Is pulmonary artery a dose-limiting organ at risk in non-small cell lung cancer patients treated with definitive radiotherapy?

PURPOSE Our previous study suggested that some pulmonary artery (PA) dosimetric parameters were associated with mortality in unresectable non-small cell lung cancer (NSCLC) treated with definitive radiotherapy. The present study aims to analyze the impact of both PA and heart dosimetric parameters on survival of patients with NSCLC treated with definitive conventional fractionated radiotherapy (CFRT) in another independent research center and further determine whether the PA should be considered a dose-limiting organ at risk (OAR) for patients receiving thoracic CFRT. METHODS We performed a retrospective analysis of successive patients with medically inoperable or unresectable NSCLC treated with definitive radiotherapy or chemoradiotherapy from August 2010 to September 2014. Clinical and pathological information, PA and heart dosimetric factors, and follow-up data were collected from each patient's records and evaluated as potential prognostic factors for survival. Survival probabilities were estimated by the Kaplan-Meier method and compared by the log rank test. Cox proportional hazards regression models were performed to determine the independent predicators of survival. The optimal cutoff points of continuous dosimetric variables were determined by Youden index in receiver operating characteristic (ROC) analysis. RESULTS This study analyzed the records of 141 patients, 50.4% had adenocarcinoma, 71.6% had stage III disease, and 55% patients received concurrent chemoradiotherapy. Radiation dose ranged from 60 to 76 Gy in 30-38 fractions. Median follow up was 16.9 months. Median overall survival (OS) was 20.5 months (95% confidence interval [CI] 10.3-30.7 months), and 1-, 2-, 3-year OS rates were 75.2%, 58.2% and 56%, respectively. Univariate and multivariate analysis showed that Karnofsky Performance Status (KPS) score, Charlson's Comorbidity Index (CCI), T and N stage, PA invasion grade and the percentage of PA volume that received 40 to 55 Gy (PA V40-55) were significantly associated with OS. No significant associations were found between heart dosimetric factors and OS. Median OS of patients with PA invasion grade 0, 1, 2, and 3 were 41.8, 27.8, 12.7 and 7.5 months, respectively (P < 0.001). PA V40, V45, V50 and V55, using thresholds of 80%, 68%, 45%, and 32%, respectively, were independent predictors for OS. CONCLUSIONS PA invasion grade and PA V40-55 appear associated with OS in patients with NSCLC treated with definitive CFRT. We propose that PA be considered as a dose-limiting OAR for such patients.