LAUSR

OBJECTIVE To examine trends of adjuvant radiotherapy choice and to examine associations between pelvic lymphadenectomy and radiotherapy choice for women with early-stage endometrial cancer. METHODS The Surveillance, Epidemiology, and End Results Program was used to identify surgically treated stage I-II endometrial cancer between 1983 and 2012 (type 1 n=79,474, and type 2 n=25,020). Piecewise linear regression models were used to examine temporal trends of intracavitary brachytherapy (ICBT) and whole pelvic radiotherapy (WPRT) use, pelvic lymphadenectomy rate, and sampled node counts. Multivariable binary logistic regression models were used to identify independent predictors for ICBT use. RESULTS There was a significant increase in ICBT use and decrease in WPRT use during the study period. ICBT use exceeded WPRT use in 2003 for type 1 stage IA, and in 2007 for type 1 stage IB and type 2 stage IA diseases. In addition, number of sampled pelvic nodes significantly increased over time in type 1-2 stage I-II diseases (mean, 7.0-12.7 in 1988 to 15.2-17.6 in 2012, all P<0.001). On multivariable analysis, extent of sampled pelvic nodes was significantly associated with ICBT use for type 1 cancer: adjusted-odds ratios for 1-10 and >10 nodes versus no lymphadenectomy in stage IA (1.38/2.40), IB (2.75/6.32), and II (1.36/2.91) diseases. Similar trends were observed for type 2 cancer: adjusted-odds ratios for stage IA (1.69/3.73), IB (2.25/5.65), and II (1.36/2.19) diseases. CONCLUSION Our results suggest that surgeons and radiation oncologists are evaluating the extent of pelvic lymphadenectomy when counseling women with early-stage endometrial cancer for adjuvant radiotherapy.