Relapse in FEV1 Decline After Steroid Withdrawal in COPD

BACKGROUND: We previously observed that 30 months of inhaled corticosteroid (ICS) treatment can attenuate FEV1 decline in COPD, but it is unclear whether withdrawal induces a relapse. We hypothesized that FEV1 decline, airway hyperresponsiveness (AHR), and quality of life (QOL) deteriorate after ICS cessation even after prolonged use. METHODS: One hundred fourteen patients with moderate to severe COPD finished randomized 6-month or 30-month treatment with fluticasone (500 mu g bid), 30-month treatment with fluticasone and salmeterol (500/50 mu g bid), or placebo (first part of the Groningen and Leiden Universities Corticosteroids in Obstructive Lung Disease [GLUCOLD] study [GL1]). The subsequent 5 years, patients were prospectively followed annually, treated by their physician (GLUCOLD follow-up study [GL2]). Postbronchodilator FEV1, AHR, and QOL were initially recorded at baseline, at 30 months (GL1), and annually during GL2. Analysis was performed by linear mixed-effects models. RESULTS: Among 101 adherent patients during GL1, 79 patients started and 58 completed GL2. Patients using ICSs during GL1, but only using ICSs 0% to 50% of the time during GL2 (n = 56 of 79), had significantly accelerated annual FEV1 decline compared with GL1 (difference GL2-GL1 [95% CI]: 30-month treatment with fluticasone and salmeterol, -68 mL/y [-112 to -25], P = .002; 30-month treatment with fluticasone, -73 mL/y [-119 to -26], P = .002), accompanied by deterioration in AHR and QOL. CONCLUSIONS: ICS discontinuation aft er 30 months in COPD can worsen lung function decline, AHR, and QOL during 5-year follow-up. This suggests that ICS treatment lacks sustained disease-modifying effect after treatment cessation