Rate of Decline of FEV1 as a Biomarker of Survival?

For decades, the natural history of chronic obstructive pulmonary disease (COPD) has been defined by the excess decline in lung function induced by tobacco smoking, with FEV1 considered the gold standard biomarker of COPD development and progression (1). Several studies have been conducted to assess the effect of smoking cessation (2) and/or pharmacological treatment on FEV1 decline with the primary aim of identifying treatments that could reduce lung function decline and thus disease progression, potentially improving survival. Unfortunately, no study with FEV1 decline as primary outcome has proven an effect of pharmacological treatment on lung function decline (3). Even in the Lung Health Study, which showed a positive effect of smoking cessation, there was no effect from ipratropium (2).

We now understand the several limitations of those studies. First, patients have been studied for a relatively short period of time (maximum 4 yr). Second, patients were typically studied late in life, when the decline slows down. Third, and more importantly, we now know that only 50% of patients who present with COPD after the age of 50 years attained high maximal lung function in their twenties and then underwent fast decline, the rest having attained low maximal lung function because of early events and who therefore developed COPD without excess decline (3, 4). The fact that these two routes are indistinguishable later in life, at the age of 50–60 years when the diagnosis of COPD is usually established, prevents the identification of the fast decliners.

Subsequent studies were performed with other primary outcomes, such as survival in TORCH (TOwards a Revolution in COPD Health) (5) and SUMMIT (Study to Understand Mortality and Morbidity in COPD) (6), with lung function decline assessed as a secondary outcome. More recently, exacerbations have been increasingly used as primary outcome, particularly in phase 3 studies such as IMPACT (Informing the Pathway of COPD Treatment) (7) and ETHOS (Efficacy and Safety of Triple Therapy in Obstructive Lung Disease) (8), which were the first studies to show an effect of pharmacological treatment on survival, although not as primary outcome.

Because post hoc analyses of some pharmacological studies had shown a positive effect on FEV1 decline, either as a secondary outcome or in subgroups of patients with COPD (6, 9, 10), in this issue of the Journal, Celli and colleagues (pp. 689–698) decided to undertake a careful systematic review of placebo-controlled pharmacological trials lasting longer than 1 year to answer the question of whether FEV1 decline can indeed be ameliorated by therapy (11). They observed an average difference in FEV1 decline of 5 ml per year between active medications and placebo and suggested that it corresponds to the benefit reported for clinically relevant outcomes (such as health status and exacerbation rates) that are considered to be improved by the same agents in the same studies. They concluded that pharmacotherapy is effective in altering the rate of lung function decline and that because the yearly absolute difference observed was similar to the treatment difference reported for clinical outcomes, current guidelines should be adjusted to reflect these findings, and future studies should include the effects on lung function decline, particularly in patients with rapid lung function decline. We agree with the authors’ conclusion on the importance of the overall positive effect of long-term pharmacologic treatment on lung function decline and support their suggestion that Global Initiative for Chronic Obstructive Lung Disease should reconsider this outcome in the assessment of response of COPD to treatment.

However, when considering the strength of this proposal, as nicely discussed by the authors, the difference versus placebo of 5–9 ml/yr is most likely too small to be used as an outcome not only in studies comparing active treatment against an active comparator but also in individual patients. By contrast, this difference may become identifiable (especially in fast decliners) if spirometry is performed in smokers from early in life (e.g., at 20–30 years of age) and at regular intervals afterward (4, 10, 12). Interestingly, as very elegantly recently reported by Marott and colleagues (12), fast decliners 1) may indeed be identified in long-term follow-up (60 ml/yr vs. 30 ml/yr) and 2) are at increased risk of death, particularly respiratory death possibly owing to continuing smoking. The simplified model of lung function decline (normal and COPD slow or fast decliners) in that study is illustrated in Figure 1. This model could provide the framework for future studies aimed to prevent excessive lung function decline and its respiratory and systemic consequences in fast decliners.

Figure 1.

Outline of the study. Participants were assigned into one of three FEV1 trajectories of interest: no chronic obstructive pulmonary disease (COPD), COPD developed through low maximally attained FEV1 trajectory, and COPD developed through normal maximally attained FEV1 trajectory based on information from the 1976–1978 or 1981–1983 examination. After the baseline 2001–2003 examination, individuals were followed for 17 years with regard to risk of severe exacerbations of COPD, repiratory disease mortality, and all-cause mortality. Reprinted by permission from Reference 12.

Regular measurement of lung function with simple spirometry including FVC would also allow an assessment of the relative value of FEV1 and FVC as biomarkers of COPD progression and/or survival.

The mechanisms of airflow limitation and excess decline are rather complex and may involve not only smooth muscle contraction, hypersecretion, and airway wall remodeling but also destruction of small airways and emphysema and systemic extrapulmonary effects (13). Thus, exploring the potential effects of pharmacological treatment on airflow limitation and decline in lung function should not be limited to long-acting bronchodilators and inhaled corticosteroids but possibly to other antiinflammatory (13, 14) and antifibrotic agents (15). In this context, lung diseases previously believed to be irreversible, such as idiopathic pulmonary fibrosis, have been shown to be sensitive to antifibrotic agents such as nintedanib and pirfenidone, both of which are effective in preventing lung function decline in patients with idiopathic pulmonary fibrosis (15). These agents might be effective in reducing lung function decline in patients with COPD dominated by airway wall remodeling. Finally, because lung function decline might reflect systemic effects of smoking, agents used for the treatment of frequent cardiovascular comorbidities of COPD might also be tested (3), or biologics that markedly improve lung function in severe asthma with airflow limitation (16), with the aim to verify whether comprehensive treatment of patients with COPD and one or more comorbidity might reduce lung function decline taken as a biomarker of survival.