ESTRO 2023 - Abstract Book

S1113

Digital Posters

ESTRO 2023

Esophageal and gastroesophageal (GE) tumors have a poor prognosis with a 5-year relative survival of around 25%. Currently, limited data exists comparing definitive chemoradiotherapy (dCRT) with neoadjuvant chemoradiotherapy (nCRT) followed by esophagectomy. We analyzed outcomes of patients treated with dCRT and nCRT, focusing on overall survival (OS), local and distant recurrences as well as on acute and late toxicity. Materials and Methods Patients with esophageal and GE tumors treated with dCRT or nCRT followed by surgery between January 2001 and December 2019 were retrospectively analyzed. Only patients treated with curative intent were included. OS, local recurrence (LR), distant metastasis free survival (DMFS) and early (<6 months) as well as late (>6 months) toxicity were determined. GLMs were used for categorical variables while Cox proportional hazard regression was performed for OS, LR and DMFS. Survival curves were analyzed using Kaplan-Meier analysis. Data analysis was performed in R and p-values below 5% were considered significant. The study was approved by the institutional ethical committee. Results A total of 300 patients were identified of which 160 underwent nCRT followed by surgery and 140 were treated with dCRT. Patient characteristics are depicted in Table 1. After a mean follow-up of 30.8 months, a tendency towards a better OS was observed in univariate analysis for patients treated with nCRT (HR 0.771; 95% CI 0.577-1.029; p-value 0.076) (Fig. 1). This tendency could not be confirmed in multivariate analysis. LR occurred in 40 patients (23 and 17 events for dCRT and nCRT resp.). No statistically significant difference in LR was observed between nCRT and dCRT (HR 0.608; 95% CI 0.325 1.138; p-value 0.120). Tumor location (mid vs. distal; HR 2.057), radiation dose ( ≥ 50 Gy vs. 50 Gy ; HR 2.097), treatment technique (VMAT vs. 3D-CRT; HR 0.366) and type of chemotherapy (Cis-5-FU vs. Carbo-Taxol; HR 2.703) appeared significant in univariate analysis, but not in multivariate analysis. Distant metastases occurred in 80 patients (30 and 50 events for dCRT and nCRT resp.). Analysis of DMFS demonstrated a tendency towards significance in univariate analysis in favor of dCRT (HR 1.511; 95% CI 0.961-2.377; p-value 0.072), which could not be confirmed after multivariate analysis. Age and histology appeared (borderline) significant in multivariate analysis in favor of female gender (HR 0.450; p-value 0.027) and squamous cell carcinoma (HR 0.627; p-value 0.074). There was a significantly higher prevalence of early respiratory toxicity in the nCRT subgroup (OR 1.849; 95% CI 1.005-3.401; p-value 0.048). Table 1: Patient characteristics

Fig. 1: OS by treatment modality