Phenotyping CLAD after single lung transplant: Limits and prognostic assessment of the 2019 ISHLT classification system

Background: Phenotyping chronic lung allograft dysfunction (CLAD) in single lung transplant (SLTX) is challenging, due to the native lung contribution to pulmonary function test (PFT). We aimed to assess the applicability and prognostic performance of International Society for Heart and Lung Transplantation (ISHLT) classification in SLTX. Methods: In this retrospective study of adult, first, SLTX performed 2009-2017, patients with persistent drop in FEV1≥20% were assessed by 2 independent adjudicators to determine CLAD status and phenotype. Interobserver agreement (IOA) was calculated (Cohen's Kappa) for CLAD, phenotype and presence of RAS (resttrictive allograft syndrome)-like opacities (RLO). Association of CLAD phenotypes with time to death or retransplant (ReTx), adjusted for age at SLTX, sex, CMV mismatch and native lung condition, were assessed using Cox proportional hazards models. Results: Of 172 SLTX recipients, 92 experienced a persistent drop in FEV1>20%. Following adjudication, 67 were diagnosed with CLAD. We noted a moderate IOA for CLAD diagnosis (Kappa 0.69) and poor IOA for phenotype adjudication (Kappa 0.52). The final phenotype adjudication was 31 bronchiolitis obliterans syndrome (BOS) (46.3%), 13 RAS (19.4%), 2 mixed (3%), 2 Undefined (3%), and 19 remained Unclassified (28.3%). Using these adjudicated phenotypes, RAS was significantly associated with a higher risk of death/ReTx compared to other groups (HR 2.98, 95%CI [1.39-6.4]). The adjudication of RLO had the best IOA (Kappa 0.73). The presence of RLO was a strong predictor of death or ReTx (HR 2.37, 95%CI [1.2-4.5]), regardless of the final phenotype. Conclusions: PFT interpretation is challenging in SLTX. A classification essentially relying on imaging, which harbored good IOA, obtained better prognostic performance than a classification using published physiological cut-offs.