25 Background: NCRT followed by surgery per the CROSS trial regimen is an accepted standard of care in the treatment of EC and GEJC. When treatments are used in the real-world setting, there are often patient, treatment and potential outcome differences compared to the original clinical trial. The study aim was to assess the real-world application and outcomes of the CROSS trial protocol. Methods: A retrospective chart review was undertaken of 83 patients (pts) with EC or GEJC who were treated from June 2012 to June 2018 with CRT. 65 pts were with NCRT intent to proceed to surgery. Pts’ demographics, clinical, pathological, treatment and surgical characteristics were assessed and exploratory analyses were conducted to review these factors and outcomes. Analyses included Chi-square, t-tests and Kaplan-Meier. Results: For pts who underwent NCRT (n = 65): median age was 68 yrs (range 52-80), male 79%, adenocarcinoma 82%, median (m) tumor length 5 cm, GERD 43%, clinical stage II/III 95%, and BMI > 30 in 37%. 80% completed CRT with RT ≥ 41.4 Gy; of these 88% had ≥ 50.4 Gy. Delay/interruption in chemotherapy occurred in 46% and in RT 37%. Pts who underwent surgery were younger (p = 0.04) and weighed more (p = 0.05). mOS was 37 months (M) v 14 M in those who started CRT ≤ 8 weeks (w) from diagnosis v > 8 w (p = 0.10). The median time from CRT to surgery was 8.9 w. 40 pts had surgery with a complete response in 38% and a R0 resection in 98%. Postoperative major and minor complications occurred in 67%. Those < 75 yrs v ≥ 75 yrs had a mOS of 32 M v 15 M respectively (log rank p = 0.46). 25 pts did not get surgery; 28% was due to death/progression. Pts who proceeded to surgery had a mOS of 35 M v 12 M in pts who did not go to surgery (log rank p = 0.002). Further correlative outcome data will be presented. Conclusions: Real-world data in our center showed patient, tumor and treatment differences compared to the CROSS trial protocol. Despite the broadening of eligibility and treatment criteria, survival in a single institution setting is maintained with trimodality therapy compared to NCRT alone. Real-world data is of value in the assessment of therapeutic validity of clinical trial data.
S3707 Pancreatic Acinar Metaplasia at the Gastroesophageal Junction in a Patient With Chronic Reflux
