Multiple-Breath Washout Outcome Measures in Adults with Bronchiectasis

Rationale: Lung clearance index (LCI) has good intravisit repeatability with better sensitivity in detecting lung disease on computed tomography scan compared with forced expiratory volume in 1 second (FEV₁) in adults with bronchiectasis. Alternative multiple-breath washout parameters have not been systematically studied in bronchiectasis. Objectives: To determine the validity, repeatability, sensitivity, specificity, and feasibility of standard LCI (LCI^2.5), shortened LCI (LCI^5.0), ventilation heterogeneity arising within proximal conducting airways (S_condVT), and ventilation heterogeneity arising within the acinar airways (S_acinVT) in a cross-sectional observational cohort of adults with bronchiectasis. Methods: Cross-sectional multiple-breath nitrogen washout data (Exhalyzer D; Eco Medics AG) from 132 patients with bronchiectasis across five United Kingdom centers (BronchUK Clinimetrics study) and 88 healthy control subjects were analyzed. Results: Within-test repeatability (mean coefficient of variation) was <5% for both LCI^2.5 and LCI^5.0 in patients with bronchiectasis, and there was no difference in mean coefficient of variation for LCI^2.5 and LCI^5.0 in patients with bronchiectasis compared with healthy volunteers. Moderate-strength correlations were seen between FEV₁ and LCI^2.5 (r = -0.54), LCI^5.0 (r = -0.53), S_condVT (r = -0.35), and S_acinVT (r = -0.38) z-scores. The proportion of subjects with abnormal multiple-breath washout (z-score > 2) but in normal FEV₁ (z-score < -2) was 42% (LCI^2.5) and 36% (LCI^5.0). Overall results from the receiver operating characteristic curve analysis indicated that LCI^2.5 had the greatest combined sensitivity and specificity to discriminate between bronchiectasis and control subjects, followed by LCI^5.0, FEV₁, and S_condVT z-scores. There was a 57% time saving with LCI^5.0. Conclusions: LCI^2.5 and LCI^5.0 had good within-test repeatability and superior sensitivity compared with spirometry measures in differentiating between health and bronchiectasis disease. LCI^5.0 is quicker and more feasible than LCI^2.5. Clinical trial registered with www.clinicaltrials.gov (NCT02468271).