Background
Radiation therapy has long been part of the treatment of endometrial cancer. Despite the long history of radiation use, prospective trials in the United States and Europe have been unable to demonstrate a survival benefit with adjuvant radiotherapy compared with observation. Whereas radiation has been associated with a decreased rate of locoregional failure, the treatment is also associated with substantial toxicity. However, a randomized trial published in 2010 demonstrated that, compared with external beam radiation therapy (EBRT), vaginal brachytherapy was less toxic and as effective in reducing locoregional relapses.
Objective
We examined patterns of use of external beam radiation therapy for women with high intermediate risk endometrial cancer.
Study Design
We examined the use of external beam radiation therapy in women registered in the National Cancer Data Base with high intermediate risk, stage I endometrial cancer treated from 2008 through 2012. High intermediate risk was defined as age > 60 years with a stage IA, grade 3 tumors or stage IB, grade 1 or 2 tumors. Multivariable models of EBRT use were developed.
Results
Among 8242 women, 915 (11.1%) received EBRT, 2614 (31.7%) were treated with brachytherapy, and 4713 (57.2%) did not receive any adjuvant radiation. The use of EBRT was 18.1% in 2008 and declined to 8.6% in 2012, whereas the use of brachytherapy rose each year from 26.5% in 2008 to 37.6% in 2012 ( P < .0001). External beam radiation was administered to 7.9% of patients with stage IA/grade 3 tumors, 8.8% of those with stage IB/grade 1 cancers, and to 15.2% of women with stage IB/grade 2 neoplasms ( P < .0001). EBRT was utilized in 10.1% of women who underwent lymphadenectomy compared with 22.0% who did not undergo lymphadenectomy ( P < .0001). In a multivariable model, black women were more likely to receive EBRT than white women (relative risk [RR], 1.33; 95% confidence interval [CI], 1.03–1.70). Similarly, patients in the eastern United States, those treated at community cancer centers and comprehensive community cancer programs, patients in metropolitan areas, and those diagnosed in earlier years were more likely to undergo EBRT. Patients with stage IB/grade 2 tumors (RR, 1.96; 95% CI, 1.65–2.32) were more likely to receive EBRT than those with stage IA/grade 3 neoplasms. Those women who did not undergo lymphadenectomy were more than twice as likely to receive EBRT compared with those who had a lymphadenectomy (RR, 2.32; 95% CI, 1.99–2.72).
Conclusion
Despite data from randomized trials, approximately 9% of women with high intermediate risk of endometrial cancer continue to receive EBRT. Performance of lymphadenectomy is associated with a lower likelihood of external beam radiation therapy.
Radiation therapy has long been part of the treatment of endometrial cancer. Most commonly, adjuvant radiation therapy has been administered after hysterectomy. Despite the long history of radiation use, prospective trials have been unable to demonstrate a survival benefit with adjuvant radiotherapy.
The Gynecologic Oncology Group (protocol 99) and the Post Operative Radiation Therapy in Endometrial Cancer (PORTEC)-1 trials compared external beam radiotherapy (EBRT) with observation in women with stage I endometrial cancer. Both trials demonstrated that EBRT reduced locoregional recurrences but did not have an impact on overall survival. Furthermore, EBRT was associated with substantial toxicity.
Although these reports failed to demonstrate an improvement in survival with EBRT, both of these studies subsequently defined a higher-risk group of patients who derived the greatest absolute benefit from EBRT in reducing locoregional relapse. Termed high intermediate risk (HIR), this group is based on a combination of factors including advanced age, grade, depth of myometrial invasion, and lymphvascular space invasion. A follow-up investigation, PORTEC-2, which was limited to early-stage, HIR patients, compared vaginal brachytherapy with EBRT. Published in 2010, PORTEC-2 found that vaginal brachytherapy was as effective as EBRT in reducing locoregional recurrences and was substantially less toxic.
Based on these studies, vaginal brachytherapy is the favored method of radiation delivery when adjuvant radiotherapy is administered for early-stage endometrial cancer. To date, few studies have examined the trends in the use of adjuvant radiotherapy since the reporting of the PORTEC-2 trial. We performed a population-based analysis to examine the patterns of use of adjuvant radiation in women with HIR, stage I endometrial cancer.
Materials and Methods
The National Cancer Data Base (NCDB) was utilized for the analysis. The NCDB is sponsored by the American College of Surgeons and the American Cancer Society and collects data on patients from across the United States on incident cancer cases from more than 1500 Commission on Cancer–affiliated hospitals. The sampling schema captures approximately 70% of all newly diagnosed cancer cases. Data on patient demographics, clinical data, tumor characteristics, staging, treatment, and overall survival are collected. The Columbia University Institutional Review Board deemed the study exempt.
Women with stage I endometrioid adenocarcinomas of the endometrium treated from 2008 through 2012 were analyzed. The cohort was limited to women with HIR tumors as defined by the PORTEC-2 study group: age > 60 years with a stage IA, grade 3 tumors or stage IB, or grade 1 or 2 tumors. Based on the revised International Federation of Gynecology and Obstetrics staging criteria, patients with < 50% myometrial invasion, including those with tumors confined to the endometrium, were classified as stage IA neoplasms.
Patients were stratified based on the receipt of radiation into the following groups: no radiation, brachytherapy, and EBRT (either alone or in combination with brachytherapy). Receipt of radiation either at the center at which a patient was diagnosed or at another facility were captured as radiation treatment.
Clinical variables analyzed included age (61–70, 71–80, or > 80 years), race (white, black, Hispanic, other, or unknown), insurance (private, Medicare, Medicaid, uninsured, or other), region of residence (eastern, south, Midwest, or west), and location (metropolitan, urban, rural, and unknown). Performance of lymphadenectomy was noted for each patient.
Comorbidity was measured using the Deyo classification of the Charlson comorbidity score (0, 1, ≥ 2). A combination of tumor stage and grade were utilized to categorize patients as stage IA/grade 3, stage IB/grade 1, or stage IB/grade 2. Hospital location was classified as metropolitan, urban, or rural. Facility type (academic/research, comprehensive community cancer center, or community cancer center) was defined by the American Cancer Society’s Commission on Cancer as academic/research hospitals (institutions affiliated with university medical schools or those designated as National Cancer Institute Comprehensive Cancer programs), community cancer centers (institutions that diagnose or treat 100–649 cancer cases annually), and comprehensive community centers (institutions that diagnose or treat ≥ 650 cancer cases annually).
Frequency distributions for categorical variables were analyzed based on receipt of external beam radiation using χ 2 tests. Multivariable generalized estimating equations with a Poisson distribution and log link function were developed to examine predictors of EBRT while adjusting for other clinical, demographic, and hospital characteristics. Results are reported with risk ratios and 95% confidence intervals (CIs). Sensitivity analyses were performed after stratifying the cohort based on performance of lymphadenectomy. All analyses were performed with SAS version 9.4 (SAS Institute Inc, Cary, NC). All analyses were two tailed and a value of P < .05 was considered statistically significant.
Results
We identified a total of 8242 women with HIR endometrial cancer treated from 2008 through 2012. Overall, 915 (11.1%) received EBRT, 2614 (31.7%) were treated with brachytherapy, and 4713 (57.2%) did not receive any adjuvant radiation. The use of external beam therapy was 18.1% in 2008, declined to 8.5% in 2010, and then rose slightly to 9.8% in 2011 and 8.6% in 2012. In contrast, utilization of brachytherapy rose each year from 26.5% in 2008 to 37.6% in 2012 ( P < .0001) ( Figure ).
Table 1 displays the clinical and demographic characteristics of the cohort. External beam radiation was administered to 7.9% of patients with stage IA/grade 3 tumors, 8.8% of those with stage IB/grade 1 cancers, and 15.2% of women with stage IB/grade 2 neoplasms ( P < .0001) ( Table 1 ). EBRT was utilized in 10.1% of women who underwent lymphadenectomy compared with 22.0% who did not undergo lymphadenectomy ( P < .0001).
| Characteristics | No radiation | External beam radiotherapy | Vaginal brachytherapy | P value | |||
|---|---|---|---|---|---|---|---|
| n | (%) | n | (%) | n | (%) | ||
| Patients | 4713 | (57.2) | 915 | (11.1) | 2614 | (31.7) | |
| Year of diagnosis | < .0001 | ||||||
| 2008 | 779 | (55.4) | 255 | (18.1) | 372 | (26.5) | |
| 2009 | 776 | (58.3) | 166 | (12.5) | 390 | (29.3) | |
| 2010 | 1105 | (60.9) | 154 | (8.5) | 556 | (30.6) | |
| 2011 | 1080 | (57.5) | 184 | (9.8) | 615 | (32.7) | |
| 2012 | 973 | (53.8) | 156 | (8.6) | 681 | (37.6) | |
| Age, y | < .0001 | ||||||
| 61–70 | 2588 | (55.2) | 542 | (11.6) | 1559 | (33.2) | |
| 71–80 | 1451 | (55.8) | 304 | (11.7) | 847 | (32.6) | |
| > 80 | 674 | (70.9) | 69 | (7.3) | 208 | (21.9) | |
| Race/ethnicity | .04 | ||||||
| White | 4000 | (56.7) | 773 | (10.9) | 2287 | (32.4) | |
| Black | 296 | (57.4) | 66 | (12.8) | 154 | (29.8) | |
| Hispanic | 204 | (63.6) | 39 | (12.1) | 78 | (24.3) | |
| Other | 140 | (60.6) | 23 | (10.0) | 68 | (29.4) | |
| Unknown | 73 | (64.0) | 14 | (12.3) | 27 | (23.7) | |
| Insurance status | .03 | ||||||
| Private insurance | 1302 | (54.8) | 261 | (11.0) | 812 | (34.2) | |
| Medicaid | 103 | (52.6) | 23 | (11.7) | 70 | (35.7) | |
| Medicare | 3110 | (58.0) | 606 | (11.3) | 1645 | (30.7) | |
| Not insured | 112 | (65.5) | 16 | (9.4) | 43 | (25.1) | |
| Other government/unknown | 86 | (61.4) | 9 | (6.4) | 45 | (32.1) | |
| Comorbidity | .16 | ||||||
| 0 | 3377 | (56.5) | 675 | (11.3) | 1928 | (32.2) | |
| 1 | 1067 | (58.3) | 196 | (10.7) | 566 | (30.9) | |
| 2 | 269 | (62.1) | 44 | (10.2) | 120 | (27.7) | |
| Lymph nodes examined | < .0001 | ||||||
| Yes | 4270 | (56.8) | 756 | (10.1) | 2493 | (33.2) | |
| No | 443 | (61.3) | 159 | (22.0) | 121 | (16.7) | |
| Stage and grade | < .0001 | ||||||
| Stage IA/grade 3 | 1313 | (60.6) | 172 | (7.9) | 680 | (31.4) | |
| Stage IB/grade 1 | 1746 | (61.2) | 252 | (8.8) | 853 | (29.9) | |
| Stage IB/grade 2 | 1654 | (51.3) | 491 | (15.2) | 1081 | (33.5) | |
| Region | < .0001 | ||||||
| East | 787 | (42.8) | 254 | (13.8) | 796 | (43.3) | |
| South | 1460 | (65.1) | 188 | (8.4) | 596 | (26.6) | |
| Midwest | 1618 | (57.8) | 353 | (12.6) | 829 | (29.6) | |
| West | 848 | (62.3) | 120 | (8.8) | 393 | (28.9) | |
| Location | .001 | ||||||
| Metropolitan | 3636 | (56.4) | 741 | (11.5) | 2073 | (32.1) | |
| Urban | 795 | (59.0) | 124 | (9.2) | 429 | (31.8) | |
| Rural | 103 | (61.7) | 13 | (7.8) | 51 | (30.5) | |
| Unknown | 179 | (64.6) | 37 | (13.4) | 61 | (22.0) | |
| Facility type a | < .0001 | ||||||
| Academic | 1784 | (54.8) | 303 | (9.3) | 1171 | (35.9) | |
| Community cancer program | 221 | (55.4) | 90 | (22.6) | 88 | (22.1) | |
| Comprehensive community cancer program | 2702 | (59.1) | 520 | (11.4) | 1353 | (29.6) | |
In a univariate analysis, EBRT was more commonly used in younger women, those with some type of insurance, those with no comorbidities, women who did not undergo lymphadenectomy, patients with stage IB/grade 2 tumors, residents of the eastern United States, and those residing in metropolitan areas and women treated at community cancer programs. Use of vaginal brachytherapy increased substantially over time.
In a multivariable model, black women were more likely to receive EBRT than white women (risk ratio [RR], 1.33; 95% CI, 1.03–1.70). Similarly, patients in the eastern United States, those treated at community cancer centers and comprehensive community cancer programs, patients in metropolitan areas, and those diagnosed in earlier years were more likely to undergo EBRT ( Table 2 ).
| Factors | Entire cohort RR (95% CI) | Lymphadenectomy performed RR (95% CI) | No lymphadenectomy RR (95% CI) |
|---|---|---|---|
| Year of diagnosis | |||
| 2008 | 2.16 (1.78–2.62) a | 2.30 (1.85–2.86) | 1.65 (1.07–2.55) a |
| 2009 | 1.49 (121–1.84) b | 1.53 (1.21–1.94) b | 1.41 (0.9–2.18) |
| 2010 | Referent | Referent | Referent |
| 2011 | 1.20 (0.97–1.49) | 1.18 (0.93–1.50) | 1.51 (0.96–2.37) |
| 2012 | 1.05 (0.83–1.34) | 1.02 (0.78–1.34) | 1.25 (0.80–1.96) |
| Age, y | |||
| 61–70 | Referent | Referent | Referent |
| 71–80 | 0.98 (0.85–1.13) | 1.02 (0.87–1.20) | 0.81 (0.59–1.12) |
| > 80 | 0.51 (0.40–0.65) b | 0.52 (0.39–0.69) b | 0.47 (0.31–0.74) a |
| Race/ethnicity | |||
| White | Referent | Referent | Referent |
| Black | 1.33 (1.03–1.70) a | 1.48 (1.13–1.93) a | 0.72 (0.37–1.41) |
| Hispanic | 1.15 (0.83–1.60) | 1.21 (0.83–1.77) | 0.91 (0.49–1.70) |
| Other | 1.04 (0.69–1.57) | 1.15 (0.74–1.77) | 0.39 (0.06–2.65) |
| Unknown | 1.25 (0.78–2.00) | 1.10 (0.63–1.93) | 1.74 (0.86–3.53) |
| Insurance status | |||
| Private insurance | Referent | Referent | Referent |
| Medicaid | 1.08 (0.74–1.57) | 1.02 (0.67–1.54) | 1.48 (0.61–3.55) |
| Medicare | 1.02 (0.87–1.20) | 0.97 (0.82–1.15) | 1.42 (0.99–2.04) |
| Not insured | 0.90 (0.59–1.36) | 0.84 (0.52–1.36) | 1.47 (0.54–3.96) |
| Other government/unknown | 0.65 (0.34–1.25) | 0.62 (0.30–1.31) | 0.78 (0.19–3.17) |
| Comorbidity | |||
| 0 | Referent | Referent | Referent |
| 1 | 0.92 (0.79–1.07) | 0.94 (0.80–1.11) | 0.84 (0.59–1.19) |
| 2 | 0.84 (0.63–1.11) | 0.84 (0.60–1.16) | 0.84 (0.52–1.35) |
| Lymph nodes examined | |||
| Yes | Referent | — | — |
| No | 2.32 (1.99–2.72) b | — | — |
| Stage/grade | |||
| Stage IA/grade 3 | Referent | Referent | Referent |
| Stage IB/grade 1 | 1.09 (0.90–1.32) | 1.07 (0.86–1.33) | 1.13 (0.76–1.68) |
| Stage IB/grade 2 | 1.96 (1.65–2.32) b | 2.01 (1.67–2.41) b | 1.63 (1.10–2.43) a |
| Region | |||
| East | Referent | Referent | Referent |
| South | 0.59 (0.47–0.75) b | 0.62 (0.47–0.80) a | 0.49 (0.32–0.75) a |
| Midwest | 0.92 (0.75–1.14) | 0.91 (0.73–1.15) | 0.97 (0.68–1.37) |
| West | 0.60 (0.47–0.78) b | 0.59 (0.45–0.78) a | 0.64 (0.40–1.03) |
| Location | |||
| Metropolitan | Referent | Referent | Referent |
| Urban | 0.75 (0.62–0.91) a | 0.75 (0.61–0.92) a | 0.73 (0.48–1.09) |
| Rural | 0.66 (0.38–1.15) | 0.68 (0.35–1.31) | 0.70 (0.30–1.62) |
| Unknown | 1.14 (0.85–1.52) | 1.19 (0.86–1.63) | 0.78 (0.37–1.65) |
| Facility type | |||
| Academic | Referent | Referent | Referent |
| Community cancer program | 2.22 (1.73–2.85) b | 2.27 (1.71–3.01) b | 1.79 (1.16–2.76) a |
| Comprehensive community cancer program | 1.27 (1.05–1.53) a | 1.29 (1.05–1.58) a | 1.18 (0.84–1.64) |
| Other | 1.84 (1.50–2.25) b | 1.91 (1.52–2.38) b | — |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
