Correlations between FEV1 and patient-reported outcomes: A pooled analysis of 23 clinical trials in patients with chronic obstructive pulmonary disease

Abstract

Background: In clinical trials of inhaled bronchodilators, chronic obstructive pulmonary disease (COPD) guidelines recommend that patient-reported outcomes (PROs) are assessed alongside lung function. How these endpoints are related is unclear.

Methods: Pooled longitudinal data from 23 randomised controlled COPD studies were analyzed (N = 23,213). Treatments included long-acting β₂ agonists, long-acting muscarinic antagonists (LABAs or LAMAs) and the LABA/LAMA combination QVA149. Outcome measures were Transition Dyspnoea Index (TDI) and St. George's Respiratory Questionnaire (SGRQ) scores, COPD exacerbation frequency and rescue medication use. Relationships between changes in trough forced expiratory volume in one second (ΔFEV₁) and outcomes following treatment were assessed using correlations of data summaries and model-based analysis: generalized linear mixed-effect regression modelling to determine if ΔFEV₁ could predict patient outcomes with different treatments.

Results: Mean age was 64 years, 73% were male, and most had moderate (45%) or severe (52%) disease. Statistically significant correlations were observed between ΔFEV₁ and each outcome measure (exacerbations Rs = 0.05; rescue medication, SGRQ, TDI, r = 0.11-0.16; all p < .001). Patients with greater improvements in trough FEV₁ had on average better SGRQ and TDI scores, fewer exacerbations, and used less rescue medication. For SGRQ and TDI scores, minimal clinically important differences were observed over the range of pooled ΔFEV₁ values. Model-based predictions confirmed the treatment effect was partly explained by changes in FEV₁ from baseline with improvements in PROs observed across all treatments when trough FEV₁ improved. Across all endpoints active treatments were better than placebo (p < .0001), and LABA/LAMA treatment resulted in numerically better treatment outcomes than either monocomponent.

Conclusions: These data suggest that FEV₁ improvements post-bronchodilation correlate with PRO improvements. Further improvements in patient outcomes may be expected by maximizing lung function improvements.

Trial registration: Registration details for the 23 randomised controlled studies used in this pooled analysis are supplied in Additional File 4.

Keywords: COPD; FEV1; PROs; Regression modelling; SGRQ; TDI.