医学
肺
内科学
心脏病学
肺移植
结构效度
移植
构造(python库)
肺活量测定
肺容积
重症监护医学
放射科
外科
肺功能
作者
John R. Greenland,Ciara M. Shaver,Krishna Pandya,Michael S. Mulligan,Michaela R. Anderson,Daniel R. Calabrese,Jason D. Christie
标识
DOI:10.1016/j.healun.2026.01.006
摘要
INTRODUCTION: Acute lung allograft dysfunction (ALAD) has been proposed as a screening criterion to identify lung transplant recipients at increased risk for graft loss. ALAD was defined formally in a 2026 International Society for Heart and Lung Transplantation (ISHLT) Delphi consensus statement but has not been empirically evaluated. We sought to test the construct validity of the ISHLT ALAD definition by testing the hypothesis that ALAD is associated with increased risk of graft failure. METHODS: We included 86,885 spirometry measurements from 2,877 participants in the 8-center Lung Transplant Outcomes Group cohort. ALAD was defined as a ≥10% decline in FEV1 from the maximum value over the preceding 180 days and coded along with 90-day follow-up categories of resolved, persistent, progressive, and rapidly progressive. We assessed whether ALAD classifications diverged in predictions of time to death or re-transplantation in multivariable-adjusted time-dependent and landmark Cox proportional hazards models. RESULTS: ALAD had a prevalence of 15%. While resolution was seen in 40%, ALAD conferred a 4.93-fold graft failure hazard (95% CI 4.47-5.45, 7.4 years median follow-up), independent of concurrent chronic or baseline lung allograft dysfunction (CLAD or BLAD) status. Resolved ALAD was not associated with increased graft failure risk, but persistent, progressive, rapidly progressive, and no follow-up FEV1 ALAD categories were associated with increasing hazards for graft failure. A 10-percent threshold for FEV1 decline and a 180-day window to set the FEV1 baseline demonstrated optimal discrimination. CONCLUSION: The 2026 ISHLT spirometric ALAD definitions identify graft failure risk, supporting their relevance in clinical interventions and further research.
科研通智能强力驱动
Strongly Powered by AbleSci AI