Radiological distribution patterns in restrictive chronic lung allograft dysfunction: Impact on survival across all phenotypes

TL;DR

This study shows that the location of lung opacities in restrictive chronic lung allograft dysfunction affects patient survival after lung transplants.

Contribution

The study identifies radiological distribution patterns as a novel prognostic factor in restrictive chronic lung allograft dysfunction.

Findings

Restrictive CLAD has significantly worse survival than bronchiolitis obliterans syndrome.

Diffuse/lower-predominant RLO distribution is associated with the poorest survival in restrictive CLAD.

The extent of RLOs does not correlate with survival within distribution patterns.

Abstract

Restrictive chronic lung allograft dysfunction (CLAD) demonstrates poor outcomes after lung transplantation. However, the impact of radiological patterns on survival within a restrictive CLAD under the new International Society for Heart and Lung Transplantation (ISHLT) criteria remains unclear.

We analyzed retrospectively 241 bilateral lung transplant recipients between 2005 and 2021. CLAD was diagnosed and classified per the 2019 ISHLT criteria. Restrictive phenotype included restrictive allograft syndrome (RAS) and mixed phenotype. In these cases, RAS-like opacities (RLOs) were evaluated both qualitatively and semiquantitatively on computed tomography at CLAD diagnosis. RLOs were classified into upper-predominant and diffuse/lower-predominant distribution groups. Overall survival after CLAD diagnosis was assessed using Kaplan-Meier method with log-rank test and Cox proportional hazards models.

Eighty-three patients were diagnosed with CLAD after transplantation. Twenty-one (25.3%) had restrictive phenotype, which showed shorter survival compared to bronchiolitis obliterans syndrome (median survival: 19.8 vs 68.1 months; hazard ratio [HR], 4.53; 95% confidence interval [CI], 1.96-10.49; p < 0.001). Within the restrictive phenotype, the upper-predominant group demonstrated longer survival than the diffuse/lower-predominant group (median survival: 61.1 vs 15.5 months; p = 0.008). The diffuse/lower-predominant group had shorter survival compared to any other CLAD phenotype (HR, 8.45; 95% CI, 3.40-21.04; p < 0.001). The extent of RLOs within each distribution pattern was not significantly associated with survival.

In restrictive phenotype CLAD, RLO distribution patterns determined survival outcomes, with diffuse/lower-predominant showing the poorest prognosis, while the extent of RLOs within each pattern did not correlate with prognosis.

Figures