, and much more commonly diagnosed in Non-Hispanic African American/Black women (referred to as NH-Black).^2,4,5 While the majority of these uterine cancers are diagnosed in postmenopausal women,⁶ approximately 25%¹ of cases occur in premenopausal women. Studies have suggested that among women <50 years, incidence rates are rising at a more rapid rate compared with older age groups.^1,7,8 However less is known about the clinical features of these cancers (i.e., histology) and how these may vary by race and ethnicity.^8–10

Understanding epidemiologic patterns and trends of uterine cancer incidence among women <50 years could reveal insights into potential age-specific differences in exposures and risk factors which may vary across different populations. Moreover, a better understanding of survival outcomes among women <50 years-old with uterine cancer could help to inform clinical practice including the use of fertility-sparing treatments¹¹ and the potential benefits of early detection approaches in this age group.

Here, we estimated age-adjusted, hysterectomy-corrected incidence rates and trends of uterine cancer, along with five-year relative survival rates of women aged 20–49 years (hereafter referred to as <50 years) by histologic subtype and race/ethnicity.

Methods

Study Population

Data Source

We utilized data from the Surveillance, Epidemiology, and End Results (SEER 22), drawing on data from 22 U.S. population-based cancer registries spanning 2000 to 2019.^12,13 The SEER 22 database represents 48% of the U.S. population from 17 states (California, Connecticut, Hawaii, Iowa, New Mexico, Washington, Utah, Georgia, Alaska, Kentucky, Louisiana, New Jersey, Idaho, New York, Massachusetts, Illinois, and Texas).¹² All data are de-identified and publicly available, thus the study was exempt from institutional review board approval and the need for informed consent.

Case Selection

The eligibility criteria are represented in Figure 1. We included women aged 20–49 years with newly diagnosed, microscopically confirmed uterine cancer between 2000 and 2019. Cases were restricted to first primary uterine cancers, and we excluded cases with unknown race (n=2,223), and cases identified using only autopsy/death certificates (n=23). Additionally, we excluded cases diagnosed in 2020 due to potential biases associated with COVID-19 pandemic.¹⁴

Figure 1.

The flow chart of incident uterine cancer case selection among U.S. women reported in the SEER 22 combined cancer registry, between the ages 20–49 years.^{12, b}

Abbreviations: Uteri NOS= Uteri Not Otherwise Specified, SEER (Surveillance. Epidemiology, and End Results).;^a Some cases have one or more overlapping exclusion criteria; thus, the total number excluded is less than the sum of individual categories of exclusion. Case selections were restricted to first matching record for each person.;^b Only the first record for each person which matches the selection criteria were included in the analyses.

Demographic and Clinical Characteristics

We categorized age at diagnosis (20–29, 30–39, and 40–49 years), and included race/ethnicity¹⁵ (NH-White, NH-Black, NH-Asian/ PI and American Indian/Alaska Natives, hereafter referred to as Native American/Alaska Native [NA/AN]).⁹ Small sample size of NA/AN women with uterine cancer (n=453), precluded us from estimating incidence trends and survival data. We grouped histologic subtypes as endometrioid, non-endometrioid, sarcomas and other cancers as shown in Table S1. To rectify potential misclassification of adenocarcinoma, NOS, Not Otherwise Specified which was more commonly diagnosed in earlier time periods, we reclassified these cases as previously described (supplemental methods).² We also included stage at diagnosis (localized, regional, and distant), which was not available for cases diagnosed before 2004, and for Massachusetts.¹⁶; Blanks (n=9,883), and not staged/unknown (n=2,229) were set as missing for a analyses evaluating rates by stage.

Statistical Analysis

Age-Adjusted Uterine Cancer Incidence Rates and Trends

Age-adjusted incidence rates and 95% confidence intervals (CIs)¹² were calculated for calendar periods 2000–2009 and 2010–2019 using SEER*Stat software version 8.4.2 (® National Cancer Institute). Incidence trends for 2000 to 2019 were presented using annual percent changes (APC) or average APC or using the Joinpoint regression program, version 5.0.2 (® National Cancer Institute).; The Average APC equals to APC when there is a single, consistent trend line. We utilized the non-parametric Empirical Quantile Confidence Interval method within the Joinpoint regression program to determine statistical significance of observed trends. Statistical significance was determined based on whether the confidence interval for each joinpoint excluded zero, with a corresponding p-value of ≤0.05.¹⁷ Age-standardized incidence rates to the 2000 U.S. population were computed, and hysterectomy-corrected incidence rates were estimated by removing the proportion of women with a hysterectomy from the denominators. The Behavioral Risk Factor Surveillance System (BRFSS)¹⁸ data were used to calculate survey-weighted hysterectomy prevalence accounting for calendar year, age, race and U.S. region.² Per SEER recommendations, rates were not shown when case count was ≤16.¹⁹

Survival Rates of Uterine Cancer

We included uterine cancer cases diagnosed between 2000 and 2018, to calculate the age-standardized five-year relative survival rates with 95% CI were calculated using SEER*Stat survival session. We assessed survival through 2019, allowing for at least one year of follow-up for those cases diagnosed in 2018. The median follow-up time for all cases was 8 years; median follow-up for cases diagnosed between 2000 and 2009 was 14 years, and 5 years for cases diagnosed from 2010 to 2018. Expected survival rates considered age-specific all-cause mortality rates within the reference population, utilizing actuarial and Ederer II methods. P-values for group differences were calculated using the z-test scores generated by SEER*Stat. These computations account for survival variations due to racial/ethnic composition, geographic area, and socioeconomic status.²⁰ We excluded registries/cases (Illinois and Massachusetts registries¹³) with missing survival data.

All statistical analyses were two sided and p values <0.05 were considered significant.

Results

Overall Uterine Cancer Incidence Rates and Trends

We included 57,128 new uterine cancer cases (diagnosed: 2000–2019) (Figure 1). The overall incidence of uterine cancer increased from 10.1 per 100,000 women in 2000–2009 to 12.0 in 2010–2019 (Table 1), with a significant annual increase of 1.74%/year (95% CI 1.53, 2.00) from 2000–2019 among all women (Figure 2). In both periods, incidence rates were highest among women aged 40–49 years (22.0 per 100,000 women in 2000–2009 and 24.9 per 100,000 women in 2010–2019) (Table 1). However, rates increased most rapidly among women <40 years of age (average annual increase of 2.95%, 95% CI 2.13, 4.03 in 20–29 years and 3.27%, 95% CI 2.67, 3.82 in 30–39 years vs. 1.27%, 95% CI 1.01, 1.54 in 40–49 years) (Figure 2).

Table 1.

Age-adjusted, Hysterectomy Corrected Incidence Rates (2000–2019) of Microscopically Confirmed Uterine Cancer Overall and Across Demographic and Clinical Characteristics, Among U.S. Women Aged 20 to 49 Years, SEER 22

Calendar Period	2000–2009		2010–2019
	Cases, N (%)	IR (95% CI)	Cases, N (%)	IR (95% CI)
Overall Incidence Rate	26,279	10.1 (9.9, 10.2)	30,849	12.0 (11.7, 12.1)
Age at Diagnosis (Years)
20–29	889 (3.4)	0.9 (0.8, 1.0)	1,402 (4.5)	1.3 (1.2, 1.4)
30–39	5,785 (22.0)	6.0 (5.8, 6.1)	7,930 (25.7)	8.0 (7.8, 8.1)
40–49	19,605 (74.6)	22.0 (21.6, 22.2)	21,517 (69.8)	24.9 (24.5, 25.1)
Race/Ethnicity
NH-White	16,321 (62.1)	10.6 (10.2, 10.6)	14,992 (48.6)	11.5 (11.1, 11.5)
NH-Black	2,203 (8.4)	7.0 (7.0, 7.3)	2,956 (9.6)	9.3 (9.0, 9.7)
NH-Asian/PI	2,322 (8.8)	10.5 (10.0, 10.9)	3,556 (11.5)	12.3 (11.9, 12.7)
NH-NA/AN	186 (0.7)	11.5 (9.9, 13.2)	267 (0.9)	17.7 (15.8, 20.2)
Hispanic	5,247 (20.0)	10.2 (10.0, 10.6)	9,078 (29.4)	13.6 (13.4, 13.9)
Histology
Endometrioid a	20,753 (79.0)	8.0 (8.0, 8.1)	24,732 (80.2)	9.6 (9.5, 9.7)
Non-Endometrioid a	1,870 (7.1)	0.7 (0.7, 0.7)	2,249 (7.3)	0.9 (0.9, 0.9)
Sarcoma	2,883 (11.0)	1.1 (1.0, 1.1)	3,022 (9.8)	1.2 (1.2, 1.3)
Other	775 (3.0)	0.3 (0.2, 0.3)	847 (2.8)	0.3 (0.3, 0.3)
Stage at Diagnosis c
Localized	11,808 (75.5)	4.5 (4.4, 4.6)	22,236 (75.7)	8.6 (8.4, 8.7)
Regional	2,804 (17.9)	1.1 (1.0, 1.2)	4,850 (16.5)	2.0 (1.8, 2.0)
Distant	1,037 (6.6)	0.4 (0.4, 0.5)	2,279 (7.8)	0.9 (0.8, 0.9)

Notes. Percentages were calculated using non-missing data. All incidence rates (IRs) were age adjusted to the 2000 US standard population and presented per 100,000 population. All IRs were also corrected for hysterectomy, using survey weighted hysterectomy prevalence data from the Behavioral Risk Factor Surveillance System (BRFSS). IR estimates were not presented for those strata with insufficient sample size (≤16)²² denoted as NS.

Abbreviations: Adeno, Adenocarcinoma; CI, Confidence Interval; IR, Incidence Rates; NA/AN, Native American/Alaska Native; NH, Non-Hispanic; NOS, Not Otherwise Specified; NS, Not Sufficient; PI, Pacific Islander.

^aIncidence rates of endometrioid and non-endometrioid cancers reflect values after re-classification of Adenocarcinoma, NOS (8140/3) into endometrioid and non-endometrioid histologic subtypes.

^cStage information was not available prior to 2004, resulting in n=9,883 missing datapoints for the period between 2000–2009.

Figure 2.

Age-adjusted incident rate trends of uterine cancer with and without hysterectomy correction among women aged 20–49 years, overall and by age groups.

Hysterectomy corrected and uncorrected Incidence Rates (IRs) for the Overall Study Population (A), and age-specific IRs (40–49 years, 30–39 years, and 20–29 years). Average annual percentages changes (AAPCs) presented for both hysterectomy corrected and uncorrected trend lines representing.; All AAPCs presented for the entire study period between 2000–2019.; Statistically significant trend p <0.05 indicated with (*).

Uterine Cancer Incidence Rates and Trends by Race/Ethnicity and Clinical Characteristics

Incidence rates of uterine cancer overall increased across all racial/ethnic groups with the highest incidence occurring among NH-NA/AN and Hispanic women (17.7 and 13.6 per 100,000 women, respectively), and the lowest among NH-Black women (9.3 per 100,000 women) in 2010–2019 (Table 1). Incidence trends rose more rapidly among underrepresented racial/ethnic groups compared to NH-White women: Average APCs of Hispanic 2.79%/year (95% CI 2.35, 3.37), NH-Black 2.69%/year (95% CI 2.19, 3.25), NH-Asian/PI 2.07%/year (95% CI 1.54, 2.49) and NH-White 0.91%/year (95% CI 0.59, 1.22) (Figure 3). Incidence rates among NH-Asian/PI and Hispanic women increased more rapidly in recent years (APC: 6.97%/year, 95% CI 2.84, 12.54 since 2016, and 3.76%/year, 95% CI 3.15, 6.38 since 2007, respectively). Please note that trend analyses for NH-NA/AN women were not shown due to the small sample size.

Figure 3.

Age-adjusted hysterectomy corrected incident rate trends of uterine cancer among women aged 20–49 years, by race/ethnicity.

Incidence rate trends lines and average percent change (APC) are shown. AAPCs represent the average rate of change for the entire study period between 2000–2019, which takes the average of different trend lines with annual percent changes (APCs).; Statistically significant trend p <0.05 indicated with (*).

In the overall population, the incidence rates of endometrioid subtypes significantly increased over time (Average APC 1.93%/year, 95% CI 1.69, 2.18) and sarcoma trends significantly increased after 2002 (APC 0.92%/year, 95% CI 0.43, 2.81) (Figure 4). In contrast, incidence rates of non-endometrioid subtypes increased significantly from 2000–2015 (APC 3.48%/year, 95% CI 2.32, 5.37) followed by a significant decline from 2016–2019 (APC −9.02%/year, 95% CI −20.37, −2.48) (Figure 4). In both time periods, the incidence rates of endometrioid subtypes were highest (8.0 per 100,000 in 2000–2009 and 9.6 per 100,000 in 2010–2019) followed by sarcomas (1.1 and 1.2 per 100,000, respectively) and non-endometrioid subtypes (0.7 and 0.9 per 100,000, respectively) (Table 1).

Figure 4.

Age-adjusted hysterectomy corrected incident rate trends of uterine cancer among women aged 20–49 years, by histologic subtype.

Incidence rate trends lines and average annual percent change (AAPC) are shown. AAPCs represent the average rate of change for the entire study period between 2000–2019, which takes the average of different trend lines with annual percent changes (APCs).; Statistically significant trend p <0.05 indicated with (*).

Endometrioid cancer trends were rising among all race/ethnic groups, with NH-Black women experiencing the steepest rise in incidence (Average APC 3.85%/year, 95% CI 3.16, 4.65), followed by Hispanic (3.17%/year, 95% CI 2.75, 3.63), NH-Asian/PI (2.31%/year, 95% CI 1.47, 2.90), and NH-White women (1.01%/year, 95% CI 0.59, 1.42) (Figure S1-S2). Across both time periods, endometrioid incidence rates were highest in NH-NA/AN women (9.2 per 100,000 in 2000–2009 and 15.2 per 100,000 in 2010–2019) and lowest in NH-Black women (4.3 and 6.4 per 100,000, respectively) (Table 2). Incidence rates of sarcomas significantly increased among Hispanic women (Average APC 1.51%, 95% CI 0.48, 2.78), but trends were stable among all other racial/ethnic groups (Figure S1 and S3). Across both time periods, NH-Black women had the highest rates of sarcomas (1.7 per 100,000 in 2000–2009 and 1.8 per 100,000 in 2010–2019), followed by Hispanic women (1.2 and 1.3 per 100,000, respectively) (Table 2). Incidence trends of non-endometrioid subtypes remained consistent over time for NH-Black and NH-Asian/PI women. However, among NH-White women, we observed a significant increase in the rates of non-endometrioid subtypes from 2000 to 2009 (APC 6.96%/year, 95% CI 4.19, 11.96) followed by a significant decline from 2009 to 2019 (APC −4.09%/year, 95% CI −7.49, −1.95) (Figure S4). In contrast, rates among Hispanic women increased significantly over time (Average APC 3.48%/year, 95% CI 2.10, 5.28).

Table 2.

Age-adjusted Hysterectomy Corrected Incidence Rates (2000–2019) of Microscopically Confirmed Corpus and Uterus, NOS, Not Otherwise Specified Cancer by Race/Ethnicity, Histologic Subtype and Stage, Among U.S. Women Aged 20 to 49 Years, SEER 22

	NH-White		NH-Black		NH-Asian/PI		NH-NA/AN		Hispanic
	Cases, N (%)	IR (95% CI)	Cases, N (%)	IR (95% CI)	Cases, N (%)	IR (95% CI)	Cases, N (%)	IR (95% CI)	Cases, N (%)	IR (95% CI)
Histology, 2000–2009 a
Endometrioid	13,205 (80.9)	8.6 (8.3, 8.7)	1,355 (61.5)	4.3 (4.1, 4.5)	1,923 (82.8)	8.7 (8.2, 9.0)	150 (80.7)	9.2 (7.8, 10.8)	4,120 (78.5)	8.0 (7.6, 8.2)
Non-Endometrioid	1,189 (7.3)	0.8 (0.8, 0.9)	187 (8.5)	0.6 (0.5, 0.8)	146 (6.3)	0.7 (0.5, 0.8)	NS	NS	339 (6.5)	0.7 (0.5, 0.8)
Sarcoma	1,557 (9.5)	1.0 (0.9, 11)	521 (23.6)	1.7 (1.5, 1.8)	204 (8.8)	0.9 (0.8, 1.1)	20 (10.8)	1.3 (0.8, 1.9)	581 (111)	1.2 (1.0, 1.3)
Other	370 (2.3)	0.2 (0.2, 0.2)	141 (6.4)	0.4 (0.4, 0.5)	49 (2.1)	0.2 (0.2, 0.3)	NS	NS	208 (4.0)	0.4 (0.3, 0.4)
Histology, 2010–2019 a
Endometrioid	12,246 (81.7)	9.4 (9.3, 9.7)	2020 (68.3)	6.4 (6.1, 6.6)	2950 (83.0)	10.2 (9.8, 10.5)	230 (86.1)	15.2(13.3, 17.3)	7286 (80.3)	10.9 (10.7, 11.1)
Non-Endometrioid	1,071 (7.1)	0.8 (0.7, 0.8)	251 (8.5)	0.8 (0.8, 0.9)	265 (7.5)	0.9 (0.8, 1.2)	20 (7.5)	1.4 (0.8, 2.2)	642 (7.1)	1.0 (1.0, 1.1)
Sarcoma	1,313 (8.8)	1.0 (0.9, 1.1)	555 (18.8)	1.8 (1.6, 1.9)	268 (7.5)	0.9 (0.9, 1.1)	NS	NS	874 (9.6)	1.3 (1.1, 1.4)
Other	362 (2.4)	0.3 (0.2, 0.3)	130 (4.4)	0.4 (0.3, 0.5)	73 (2.1)	0.3 (0.2, 0.3)	NS	NS	277 (3.1)	0.4 (0.3, 0.4)
Stage at Diagnosis, 2000–2009 b
Localized	7339 (76.9)	4.8 (4.6, 4.8)	873 (68.0)	2.8 (2.6, 2.9)	1085 (74.1)	4.9 (4.6, 5.1)	80 (72.7)	0.9 (3.9, 6.1)	2431 (74.6)	4.7 (4.5, 5.0)
Regional	1643 (17.2)	1.1 (1.0, 1.1)	241 (17.6)	0.8 (0.7, 0.8)	283 (19.3)	1.3 (1.2, 1.4)	25 (22.7)	1.5 (1.0, 2.2)	612 (18.8)	1.2 (1.1, 1.4)
Distant	559 (5.9)	0.4 (0.4, 0.4)	163 (14.3)	0.5 (0.5, 0.6)	96 (6.6)	0.4 (0.3, 0.5)	NS	NS	214 (6.6)	0.4 (0.3, 0.5)
Stage at Diagnosis, 2010–2019
Localized	11111 (77.4)	8.6 (8.2, 8.6)	1894 (68.0)	6.0 (5.7, 6.2)	2534 (74.5)	8.8 (8.5, 9.1)	198 (77.0)	13.0 (11.4, 15.1)	6499 (75.9)	9.7 (9.6, 10.0)
Regional	2297 (16.0)	1.8 (1.6, 19)	491 (18.0)	1.6 (1.4, 1.8)	612 (18.0)	2.1 (2.0, 2.3)	42 (16.3)	2.9 (2.1, 4.1)	1408 (16.4)	2.2 (2.0, 2.3)
Distant	952 (6.6)	0.7 (0.7, 0.8)	399 (14.0)	1.3 (1.2, 1.4)	256 (7.5)	0.9 (0.7, 1.0)	17 (6.6)	1.1 (0.6, 1.8)	655 (7.7)	1.0 (0.9, 1.1)

Notes. Percentages were calculated using non-missing data. All incidence rates (IRs) were age adjusted to the 2000 US standard population and presented per 100,000 population. All IRs were also corrected for hysterectomy, using survey weighted hysterectomy prevalence data from the Behavioral Risk Factor Surveillance System (BRFSS). IR estimates were not presented for those strata with insufficient sample size (≤16)¹⁹ denoted as NS.

Abbreviations: Adeno, Adenocarcinoma; CI, Confidence Interval; IR, Incidence Rates; NA/AN, Native American/Alaska Native; NH, Non-Hispanic; NOS, Not Otherwise Specified; NS, Not Sufficient; PI, Pacific Islander.

^aIncidence rates of endometrioid and non-endometrioid cancers reflect values after re-classification of Adenocarcinoma, NOS (8140/3) into endometrioid and non-endometrioid histologic subtypes.

^bStage information was not available prior to 2004, resulting in n=9,883 missing datapoints for the period between 2000–2009.

Incidence rates have increased for all stages across all racial/ethnic groups over time, reflecting the overall increase in uterine cancer incidence (Table 2). Overall, the highest incidence trends were observed for distant tumors, steadily increasing by 3.38%/year (95% CI 2.56, 4.24) between 2004 and 2019 (Figure S5). Localized disease showed a significant rise after 2011 (APC 2.60%/year, 95% CI 2.15, 3.52), averaging 1.82%/year (95% CI, 1.58, 2.06) from 2004 to 2019 (Figure S5). Conversely, regional disease increased significantly until 2014 (APC 1.77%/year, 95% CI 0.64, 7.86) and declined thereafter (−3.12%/year, 95% CI −11.48, −0.14). The incidence trends of the local and distant endometrioid subtypes follow a similar pattern to the overall stage trends (Average APC 2.0%/year, 95% CI 1.69, 2.32 for localized, 4.4%/year, 95% CI 3.16, 5.31 for distant and stable trends for regional disease).

Five-Year Relative Survival Overall, and by Race/Ethnicity, and Clinical Characteristics

The overall five-year relative survival rate in women with uterine cancer was 88.6% (95% CI 88.1%, 89.0%) in 2010–2019 and did not change over time in the full study population (Table 3). NH-Black women with uterine cancer had worse overall survival rates compared other racial/ethnic groups (79.7% vs. 88.6–90.3%, respectively in 2010–2019) (Table 3).

Table 3.

Five-Year Relative Survival Rates (2000–2019) of Microscopically Confirmed Uterine Cancer Diagnosed between 2000–2018, Overall and Across Demographic and Clinical Characteristics, Among U.S. Women Aged 20 to 49 Years, SEER 22

Calendar Period	2000–2009a		2010–2019 a,b
	Cases, N (Deaths, N)	Five-Year Relative Survival Rate% (95% CI)	Cases, N (Deaths, N)	Five-Year Relative Survival Rate% (95% CI)
Overall Survival Rate	22,277 (2,652)	88.9 (88.4, 89.3)	23,891 (2,523)	88.6 (88.1, 89.0)
Race/Ethnicity
NH-White	13,297 (1,423)	90.3 (89.7, 90.8)	11,261 (1,087)	90.1 (89.5, 90.8)
NH-Black	1,860 (420)	78.5 (76.5, 80.4)	2,269 (425)	79.7 (77.7, 81.5)
NH-Asian/PI	2,163 (222)	90.0 (88.6, 91.2)	2,870 (282)	88.8 (87.4, 90.0)
NH-NA/AN	183 (27)	86.7 (80.3, 91.1)	224 (22)	90.3 (84.6, 93.9)
Hispanic	4,774 (560)	88.4 (87.4, 89.3)	7,267 (707)	88.6 (87.7, 89.5)
Histology
Endometrioid without Adeno, NOS c	14,524 (1,074)	93.4 (93.0, 93.9)	17,920 (1,040)	93.9 (93.4, 94.3)
Endometrioid with Adeno, NOS c	17,813 (1,360)	93.2 (92.8, 93.6)	19,130 (1,157)	93.7 (93.2, 94.1)
Non-Endometrioid	1,317 (363)	73.2 (70.6, 75.6)	1,699 (430)	73.2 (70.8, 75.5)
Sarcoma*	2,465 (759)	69.8 (67.9, 71.6)	2,367 (711)	66.4 (64.2, 68.5)
Other*	682 (170)	74.6 (71.0, 77.9)	695 (225)	65.0 (61.0, 68.6)
Stage at Diagnosis d
Localized	9,958 (463)	96.2 (95.8, 96.6)	17,088 (634)	96.2 (95.8, 96.5)
Regional	2,371 (469)	80.9 (79.2, 82.5)	3,844 (634)	82.1 (80.7, 83.4)
Distant	915 (642)	29.5 (26.5, 32.6)	1,784 (1,120)	31.8 (29.4, 34.3)

Notes. Percentages were calculated using non-missing data. All five-year relative survival estimates were age standardized using the SEER recommended variable “International Cancer Survival Standards (ICSS) 1”.

Abbreviations: Adeno, Adenocarcinoma; CI, Confidence Interval; NA/AN, Native American/ Alaska Native; NH, Non-Hispanic; NOS, Not Otherwise Specified; PI, Pacific Islander.

^aP-values (two-tailed) indicate the statistical difference in 5-year relative survival estimates when comparing two periods, with 2000 to 2009 as the reference. P-values are calculated from the z-statistic with the formula p=2×(1−Φ(z)), where Φ(x)= standard normal distribution’s cumulative distribution function, and z = absolute z-score. P <0.05 considered statistically significant and denoted with a (*).

^bSurvival was estimated until the year 2019, but cases diagnosed in year 2019 were excluded from this time period.

^cFive-year relative survival of women with endometrioid and non-endometrioid histologic subtype, without Adenocarcinoma, NOS (8140/3)

^dStage information was not available prior to 2004, resulting in n=9,883 missing datapoints for the period between 2000–2009.

Of defined histologic subtypes, women diagnosed with sarcomas had the lowest survival rates (66.4% in 2010–2019) followed by women diagnosed with non-endometrioid (73.2%) and endometrioid (93.7%) subtypes (Table 3). Survival rates among women diagnosed with sarcoma decreased from 69.8% in 2000–2009 to 66.4% in 2010–2019 (p-value=0.01), whereas survival rates for other histologic subtypes were stable over time. Overall survival rates decreased with more advanced stage at diagnosis (localized 96.2%, regional 82.1% and distant 31.8%), and were stable over time (Table 3). NH-Black women with uterine cancer had the lowest survival rates across all histologic subtypes and tumor stages, when compared to other racial/ethnic groups (Table S2).

Comment

Principal findings

The 1.7% annual increase of uterine cancer among women <50 years is primarily driven by increasing rates of endometrioid cancers, which constitute 80% of cases in this age range and were rising most steeply in underrepresented racial/ethnic groups. NH-NA/AN women had the highest incidence rates of endometrioid subtype, and NH-Black women had the lowest. Sarcomas were the second most common subtype, comprising 10% of cases Sarcoma incidence rates were highest among NH-Black women, with NH-Asian/PI women having the lowest rates. While sarcoma incidence rates appeared to be increasing slightly in the full study population, race-stratified analyses revealed that this increase was only significant among Hispanic women. For non-endometrioid subtypes, incidence rates were highest among NH-NA/AN women and lowest among NH-Black women. Overall trends of non-endometrioid subtypes increased from 2000–2015, followed by a sharp decline from 2016–2019; however, stratified analyses revealed heterogeneity in these trends across racial/ethnic groups. Notably, NH-White women experienced a significant increase of non-endometrioid subtype until 2009 followed by a decline, and Hispanic women showed a steady increase throughout the period, whereas trends in other groups were stable. Currently the reasons for these differential trends are not well understood. Survival rates were lowest for sarcoma cases and decreased over time, while remaining relatively stable for other subtypes.

Results in the context of what is known.

Uterine cancers diagnosed among women <50 years show distinct differences in terms of population trends and clinical characteristics compared to the broader population where cancer cases predominantly occur in women aged ≥50 years.^2,8,21 While the increasing incidence of uterine cancer in <50 year-old women has been previously demonstrated in the U.S.^9,22–24 and other countries (New Zealand,²⁵ Canada,^26,27 England²⁸), the incidence rate differences by histology across racial/ethnic groups have not been investigated before. Our hysterectomy corrected incidence trends align with these previous studies, showing that although uterine cancer rates are relatively low in women <50 years compared to ≥50 years of age, rates are increasing at a more rapid rate compared with the broader population.^2,7

In a previous study conducted within SEER, rising uterine cancer incidence rates in women aged 30–79 years were mainly associated with increasing rates of non-endometrioid subtypes among women in all racial and ethnic groups.² Conversely, among women <50 years, increasing incidence rates appear to be related to rising rates of endometrioid cancers, with the most rapid increases occurring in underrepresented racial/ethnic groups. Rates of sarcomas also increased between 2002 to 2019 among all women <50 years, but these trends were only significant among Hispanic women when stratified by race/ethnicity. Unlike in the overall population of women aged 30 to 79 years,² non-endometrioid subtypes were rarer in women <50 years, with incidence rates of sarcomas, predominated by leiomyosarcoma, being higher in this age group. Notably, while incidence rates of non-endometrioid subtypes are twice as high among NH-Black women in the overall population (aged 30–79 years) compared with other racial/ethnic groups,² rates of these tumors were lowest among NH-Black women <50 years. With respect to trends, rates of non-endometrioid cancers in women <50 years increased from 2000 to 2015, aligning with trends in the overall population (aged 30–79 years); however, this was followed by a decline from 2016 to 2019. Unlike other racial/ethnic groups incidence trends of non-endometrioid cancers in Hispanic women closely resemble the trends of this subtype in the overall population (aged 30–79 years).

Survival data on women <50 years with uterine cancer of age are scarce²⁹ and have not been reported by race/ethnicity and histology. We found no differences in survival over time in most racial/ethnic groups by histology, except for women with sarcoma whose survival rates worsened over time. Consistent with findings from the overall population, NH-Black women had the lowest survival rates, irrespective of stage at diagnosis and histology in our study.²

Potential explanations

The rising trends of uterine cancer in women <50 years, particularly endometrioid subtypes, could be associated with increasing life course obesity prevalence among women in the U.S.^30–33 Women aged ≥20 years showed a significant rise in obesity prevalence since 1999, reaching 42% in 2017–2018,^34,35 and the prevalence of Class III obesity being highest among 40–59 age group.³⁴ Obesity in early adulthood has been shown to be independently associated with uterine cancer risk and is more strongly associated with risk in pre- and perimenopausal women and for endometrioid subtypes.³⁶ Given that nulliparity is associated with an increased risk of uterine cancer in all women,³³ the rising prevalence of nulliparity³⁷ and delayed age at parity³⁸ may also contribute to shaping uterine cancer population trends in this age group. At least to a lesser extent uterine cancer trends may also be impacted by the declining uptake of hysterectomy,^18,39,40 and preference of fertility preserving treatment options for premalignant or benign uterine conditions over hysterectomy among reproductive age women.⁴¹ While still rare, sarcomas were the second most common subtype among women <50 years of age and rates appear to be increasing among Hispanic women. Although risk factors for sarcomas are not well understood, our data suggests a potential role for exposures occurring earlier in life (i.e., premenopause) and/or genetic factors. In contrast, it appears that risk factors for non-endometrioid subtypes may increase with age, as rates of these subtypes were lowest among women <50 years. Future studies are warranted to understand age-specific differences and risk factors for both sarcomas and non-endometrioid subtypes including lifestyle and environmental exposures that may vary by age across populations.

Clinical implications

Increasing trends of uterine cancers in women <50 years indicate a need for better early detection methods in this age group. The primary symptom associated with uterine abnormalities, whether benign or malignant, is abnormal uterine bleeding. The American College of Obstetricians and Gynecologists (ACOG) clinical guidelines recommend that nonpregnant reproductive age women over 45 with abnormal uterine bleeding and women under 45 with a history of unopposed estrogen exposure should undergo endometrial tissue sampling as a first line test.⁴² However, in premenopausal women, distinguishing signs and symptoms of uterine cancer is challenging, and only about 1–2% of women with abnormal bleeding are diagnosed with uterine cancers.⁴³ Therefore, improved risk assessment strategies are needed to guide clinical decision making.

Research implications

Our findings raise interesting and novel questions with potential clinical and research implications. For example, it is important to understand the factors contributing to the decline of non-endometrioid rates in NH-White women in recent years and why among NH-Black women rates of these tumors are lowest compared with other groups <50 years, but twice as high among NH-Black women in older ages compared with other groups. Risk factors for non-endometrioid subtypes are not well understood, and few studies have been conducted within racially and ethnically diverse populations. Given the poor prognosis of these subtypes, along with uterine sarcomas which we found to be more common in women <50 years, more studies are needed to investigate lifestyle, environmental, and other risk factors associated with these cancers. Additionally, more research is needed to explore the reasons behind worsened survival rates in women with sarcoma over time, as well as the absence of discernible improvements across other histologic subtypes. Due to the rarity in the overall population, cases with sarcomas are often excluded from clinical trials.

It is also important to note that incidence rates of uterine cancer were highest in NH-NA/AN women but we were unable to assess trends due to the small sample size. Given the limited data on uterine cancer incidence and risk factors among NH-NA/AN women, our findings underscore the need for increased research efforts in this population, mirroring the requirements for other underrepresented racial/ethnic groups in this age bracket.

Strengths and limitations

To the best of our knowledge, ours is the first study to investigate hysterectomy-corrected uterine cancer incidence rates and trends by race/ethnicity and histology among U.S. women <50 years. We incorporated the latest data from SEER 22 representing 48% of the U.S. population¹² and addressed potential misclassification of adenocarcinoma, NOS, Not Otherwise Specified over time.

There are several limitations. We lacked information on tumor stage prior to 2004, thereby limiting our trend analyses by tumor stage from 2004 to 2019. We also lacked individual-level data on family history of cancer, which is particularly relevant for women <50 years. However, considering that approximately 9% of cases are attributed to the most common hereditary uterine cancer condition, Lynch syndrome, among women <50 years,⁴⁴ we infer that hereditary factors contribute only minimally compared to those caused by non-hereditary factors. Individual-level risk factors, treatment related factors, and investigations into the management of benign conditions were not available in SEER but could offer valuable insights into the observed trends.

Conclusion

Uterine cancers diagnosed among women <50 years show distinct clinical profile and population trend differences when compared to the trends of the overall population (aged 30–79 years) overall and across racial/ethnic groups. Our study shows that the rising trends of uterine cancer in this age group is primarily attributed to the rising rates of cancers with endometrioid histology. Further studies are warranted to elucidate the reasons and underlying factors explaining these population trends across racial/ethnic and histologic groups.

Supplementary Material

1

Figure S1. Age-adjusted hysterectomy corrected incident rate trends of uterine cancer among women aged 20–49 years, by race/ethnicity and histologic subtype. Incidence rate trends shown for (A) Endometrioid, (B) Non-Endometrioid, and (C) Sarcoma histologic subtypes, by race/ethnicity.

Figure S2. Age-adjusted hysterectomy corrected incident rate trends of endometrioid subtype of uterine cancer among women aged 20–49 years, by race/ethnicity.

Figure S3. Age-adjusted hysterectomy corrected incident rate trends of sarcoma subtype of uterine cancer among women aged 20–49 years, by race/ethnicity.

Figure S4. Age-adjusted hysterectomy corrected incident rate trends of non-endometrioid subtype of uterine cancer among women aged 20–49 years, by race/ethnicity.

Figure S5. Age-adjusted hysterectomy corrected incident rate trends of uterine cancer, stage at diagnosis for the overall study population of women aged 20–49 years (A), and among those with Endometrioid subtype (B).

AJOG at a Glance.

Why was this study conducted?

Uterine cancer is one of the few cancers with increasing incidence and mortality in the U.S. Given the increasing rates of uterine cancer in women <50 years, we wanted to understand the clinical characteristics of these cancers overall and by race/ethnicity and whether they have changed over time.

Key findings:

Uterine cancer is increasing in women <50 years, primarily due to rising rates of endometrioid subtype, with a sharper rise observed in underrepresented racial/ethnic groups. Five-year relative survival rates have remained stable in this age group over time for endometrioid and non-endometrioid subtypes, whereas survival among women with sarcoma declined.

What does this add to what is known?

In women >50 years, rising trends are primarily associated with non-endometrioid subtypes, whereas in women <50 years, rising trends are primarily associated with endometrioid subtype, suggesting potential age-specific differences in exposures/risk factors.