ESTRO 36 Abstract Book

S289 ESTRO 36 2017 _______________________________________________________________________________________________

Purpose or Objective The optimal adjuvant management of Stage IA endometrial cancer with serous or clear cell histology is controversial. The objective of this study is to explore the role of adjuvant therapy in this population and identify patient characteristics of those suitable for observation. Material and Methods Retrospective chart reviews for consecutive patients who underwent hysterectomy for FIGO Stage IA endometrial cancer with serous or clear cell histology between 2004- 2015 were conducted in 6 academic centres. After excluding patients who were upstaged following surgery, a pooled analysis was performed for relevant endpoints. Results A total of 414 patients with a median age of 67 years(range 41-90) met the inclusion criteria. The most common histology were pure serous (64%,n=266) followed by mixed (27%,n=112) and pure clear cell (9%,n=36). Myometrial invasion was identified in 54%(n=222). Most patients underwent some surgical staging including sampling of pelvic lymph node (LN) (81%,n=335), para-aortic LN (45%,n=146), omentum (58%,n=239) and peritoneal washing (62%,n=219). Only 23% of patients (n=95) were considered to have adequate staging (pre-defined as ≥10 pelvic LN, sampled para-aortic LN and omentum). Thirty- four percent of patients (n=140) received adjuvant chemotherapy and carboplatin/paclitaxel was most commonly used (77%, n=108). Adjuvant RT was delivered to 47%(n=178) of patients (external beam alone 16%,n=29; brachytherapy alone 56%, n=99; both 28%,n=50). The median follow-up was 2.7 years (range 0-12). Among patients who did not received any adjuvant treatment, 5- year actuarial estimates were as follows: local control(LC) 88%, regional control(RC) 93%, distant failure(DF) 15%, disease free survival(DFS) 74%, cancer specific survival(CSS) 91% and overall survival(OS) 82%. Adequate staging was associated with better CSS in patients who did not have adjuvant treatment (100% vs. 87%, logrank p=0.0035). Adjuvant RT was associated with better LC(5- year 96% vs. 84%, HR 0.32,logrank p=0.014). The 5-year RC was 93%, which was not found to be significantly improved by external beam RT. Adjuvant chemotherapy was associated with better LC(5-year 94% vs. 84%,HR 0.29, logrank p=0.0079), DFS(5-year 79% vs. 71%, HR 0.47, logrank p=0.033), but not for RC, DF or CSS. The delivery of at least 5 cycles of chemotherapy, was associated with a trend towards better LC (4-year 98% vs. 88%, HR 0.18,logrank p=0.090). Myometrial invasion, lymphovascular invasion, histological subtypes and the proportion of type II component were not found to be significant prognostic factors for LC, RR, DF or CSS. Conclusion Adjuvant radiotherapy and chemotherapy were associated with better LC and DFS but not for RC, DF or CSS for stage IA serous and clear cell uterine cancer. Observation may be an acceptable strategy in patients who have had adequate surgical staging. PV-0549 National Cancer Data Base Analysis of SBRT, IMRT, and Brachytherapy Boost for Cervical Cancer M. Ludwig 1 , M. Bonnen 1 , J. Shiao 2 , B. O'Donnell 3 , T. Pezzi 1 , N. Waheed 1 , S. Sharma 1 1 Baylor College of Medicine, Radiation Oncology, Houston, USA 2 University of Texas San Antonio, Radiation Oncology, San Antonio, USA 3 University of Texas Houston, Radiation Oncology, Houston, USA Purpose or Objective Brachytherapy is a vital component of curative radiotherapy for cervical cancer, but in certain patients, brachytherapy is not feasible because of inaccessibility of brachytherapy comorbidities, patient refusal, or unfavorable anatomy. Our objective was to determine if

stereotactic body radiotherapy (SBRT) and intensity- modulated radiation therapy (IMRT) boost techniques have comparable overall survival (OS) with brachytherapy boosts for patients with cervical cancer when adjusted for known prognostic factors. Material and Methods We used the National Cancer Data Base to study women who received radiation between 2004 and 2013 diagnosed with squamous cell carcinoma, adenocarcinoma or adenosquamous carcinoma of the cervix. Biological effective doses (BED) were calculated for both SBRT and IMRT treatments. A logistic regression model was built to identify factors associated with the receipt of SBRT and IMRT. Correlates of OS were determined using Kaplan- Meier and propensity score matching. Results

Of all 15,905 patients, 14,394 (90.5%) received brachytherapy, 42 (0.8%) received SBRT, and 1468 (9.2%) received IMRT. A multivariable binary logistic regression model identified the following factors associated with receipt of SBRT: advancing age, higher income status, Asian ethnicity, and FIGO Stage III cervical cancer. The following factors were associated with receipt of IMRT: advancing age, treatment at an academic/research program, treatment at an integrated network cancer program, private insurance, lower income status, FIGO Stage III, IVA, and IVB cervical cancer, positive nodal status, metastatic disease, and not receiving chemotherapy as part of the first course of treatment. Median BED was 63.7 Gy for patients treated with IMRT and 75.5 Gy for SBRT (p < 0.001). The median follow-up was 4.8 years. Median survival was 99.1 months (95% CI: 94.4 – 103.8 months), 30.6 months (95% CI: 20.5 – 40.6. months), and 29.8 months (95% CI: 26.0 – 34.7 months) for patients who received brachytherapy, SBRT, and IMRT boost, respectively. Using propensity score matching, there was no significant difference in OS for patients who received SBRT boost vs. brachytherapy boost (HR 1.477; 95% CI 0.746 – 2.926; p = 0.263). IMRT boost patients were also matched and did significantly worse than those who received brachytherapy boost (HR 1.455, 95% CI 1.300 - 1.628; p <0.001).