Incidence of and Risk Factors for Immune Checkpoint Inhibitor-Related Pneumonitis in Thoracic Malignancies: A Prospective Multicenter Study

Background Immune checkpoint inhibitors (ICIs) have fundamentally changed cancer treatment, but ICI-related pneumonitis (ICI-P) remains a major clinical concern. Although numerous studies have reported its incidence and risk factors, the retrospective nature of these studies makes accurate assessment difficult in a real-world clinical setting. Patients and Methods This prospective multicenter observational study enrolled patients with advanced or recurrent thoracic malignancies treated with nivolumab, pembrolizumab, or atezolizumab. Thoracic high-resolution computed tomography (HRCT) examinations were performed at the start of ICI therapy and at the onset of ICI-P. The primary endpoint was ICI-P incidence. The secondary endpoints included evaluation of risk factors, with a focus on the lung lesions present at ICI initiation. Fine–Gray analyses with three models were used to identify risk factors. Results Among 614 eligible patients, 78 (12.7%) developed ICI-P. During a median observation period of 14.4 months, 29 patients developed grade ≥3 pneumonitis and 7 patients died owing to ICI-P. In the multivariable Fine–Gray analyses, independent risk factors for ICI-P included heavy smoking history (>50 pack-years; subdistribution hazard ratio [SHR]: 1.90; 95% confidence interval [CI]: 1.17–3.10), baseline interstitial lung abnormalities (ILAs; SHR: 3.42; 95% CI: 2.05–5.71), use of PD-1 inhibitors (SHR: 2.15; 95% CI: 1.28–3.63), and immunochemotherapy (SHR: 2.16; 95% CI: 1.17–3.98). Conclusion In this prospective real-world cohort, the incidence of ICI-P was relatively high, and ILAs detected by HRCT, heavy smoking history, and treatment with a PD-1 inhibitor alone or in combination with cytotoxic chemotherapy were identified as independent risk factors.