Predictors of chronic obstructive pulmonary disease progression

Proposal
1692

Title of Proposed Research

Lead Researcher

Akihiro Hisaka

Affiliation

Laboratory of Clinical Pharmacology and Pharmacometrics, Graduate School of Pharmaceutical Sciences, Chiba university, Chiba, Japan

Funding Source

None

Potential Conflicts of Interest

None

Data Sharing Agreement Date

13 September 2017

Lay Summary

Chronic obstructive pulmonary disease (COPD) is a chronic inflammatory lung syndrome that affects millions of people, and it is estimated to become the third leading cause of death worldwide in 2030. Although it is well known that COPD is characterized by a progressive and life-long airflow limitation, the rate of decline in lung function is extremely variable across patients.
The lung function of patients with COPD is confirmed using a pulmonary function test called “spirometry”. A key value of spirometry is the forced expiratory volume in 1 s (FEV1), which is used to stage COPD severity according to the Global Initiative for Chronic Obstructive Lung Disease guidelines. People with COPD demonstrate a decline in FEV1 over time, and the FEV1 can be crucial in choosing the best treatment option and predicting prognosis.

Previously, a number of clinical studies have been conducted to analyze FEV1 change and its influencing factors in patients with COPD. However, few studies have focused on lifelong FEV1 change because it is practically difficult to perform a cohort study over several decades. Moreover, although some investigations of an FEV1 prediction model have been performed, no model with the ability to predict lifelong changes in FEV1 in patients with COPD has been developed. Therefore, the aim of this study is to develop a lifelong FEV1 model and identify its influencing factors. Because it is practically difficult to perform a cohort study over several decades, we are planning to assume decades-long FEV1 longitudinal change by using a model-based meta-analysis with Statistical Restoration of Fragmented Time-course (SReFT). SReFT is a new method we previously developed to restore long-term time courses from numerous short fragments by using an extended nonlinear mixed-effects estimation method.

This study would allow the prediction of lifelong FEV1 change in COPD patients based on their background. It would provide useful information in predicting prognosis and choosing the best treatment option for each patient with COPD. In addition, this study is expected to propose a novel and efficient approach for understanding disease progression in chronic diseases. SReFT can be applied not only to COPD but to almost all chronic diseases. With SReFT, we will be able to assume long-term disease progression without conducting long-term clinical studies.

Study Data Provided

[{ "PostingID": 1413, "Title": "BI-205.372", "Description": "Tiotropium / Respimat One Year Study in COPD." },{ "PostingID": 3876, "Title": "BI-1237.5", "Description": "Tiotropium+Olodaterol Fixed Dose Combination (FDC) Versus Tiotropium and Olodaterol in Chronic Obstructive Pulmonary Disease (COPD)" },{ "PostingID": 3877, "Title": "BI-1237.6", "Description": "Tiotropium +Olodaterol Fixed Dose Combination (FDC) Versus Tiotropium and Olodaterol in Chronic Obstructive Pulmonary Disease (COPD)" },{ "PostingID": 4109, "Title": "GSK-HZC113782", "Description": "A Clinical Outcomes Study to compare the effect of Fluticasone Furoate/Vilanterol Inhalation Powder 100/25mcg with placebo on Survival in Subjects with moderate Chronic Obstructive Pulmonary Disease (COPD) and a history of or at increased risk for cardiovascular disease" }]

Statistical Analysis Plan

Development of a lifelong FEV1 change modelWe will model decades-long FEV1 change using SReFT. SReFT is a novel method we previously developed to restore long-term time courses from numerous short fragments by extending the nonlinear mixed-effects model (NLMM).Concept of SReFTThe fitting of the “hyperparameters” (the parameters that define the distribution in the population, such as the means or variances of the parameters of the nonlinear function, or variances of the observation noise) in SReFT is achieved by maximizing the marginal likelihood function, as in NLMM. In addition to the hyperparameters, SReFT also estimates the "disease time" for the subject , i.e., the most likely time-point of an observed fragment for each subject in the disease progression. During the iterative optimization, subjects are shifted back and forth along the temporal axis to maximize the likelihood function. Assuming a lack of correlation across the disease progression of different subjects, this is performed for each subject individually. SReFT uses an identical objective function for both the estimation of disease time and estimation of the hyperparameters; thus, SReFT is formalized as a maximum likelihood estimator. We utilized mathematically simple functions to describe the evolution of all measured biomarkers during COPD progression. Functions should offer good flexibility for describing time-dependent changes in biomarkers, such as linear-like, exponential-like, and sigmoidal changes. SReFT is able to easily accommodate any nonlinear functions. Sub-group analysisWe will fit our models to the datasets of individual patients. This will allow us to gain an overview of the variability among different patients with COPD and thus help us to interpret the uncertainty of parameter estimates. Moreover, we will be able to analyze variability in the following sub-categories:·Change in biomarkers, such as the patient-reported outcome scores or vital signs·Characteristics of the patient, such as the age group, sex, and where they live·Presence or absence and types of co-morbidities·Presence or absence and types of concomitant medicationsValidation ProcessOnce the parameters are determined, we will perform a bootstrap analysis with numerous randomly resampled data sets to confirm the reproducibility of the parameters estimated by the final covariate model.

Publication Citation

Kawamatsu, S.; Jin, R.; Araki, S.; Yoshioka, H.; Sato, H.; Sato, Y.; Hisaka, A. Scores of Health-Related Quality of Life Questionnaire Worsen Consistently in Patients of COPD: Estimating Disease Progression over 30 Years by SReFT with Individual Data Collected in SUMMIT Trial. J. Clin. Med. 2020, 9, 2676

https://doi.org/10.3390/jcm9082676