Acute esophageal toxicity in non-small cell lung cancer patients after high dose conformal radiotherapy.

BACKGROUND AND PURPOSE To correlate acute esophageal toxicity with dosimetric and clinical parameters for non-small cell lung cancer (NSCLC) patients treated with radiotherapy (RT) alone or with chemo-radiotherapy (CRT). PATIENTS AND METHODS We analyzed the data of 156 patients with medically inoperable or locally advanced NSCLC. Seventy-four patients were irradiated with high dose RT only, 45 patients with sequential CRT (Gemicitabine/Cisplatin) and 37 patients with concurrent CRT (Cisplatin daily 6 mg/m(2)). The radiation dose delivered ranged from 49.5 to 94.5 Gy (2.25-2.75 Gy per fraction) with an overall treatment time of 5-6 weeks. For all patients the maximal acute esophageal toxicity (RTOG/EORTC criteria) was scored and related to dose-volume parameters, as well as to clinical and treatment-related parameters. All parameters were tested univariable and multivariable in a binary logistic regression model. The toxicity data of a homogeneous subgroup was fitted to the Lyman-Kutcher-Burman model. RESULTS Grade 2 acute esophageal toxicity or higher occurred in 27% (n=42) of the patient population of which nine patients developed grade 3 toxicity and one patient grade 4. All 10 patients with grade>or=3 esophageal toxicity received concurrent CRT. At multivariable analysis, the most significant clinical parameter to predict acute esophageal toxicity was the concurrent use of CRT. The most significant dosimetric parameter was the esophagus volume that received at least 35 Gy. The data of the patients who did not receive concurrent CRT were well described by the Lyman-Kutcher-Burman normal tissue complication probability model. The optimal fit of the data of non-concurrent treated patients to this model was obtained using the following values for the parameters: TD(50)=47 Gy (41-60 Gy), n=0.69 (0.18-6.3) and m=0.36 (0.25-0.55) where the numbers between brackets denote the 95% confidence interval. Acute esophageal toxicity was not significantly increased for patients treated with sequential CRT. CONCLUSION Both concurrent CRT and the volume that receives at least 35 Gy were predictors of acute esophageal toxicity.