ObjectivesConcurrent chemoradiotherapy is the mainstay treatment for NSCLC stage III disease. To investigate whether radiation dose escalation based on individual normal tissue constraints can improve outcome, the Swedish lung cancer study group launched this randomized phase II trial. Materials and MethodsNSCLC patients with stage III disease, good performance status (0–1) and adequate lung function (FEV1 > 1.0 L and CO diffusion capacity > 40%) received three cycles of cisplatin (75 mg/m2 day 1) and vinorelbine (25 mg/m2 day 1 and 8) every third week. Radiotherapy started concurrently with the second cycle, with either 2 Gy daily, 5 days a week, to 68 Gy (A) or escalated therapy (B) based on constraints to the spinal cord, esophagus and lungs up to 84 Gy by adding an extra fraction of 2 Gy per week. ResultsA pre-planned safety analysis revealed excessive toxicity and decreased survival in the escalated arm, and the study was stopped. Thirty-six patients were included during 2011–2013 (56% male, 78% with adenocarcinoma, 64% with PS 0 and 53% with stage IIIB). The median progression-free survival (PFS) and overall survival (OS) were 11 and 17 months in arm B compared to the encouraging results of 28 and 45 months in the standard arm. The 1- and 3-year survival rates were 56% and 33% (B) and 72% and 56% (A), respectively. There were seven toxicity-related deaths due to esophageal perforations and pneumonitis: five in the escalated group and two with standard treatment. ConclusionDose-escalated concurrent chemoradiotherapy to 84 Gy to primary tumor and nodal disease is hazardous, with a high risk of excessive toxicity, whereas modern standard dose chemoradiotherapy with proper staging given in the control arm shows a promising outcome with a median survival of 45 months and a 3-year survival of 56% (NCT01664663).