Predictors and prognostic stratification of bronchiolitis obliterans syndrome after pediatric allogeneic hematopoietic stem cell transplantation

Abstract

Abstract Background Bronchiolitis Obliterans Syndrome (BOS) following pediatric allogeneic hematopoietic stem cell transplantation (allo-HSCT) is associated with high morbidity and mortality, largely due to diagnostic delays and static prognostic assessments. This study aimed to validate the value of a strategy based on dynamic pulmonary function monitoring for early warning and prognostic stratification in the management of pediatric BOS. Methods This single-center, retrospective cohort study included 201 children who underwent allo-HSCT between January 2015 and December 2024, comprising 29 patients with BOS and 172 non-BOS controls. Independent risk factors for BOS were identified using multivariate Cox regression analysis. We evaluated the diagnostic performance of an "at-risk stage" warning signal based on the relative decline in Forced Expiratory Volume in 1 second (FEV1). In the longitudinal cohort of BOS patients, the independent predictive value of the FEV1 decline rate within the first 3 months post-diagnosis for progression-free survival (PFS) was investigated. The longitudinal trajectories of pulmonary function and high-resolution computed tomography (HRCT) imaging scores were also compared. Results The cumulative incidence of pediatric BOS at our center was 6.9%. Multivariate analysis identified grade III-IV acute graft-versus-host disease (aGVHD) (HR = 15.12), chronic GVHD (cGVHD) (HR = 6.94), dry cough (HR = 11.77), and an early post-transplant FEV1/FVC decline of ≥ 10% (HR = 8.70) as independent risk factors for BOS development. Using a relative FEV1 decline of ≥ 10% (termed BOS stage 0p) as an early warning signal advanced the diagnostic window by a mean of 176 days and demonstrated a high negative predictive value of 84.95%. For prognosis, a rapid FEV1 decline (≥ 25%) within the first 3 months post-diagnosis was a strong independent predictor for inferior PFS (adjusted HR = 30.68, 95% CI: 3.39-277.79, P = 0.002), independent of the NIH severity grade at diagnosis. Furthermore, a significant "radio-functional discordance" was revealed, where HRCT structural damage scores continued to worsen even as pulmonary function stabilized in some patients (P < 0.001). Conclusion This study confirms that a strategy based on dynamic pulmonary function monitoring effectively addresses key challenges in pediatric BOS management. Monitoring the relative decline in FEV1 serves as a sensitive early warning tool, significantly advancing the diagnostic window. Moreover, assessing the early FEV1 decline rate post-diagnosis provides a powerful dynamic tool for achieving precise prognostic stratification and guiding individualized treatment escalation. The evaluation of BOS treatment response should concurrently address functional improvement and potential underlying structural deterioration.