Reconceptualizing Disease Progression in Pulmonary Fibrosis - PROJECT SUMMARY/ABSTRACT Pulmonary Fibrosis (PF) is a progressive, fatal lung disease. Two FDA-approved treatments modestly reduce functional decline but are poorly tolerated and do not relieve symptoms. A major barrier to developing new and better treatments is that the primary endpoint in clinical trials, change in the forced vital capacity, varies significantly and is not suitable for early-stage disease. To gain a more complete understanding of progression across the entire disease development history, the Familial Pulmonary Fibrosis (FPF) Registry was established. The Registry enrolls proband with clinical (symptomatic) PF and their asymptomatic relatives, who are “At-Risk” to develop FPF. At-risk relatives participate in an ongoing, prospective study to screen for progressive subclinical PF. In a preliminary analysis of FPF Registry data, collectively representing early subclinical through advanced PF, the progressive decline in several pulmonary function tests (PFTs) was evaluated after disease stage alignment using a Bayesian approach. Counter to the prevailing view that PF progression is individual-inherent, our data suggests that the rate and timing of progression on various PFTs is conserved across individuals and highly stage-specific. The diffusion capacity of lung for carbon monoxide begins declining 10 years before clinical disease onset, while forced vital capacity is normal initially but declines precipitously 5 years after clinical disease onset. Notably, no PFT parameter declines uniformly across the disease development history, so any trial whose analysis ignores disease stage will be inherently inefficient. A Bayesian approach can also be used to conduct a treatment efficacy analysis, measuring the proportional treatment effect on progression of one or more endpoints, after accounting for disease stage. Experience with other rare, slowly progressive diseases suggests that this approach can have a huge impact on trial efficiency. Our hypothesis is that patients with PF experience a conserved, sequenced, and stage- specific progression trajectory. Specific Aim 1) Fit and validate a Bayesian disease progression model defining the conserved progression in PF. Domains on which progression is evaluated will include physiological (PFTs), anatomical (quantitative lung imaging), and biological (circulating biomarkers). Aim 2) Test the utility of incorporating a Bayesian approach to PF trial design. The results of Aim 1 will be used to design trials with stage-specific population selection criteria, stage-appropriate endpoint(s), and/or a Bayesian statistical approach to account for disease stage. The operating characteristics (trial size, type 2 error) will be simulated and compared to traditional trial designs. Trial designs applicable in both subclinical and clinical disease will be simulated. These transformative results will ultimately lead to more effective therapy for this devastating disease by: 1) proposing novel trial designs that substantially improve trial efficiency; 2) enabling disease prevention trials; and 3) providing a template to understand the optimal timing of treatment initiation, ultimately supporting earlier diagnosis of PF.
