Résumé
Background International guidelines for locally advanced esophageal squamous cell carcinoma (ESCC) recommend both neoadjuvant chemoradiotherapy (nCRT) followed by planned surgery (P-surg) or definitive chemoradiotherapy (dCRT) with salvage surgery (S-surg). However, the optimal treatment remains unknown, and the ongoing pragmatic NEEDS randomized trial is assessing the efficacy of an organ-preserving strategy after dCRT (Nilsson, Front Oncol 2022). Real-world data on the management of non-metastatic ESCC is limited. This study aims to describe the treatment strategies and prognosis of patients with non-metastatic ESCC in a real-world setting. Methods We conducted a retrospective cohort study of patient treated for non-metastatic ESCC between 2014 and 2022. Data were extracted from the FREGAT database. For the entire population, different treatment strategies were described. We compared patients who received nCRT with those who received dCRT under real-life conditions (“Real-life” population) and according to the NEEDS randomized trial criteria (“NEEDS-like” population). Five-year overall survival (OS) and relapse-free survival (RFS) were compared between nCRT and dCRT patients using a Cox proportional hazard model adjusting for major confounders. In operated patients, the impact of surgery on OS and RFS was assessed using a time-dependent Cox proportional hazard model. Postoperative outcomes were compared between P-surg and S-surg using Fisher’s exact test. Results From 971 patients diagnosed with non-metastatic ESCC, 867 (89.3%) received a curative treatment including dCRT (41.0%), nCRT (28.0%), primary esophagectomy (9.8%), neoadjuvant chemotherapy (3.4%), or endoscopic resection (1.0%). Fifty-nine patients (6,1%) received supportive care or chemotherapy alone. Forty-five patients (4.6%) died during or before treatment. In the “real-life” population (n=659), 269 patients received nCRT and 390 patients received dCRT. A total of 229 patients (93.5%) in nCRT group underwent P-surg and S-surg was performed in 44 patients (11.3%) in dCRT group. After adjustment, dCRT was associated with a significantly poorer 5-year OS and RFS (respectively HR=1.83, 95%CI 1.44-2.33; HR=1.98, 95%CI 1.57-2.49). A complete clinical response (cCR) was observed in 21.6% and 29.5% of the nCRT and dCRT groups, respectively. Among cCR undergoing resection, 35.8% had residual disease on pathological specimen. Time between CRT and surgery was 58 days (IQR44-72) vs. 109 days (IQR71-278) for P-surge and S-Surg respectively. The 90-day mortality, major morbidity (Clavien-Dindo III/IV) and pulmonary complication rates were similar in both groups, except for a higher risk of anastomotic leakage after S-surg (p=0.05). In the “NEEDS-like” population (n=598), findings were consistent and comparable to those observed in the “real-life” population. Conclusion Among non-mestastatic ESCC included in the FREGAT database, 10.6% had a non-curative approach or died prematurely. Most patients underwent dCRT, with S-surg rate of only 11.3%. Neoadjuvant CRT followed by P-surg was associated with the best outcome in terms of OS and RFS in both “rela life” and “Needs-like” populations. In operated patients, postoperative outcomes were similar after nCRT and dCRT, except for a higher incidence of anastomotic leakage after S-surg. This study highlighted the therapeutic strategies used and their results in real-life conditions for non-metastatic ESCC inside the FREGAT database. The treatment option of nCRT followed by P-surg appears to remain the optimal curative treatment regimen.