Chemoradiation-induced pneumonitis in patients with unresectable stage III non-small cell lung cancer: A meta-analysis.

Abstract

e20532 Background: Recent clinical trials have shown positive results for therapies combining concurrent chemoradiation therapy (cCRT) and checkpoint immunotherapy in unresectable non-small cell lung cancer (NSCLC). cCRT is associated with an increased risk of pneumonitis, a severe and life-threatening inflammation of the lungs. To further inform clinical decision-making and support the evaluation of new therapies combining immunotherapies with cCRT, it is important to understand the baseline risk of pneumonitis associated with cCRT alone. The objective of this study was to quantify the incidence of cCRT-induced grade 3−5 pneumonitis (immune-mediated and radiation pneumonitis) in unresectable stage III NSCLC patients. Methods: A systematic literature review and meta-analysis were performed in accordance with PRISMA guidelines. MEDLINE, Embase, and the Cochrane Central Register were searched from 2014 to April 24, 2020. Chemotherapies of interest were cisplatin, pemetrexed, etoposide, carboplatin, and paclitaxel. Randomized controlled trials (RCTs), observational studies, and non-randomized trials were included. Bayesian meta-analysis using a binomial model random effects model was conducted with SAS 9.4. Results: Among 1,889 records identified from the search, 17 studies (6 RCTs, 8 observational studies, 3 single-arm trials) met inclusion criteria. Eleven studies were included in the meta-analysis (5 RCTs, 6 observational studies; 1,788 patients). All studies specified radiation-related pneumonitis (RP), although this is clinically indistinguishable from immune-mediated pneumonitis. Patient populations were comparable across studies; the most common chemotherapies were paclitaxel + carboplatin (n = 6) and pemetrexed + cisplatin (n = 5), and radiation doses ranged from 60–74 Gy. There was variation across studies in intervention, outcome reporting, and follow-up (median range: 12–73 months), but this variation was considered acceptable based on sensitivity analyses. The estimated pooled incidence of grade 3−5 RP in cCRT-treated unresectable stage III NSCLC patients was 3.62% [95% confidence interval (CI): 1.65−6.21] in RCTs and 5.98% [95% CI: 2.26−12.91] in observational studies. The pooled incidence of fatal (grade 5) RP was 0.37% [95% CI: 0−2.78] in RCTs and 1.73% [95% CI: 0.53−4.33] in observational studies. Conclusions: This study estimates that 3.62–5.98% of patients with unresectable stage III NSCLC develop grade 3−5 RP when treated with cCRT, with incidence varying by study design. Estimates of RP incidence were higher in the real-world setting compared to RCTs. These results can be used to contextualize the baseline risk of cCRT-induced pneumonitis in unresectable stage III NSCLC to better understand the adverse event of pneumonitis associated with novel immunotherapy treatments indicated for concomitant use with this modality.