Mitigation and treatment of radiation-induced thoracic injury with a cyclooxygenase-2 inhibitor, celecoxib

Purpose: To test whether a cyclooxygenase-2 inhibitor (celecoxib) could reduce mortality resulting from radiation-induced pneumonitis. Methods and Materials: Celecoxib was given to mice twice daily for 40 consecutive days starting on the day of local thoracic irradiation (LTI) or 40 or 80 days later. C3Hf/KamLaw mice were observed for morbidity, and time to death was determined. Results were analyzed using the Cox proportional hazards model. Results: Timing of celecoxib relative to LTI determined efficacy. A significant reduction in time to death was achieved only when celecoxib was started 80 days after LTI, corresponding to the time when pneumonitis is expressed. For these mice the reduction in mortality was quantified as a hazard ratio for mortality of treated vs untreated of 0.36 (95% confidence interval [CI] 0.24-0.53), thus significantly less than 1.0. Correspondingly, the median lethal dose for treated mice (12.9 Gy; 95% CI 12.55-13.25 Gy) was significantly (P=.026) higher than for untreated mice (12.4 Gy; 95% CI 12.2-12.65 Gy). Conclusions: Celecoxib significantly reduced lung toxicity when administered months after LTI when the deleterious effects of radiation were expressed. The schedule-dependent reduction in fatal pneumonitis suggests that celecoxib could be clinically useful by reintroduction of treatment months after completion of radiation therapy. These findings may be important for designing clinical trials using cyclooxygenase-2 inhibitors to treat radiation-induced lung toxicity as a complement to concurrent radiation therapy of lung cancers.