Pulmonary epithelial markers in phenotypes of chronic lung allograft dysfunction

Background: Airway epithelial injury is thought to be a key event in the pathogenesis of chronic lung allograft dysfunction (CLAD). We investigated whether markers of epithelial activity and injury in bronchoalveolar lavage fluid (BAL) correlate with CLAD diagnosis and major CLAD phenotypes: bronchiolitis obliterans syndrome (BOS) vs restrictive allograft syndrome (RAS)-related phenotypes (including RAS, mixed phenotype, and all other patients with RAS-like opacities).

Methods: CLAD status and phenotypes were retrospectively determined in a cohort of all consecutive adult, first, bilateral lung transplants performed 2010-2015, with available BAL samples. All patients with RAS-related phenotypes were included and 1:1 matched with BOS patients based on the time from transplant to CLAD-onset. Subjects who were CLAD-free for a minimum of 3 years post-transplant were 1:1 matched to CLAD patients and included as controls. Proteins that maintain the barrier function of the airway epithelial mucosa (club cell secretory protein, surfactant protein-D and epithelial mucins: MUC1, MUC5AC, MUC5B, MUC16), as well as epithelial cell death markers (M30&M65 representing epithelial cell apoptosis and overall death, respectively), were measured in BAL obtained within 6-months post CLAD onset using a double-sandwich ELISA or a multiplex bead assay. Protein levels were compared using Mann-Whitney-U-test. Association between protein levels and graft survival was assessed using Cox proportional hazards models, adjusted for CMV serology mismatch status and CLAD phenotype.

Results: Fifty-four CLAD (27 BOS, 11 RAS, 7 mixed, 9 others with RAS-like opacities) patients and 23 CLAD-free controls were included. Median BAL levels were significantly higher in patients with CLAD compared to CLAD-free controls for M30 (124.5 vs 88.7 U/L), MUC1 (6.8 vs 3.2 pg/mL), and MUC16 (121.0 vs 30.1 pg/mL). When comparing CLAD phenotypes, M30 was significantly higher in patients with RAS-related phenotypes than BOS (160.9 vs 114.6 U/L). In multivariable models, higher M30 and MUC5B levels were associated with decreased allograft survival after CLAD onset independent of phenotype (p < 0.05 for all).

Conclusions: Airway epithelial mucins and cell death markers are enhanced in the BAL of patients with CLAD and can assist in differentiating between CLAD phenotypes and post-CLAD outcomes. Abnormal airway mucin expression and epithelial cell death may be involved in the pathogenesis of CLAD, and therefore their detection may aid in future selection of targeted therapies.