To evaluate the effect of radiation dose escalation on overall survival (OS) for patients with nonmetastatic esophageal cancer treated with concurrent radiation and chemotherapy.

Patients diagnosed with stage I to III esophageal cancer treated from 2004 to 2012 were identified from the National Cancer Data Base. Patients who received concurrent radiation and chemotherapy with radiation doses of ≥50 Gy and did not undergo surgery were included. OS was compared using Cox proportional hazards regression and propensity score matching.

A total of 6854 patients were included; 3821 (55.7%) received 50 to 50.4 Gy and 3033 (44.3%) received doses >50.4 Gy. Univariate analysis revealed no significant difference in OS between patients receiving 50 to 50.4 Gy and those receiving >50.4 Gy (P=.53). The dose analysis, binned as 50 to 50.4, 51 to 54, 55 to 60, and >60 Gy, revealed no appreciable difference in OS within any group compared with 50 to 50.4 Gy. Subgroup analyses investigating the effect of dose escalation by histologic type and in the setting of intensity modulated radiation therapy also failed to reveal a benefit. Propensity score matching confirmed the absence of a statistically significant difference in OS among the dose levels. The factors associated with improved OS on multivariable analysis included female sex, lower Charlson-Deyo comorbidity score, private insurance, cervical/upper esophagus location, squamous cell histologic type, lower T stage, and node-negative status (P<.01 for all analyses).

In this large national cohort, dose escalation >50.4 Gy did not result in improved OS among patients with stage I to III esophageal cancer treated with definitive concurrent radiation and chemotherapy. These data suggest that despite advanced contemporary treatment techniques, OS for patients with esophageal cancer remains unaltered by escalation of radiation dose >50.4 Gy, consistent with the results of the INT-0123 trial. Furthermore, these data highlight that many radiation oncologists have not embraced the concept that dose escalation does not improve OS. Although local control, not investigated in the present study, might benefit from dose escalation, novel therapies are needed to improve the OS of patients with esophageal cancer.