Longitudinal changes in pulmonary function following multiple courses of thoracic radiotherapy.

Abstract

e20103 Background: Pulmonary function decline following thoracic radiotherapy (RT), particularly DLCO, is well documented after single courses of RT in lung and esophageal cancer cohorts. However, in current practice patients increasingly undergo multiple distinct thoracic RT courses. Re-irradiation analyses to date focus on feasibility, toxicity, and dosimetry rather than functional longitudinal PFT endpoints linked to prior RT exposure at the time of testing. We investigated DLCO changes following multiple courses of thoracic RT. Methods: We performed a retrospective chart review of patients at our institution who received ≥2 distinct courses of thoracic RT, and we abstracted baseline and follow-up PFTs. Among 167 patients who underwent thoracic re-irradiation, 35 patients had paired pre- and post-treatment PFT data and formed our sample population (25 with 2 thoracic RT courses; 10 with 3 thoracic RT courses). Thoracic RT courses in this cohort were delivered between 2007 and 2024. Individual patient changes were assessed using paired tests (paired t-test or Wilcoxon signed-rank, as appropriate). We evaluated predictors of DLCO change using multivariable linear regression while adjusting for baseline DLCO %pred, age at baseline PFT, and sex. We also evaluated clinically meaningful DLCO decline, defined as DLCO ≥10 %-pred point decline or ≥15% relative decline, using binary endpoints. Results: Baseline PFTs preceded first RT completion by a median of 91 days (IQR 134), and follow-up PFTs were obtained a median of 620 days (IQR 966) after the most recent RT course preceding the post-RT PFT. Given variable follow-up timing, analyses were stratified by RT courses completed before the post-RT PFT. Patients with paired pre- and post-RT PFTs showed declines in DLCO (−9.8±18.8 %pred; n = 28; p = 0.010), FEV1 (−12.3±18.6 %pred; n = 35; p = 0.0004), and FVC (−8.2±17.0 %pred; n = 35; p = 0.0074) from baseline to their most recent follow-up. Unadjusted declines in DLCO were greater when PFTs were obtained following ≥2 RT courses versus 1 RT course (DLCO: −14.2±19.7 vs −2.8±15.6 %pred). In multivariable linear regression adjusting for baseline DLCO, age at baseline PFT, and sex, undergoing ≥2 thoracic RT courses was associated with a trend towards greater DLCO decline (β = −11.0 %pred points; 95% CI −22.6 to 0.5; p = 0.061). Clinically meaningful DLCO decline was also more frequent after ≥2 courses at the time of follow-up testing (DLCO ≥10 %-pred point decline: 58.8% vs 36.4%; DLCO ≥15% relative decline: 52.9% vs 27.3%). Conclusions: Greater longitudinal DLCO decline was observed with successive courses of thoracic radiotherapy. This study adds novel patient-level longitudinal pulmonary function data to the thoracic re-irradiation literature. Furthermore, these findings underscore the importance of pulmonary function surveillance strategies and can guide risk stratification with PFT thresholds for repeat thoracic RT.