A risk model for lung complication combining radiation therapy and chronic obstructive pulmonary disease

To develop a multi-variate risk model of lung complication for the treatment of locally advanced lung cancer combining radiation dosimetry and patient-specific risk factors including chronic obstructive pulmonary disease (COPD).

A retrospective study of 89 advanced lung cancer patients including clinical variables and radiation dose-volume metrics was correlated to overall survival (OS), local progression-free survival (LPFS), and lung complications. Lung toxicities were defined as grade 2 or higher lung complication including persistent cough requiring narcotic or antitussive agents, and dyspnea with minimal effort but not at rest. Grade 2 or higher radiation pneumonitis (RP) was defined clinically or radiographically. Each observed complication was correlated to the set of patient-specific factors via regression.

Age was the dominant factor in survival analysis; in patients > 70, 2-year survival was 28% (N = 18) vs. 59% in patients < 70 (N = 71, HR = 3.9, p = 0.004). OS and LPFS were not significantly different based on with COPD status (p = 0.2, HR = 1.5) or radiation dose (range 40–74 Gy, median = 60 Gy, p > 0.5, HR < 1.3). In predicting all observed lung complications, multi-variate stepwise logistic regression revealed COPD status (p = 0.01) and a total lung V50 (p = 0.02) as significant. Separation of G2 or higher lung complication and RP showed a different trend, with G2 or higher lung complication statistically associated with age (p = 0.02) and COPD (p = 0.03). Radiation pneumonitis was only associated with total lung volume at 20 Gy (V20, p = 0.003), with a resulting model risk (RP) = 1.4 (V20-0.21).

COPD status was not associated with survival or radiation pneumonitis. Age was the dominant factor in survival, and total lung V20 was associated with risk of RP.