Dose-Escalated Definitive Chemoradiotherapy with Intensity-Modulated Radiation Therapy Versus Esophagectomy in Locally Advanced Resectable Esophageal Squamous Cell Carcinoma

W. H. Lee,D. J. Kim,S. Y. Park,Y. C. Lee,S. K. Lee,S. K. Shin,B. C. Cho,H. R. Kim,M. H. Hong,B. M. Lee,A. J. Yang,H. K. Byun,T. H. Kim,C. G. Lee

DOI: https://doi.org/10.1016/j.ijrobp.2020.07.1965

2020-01-01

Abstract:Esophagectomy is the main treatment modality for resectable locally advanced esophageal squamous cell carcinoma (ESCC). Dose-escalated definitive concurrent chemoradiotherapy (dCRT) studies using outdated techniques have failed. In this study, we compared dose-escalated dCRT with IMRT to esophagectomy in locally advanced, resectable and medically operable ESCC. We identified 124 patients of potentially resectable (American Joint Committee on Cancer [AJCC] 8th edition stage II-IVA) and medically operable locally advanced ESCC from July 2015, when IMRT was first applied for ESCC, to June 2018. Those who refused surgery, considered as high-risk for surgery, or diagnosed with cervical esophageal tumors received dCRT. Radiation field included the primary tumor with 2-3 cm proximal/distal margin and involved lymph nodes with simultaneous integrated boost. The survival outcomes and treatment-related toxicities were primary endpoints. Among 124 patients, 39 patients (31.5%) and 85 patients (68.5%) received dCRT with IMRT and esophagectomy, respectively. The dCRT group consisted of patients with older age, poorer ECOG PS, more ≥T3 stage, and more cervical esophageal tumors compared to the esophagectomy group (all p<0.05). In the esophagectomy group, 64 patients received preoperative and/or adjuvant chemotherapy and/or radiation. In the dCRT group, 69.2% received radiation dose of ≥60.0 Gy, with median dose of 63.0 Gy (range, 50.4-78.0). With median follow-up of 18.5 months (range, 2.3-47.8), the 2-year overall survival (OS) and locoregional failure (LRF) for dCRT (vs. esophagectomy) were 55.0% (vs. 66.9%, p = 0.14) and 32.4% (vs. 30.7%, p = 0.32), respectively. In multivariate Cox regression, significant adverse predictors for OS were ≥T3 stage (p = 0.05) and ECOG PS 1 (vs. 0) (p = 0.01), while only ≥T3 stage (p = 0.05) was predictive of LRF. Treatment modality was not predictive of either OS or LRF. The differences in patterns of first failure were insignificant. Treatment-related mortality and any ≥G3 acute toxicities was reported in 2 patients (5.1%) and 8 patients (20.5%) for dCRT, and 5 patients (5.9%) and 30 patients (35.3%) in esophagectomy group, respectively. Main ≥G3 late toxicities after dCRT were esophageal stricture (12.8%) followed by fistula or perforation (10.3%). After esophagectomy, ≥G3 late toxicities were mainly small bowel obstruction (7.1%), anastomosis leakage (5.9%), chyloperitoneum/thorax (5.9%), and esophageal stricture (9.4%). Dose-escalated dCRT with IMRT showed comparable survival outcomes to esophagectomy with tolerable toxicities, despite unfavorable baseline characteristics. This treatment is worth considering to offer chance for organ preservation in patients with locally advanced ESCC.