Lung transplantation (LT) is the only therapeutic option that can improve survival and quality of life of selected patients with severe chronic respiratory disease. Despite constant improvements, the survival of patients after lung transplantation remains limited and several studies have questioned the benefit of lung transplantation on the survival of patients with cystic fibrosis, as well as other chronic respiratory diseases, although the statistical methods used were arguable. Assessing the survival benefit of lung transplantation and factors associated with this benefit thus remains an important objective for patients with cystic fibrosis, even in a context where graft allocation scores have been implemented to optimize transplantation. The main issues are to determine which patients should be registered on the waiting list for a lung transplant, the adequate time of registration, which patients should be transplanted and the adequate time of transplantation.
The main objective of this project is to determine the parameters associated with the benefit of lung transplantation in patients with cystic fibrosis.
The secondary objectives are:
– To quantify the marginal benefit of LT in patients with cystic fibrosis
– To estimate the benefit of LT on quality of life-adjusted survival
– To compare different statistical methods to achieve these objectives
In the absence of randomized clinical trials, the evaluation of the benefit of lung transplantation relies on statistical modeling of observational data. We will use data from the International Society for Heart & Lung Transplantation (ISHLT) for that purpose. The methods used must however take into account the observational nature of the data, while producing results that may have a causal interpretation, i.e. that allow concluding at a benefit of lung transplantation by itself, and not of lung transplantation associated with the decision to transplant. In this context, we identified six approaches to be compared including models for the hazard of death (joint models for longitudinal and survival data, marginal structural models, structural nested models, sequential stratification) and two types of models for the mean survival. Then a measure of quality-adjusted survival will be derived to account quality-of-life of patients when estimating the benefit of LT.
From a clinical point of view, this project will assess the benefit of lung transplantation and the factors associated with this benefit, using the most up-to-date methods adapted to the nature of the data. From a methodological point of view, it will enable to determine the advantages and limitations of the proposed approaches for analysis of the benefit of lung transplantation.