Clinical and Dosimetric Factors Predicting Grade ≥2 Radiation Pneumonitis After Postoperative Radiotherapy for Patients With Non-Small Cell Lung Carcinoma.

PURPOSE:To identify clinical and dosimetric factors that would predict grade ≥2 radiation pneumonitis (RP) for patients undergoing postoperative radiation therapy (PORT) for non-small cell lung cancer (NSCLC); and to use the factors identified to generate a predictive model to quantify risk of RP in such patients. METHODS AND MATERIALS:We retrospectively reviewed radiation therapy, radiographic, and clinical data from 199 patients who had received PORT, with or without chemotherapy, for NSCLC. Potential associations between dosimetric and clinical factors and RP were evaluated in univariate and multivariate Cox regression hazard models and competing risk analysis. Kaplan-Meier analysis was used to estimate overall survival and the cumulative incidence of RP, and receiver operating characteristic curve analysis to identify cutpoints for variables found to influence RP risk. The endpoint was grade ≥2 RP (symptomatic, requiring steroids or limiting instrumental activities of daily living). RESULTS:Thirty-seven patients (19%) developed grade ≥2 RP. Patient-related factors, type of surgery or chemotherapy, and radiation therapy-related factors were not associated with grade ≥2 RP; only lung V10 > 30% and lung V20 > 20% predicted grade ≥2 RP. Risk groupings were as follows: high risk, V10 > 30% and V20 > 20% (24 of 72 patients, 33%); intermediate risk, V10 > 30% and V20 ≤ 20% or V10 ≤ 30% and V20 > 20% (6 of 26 patients, 23%); and low risk, V10 ≤ 30% and V20 ≤ 20% (6 of 101 patients, 6%) (P < .0001). In a subgroup analysis of patients who had had lobectomy, corresponding incidences of RP were as follows: high risk, 20 of 59 (34%); intermediate risk, 5 of 22 (23%); and low risk, 6 of 70 (9%) (P = .001). CONCLUSIONS:The lung dose-volume variables V10 and V20 predicted risk of grade ≥2 RP among patients who underwent PORT for NSCLC.