Assessing treatment outcomes in CLAD: The Hannover-extracorporeal photopheresis model

Abstract

Background: Chronic lung allograft dysfunction (CLAD) is a leading cause of graft loss in lung transplantation. Despite this, convincing treatment data is lacking, and protocols vary widely between centers. CLAD phenotypes exist, but phenotype transitioning has increased the challenge of designing clinically relevant studies. Extracorporeal photopheresis (ECP) has long been a suggested salvage treatment, but efficacy appears unpredictable. This study describes our experiences with photopheresis, using novel temporal phenotyping to illustrate the clinical course.

Methods: Retrospective analysis of patients completing ≥3 months of ECP for CLAD between 2007 and 2022 was performed. A latent class analysis employing a mixed-effects model was performed, deriving patient subgroups based on spirometry trajectory over the 12 months prior to photopheresis until graft loss or 4 years post photopheresis initiation. The resulting temporal phenotypes were compared in terms of treatment response and survival outcomes. Linear discriminatory analysis was used to assess phenotype predictability, relying solely on data available at photopheresis initiation.

Results: Data from 5,169 outpatient attendances in 373 patients was used to construct the model. Five trajectories were identified, with uniform spirometry changes evident following 6 months of photopheresis. Outcomes were poorest in Fulminant patients (N = 25, 7%) with median survival of 1 year. In the remainder, poorer lung function at initiation led to poorer outcomes. The analysis revealed important confounders, affecting both decision-making and outcome interpretation.

Conclusions: Temporal phenotyping provided novel insights into ECP treatment response in CLAD, particularly the importance of timely intervention. Limitations in % Baseline values in guiding treatment decisions warrant further analysis. Photopheresis may have a more uniform effect than previously thought. Predicting survival at ECP initiation appears feasible.

Keywords: CLAD; lung transplant; mixed-modelling; phenotypes; photopheresis.