Effect of Contralateral Esophagus Sparing in Hypofractionated Radiotherapy for Ultra-Central NSCLC

Abstract

Esophageal toxicity is a known complication of radiotherapy (RT) for ultra-central lung tumors, with ∼5 - 15% of patients developing grade ≥3 esophagitis. Despite increasing adoption of hypofractionated regimens, there are no consensus recommendations for reducing esophageal toxicity in these patients. We previously demonstrated the utility of a novel contralateral esophagus sparing technique (CEST) in a prospective trial of patients with locally-advanced NSCLC receiving concurrent chemoRT (70 Gy/7 weeks). Whether CEST is effective for hypofractionated RT is unknown. The aims of this study were to identify outcomes and toxicities associated with use of CEST in patients receiving definitive hypofractionated RT to ultra-central lung tumors. We retrospectively analyzed the records of consecutive patients with ultra-central NSCLC who received definitive RT to a dose of ≥60 Gy with fraction size ≥3 Gy per day, from 1/1/2015 to 6/30/2022. The esophageal wall contralateral to the tumor was contoured as an avoidance structure. VMAT planning was used to ensure steep dose fall-off across the esophagus while maintaining full tumor coverage. Treatment was delivered using daily cone-beam CT guidance. Toxicity was graded using CTCAE v 4.0. Descriptive statistics were used to analyze outcomes. The study was approved by the IRB. We identified 45 patients with ultra-central NSCLC treated with definitive RT, with median follow-up of 27.3 months. Median age was 74.5 years old [57.6 - 88.2]. Median ECOG PS was 2 [1 - 3]. Twenty-three patients had PTVs located within 1 cm of the esophagus. Median ITV and PTV volumes were 17 cc [1.6 - 154 cc] and 43.6 cc [8.4 - 292 cc], respectively. Median total dose was 60 Gy [60 - 67.5 Gy] in 20 [10 - 20] fractions. Median esophageal Dmax was 24 Gy [2 - 66.7 Gy] for all patients and 36 Gy [12.3 - 66.7 Gy] for patients with tumors located within 1 cm of the esophagus. For the CE, median Dmax, V20 Gy and V25 Gy were 35.7 Gy [17 - 45.2 Gy], 2.1 cc [0 - 6.7 cc] and 0.95 cc [0 - 3 cc], respectively. There were no episodes of grade ≥2 esophagitis (95% CI, 0-7.9%). The rate of grade 1 esophagitis was 13.3% (6.2 - 26.1%). Esophageal Dmax was higher in patients who developed esophagitis (44.5 vs. 26.1 Gy, t = 2.5, p = 0.038). Two patients developed grade 2 pneumonitis (4.4%, 95% CI 0 - 10.5%). There were no episodes of grade ≥3 pulmonary toxicity. There were 3 (6.7%) local recurrences, with 2-year local recurrence rate of 4.4%. Use of CEST to minimize esophageal dose for treatment of ultra-central NSCLC is associated with reduced rates of esophagitis compared with historical controls, without compromising local control. This technique is readily adaptable for clinical practice. More data are needed to refine the empirically derived CE dose-volume constraints, especially for tumors directly abutting the esophagus.