Cumulative incidence of and predictive factors for lung cancer in IPF

ABSTRACT Background and objective: Previous studies have indicated a high incidence of lung cancer in IPF, and some have identified its risk factors. However, those studies were retrospective and the clinical characteristics of IPF patients developing lung cancer were evaluated only when those patients had developed the cancer. The true cumulative incidence of lung cancer after the diagnosis of IPF and its predictive factors at the initial diagnosis of IPF remain unknown. The present study was conducted to elucidate the cumulative incidence and risk factors for lung cancer in IPF patients by retrospective longitudinal cohort analysis. Methods: The study group consisted of 103 IPF patients without lung cancer at the time of their initial diagnosis. The cumulative incidence of lung cancer was estimated using the Kaplan–Meier method. The strength of association between several variables present at the initial diagnosis of IPF and the development of lung cancer was assessed using Cox proportional hazards regression analysis. Results: A total of 21 (20.4%) patients with IPF developed lung cancer during the observation period. The cumulative incidence rate of lung cancer increased as the duration of follow up increased (3.3%, 15.4% and 54.7% at 1, 5 and 10 years, respectively). Univariate Cox proportional hazards regression analysis showed that age and smoking at the initial diagnosis of IPF were significantly associated with lung cancer. Multivariate Cox proportional hazards regression analysis indicated that age at initial diagnosis was an independent significant factor predicting lung cancer. Conclusions: The cumulative incidence of lung cancer increased over time and age at diagnosis of IPF was independently associated with development of lung cancer.