PT - JOURNAL ARTICLE AU - Helen Stone AU - Gillian McNab AU - Robert Stockley AU - Elizabeth Sapey TI - Variability of sputum inflammatory mediators in alpha-1-antitrypsin deficiency and usual COPD DP - 2011 Sep 01 TA - European Respiratory Journal PG - p4033 VI - 38 IP - Suppl 55 4099 - http://erj.ersjournals.com/content/38/Suppl_55/p4033.short 4100 - http://erj.ersjournals.com/content/38/Suppl_55/p4033.full SO - Eur Respir J2011 Sep 01; 38 AB - Introduction: There is inherent variability in the concentrations of inflammatory mediators in stable state sputum of usual COPD patients. Patients with alpha-1-antitrypsin deficiency (ATD) have a similar spectrum of lung disease and more inflammation but variability has not been assessed. Evidence for the efficacy of treatment in A1AT is lacking; proof of concept (POC) studies indicate that augmentation increases AAT levels and reduces local mediator release. Our aims were to assess the daily variability of mediators in sputum in ATD, compared to usual COPD and to study the effects of sequential sampling to determine sample size for POC studies.Methods: Interleukin 8 (IL8), myeloperoxidase (MPO) and leukotriene B4 (LTB4) were measured in the stable state spontaneous sputum of 12 patients with ATD and 12 usual COPD patients on 9 days over 1 month. The intra-patient variability was calculated (CV), compared between the groups, and the effects of combining results from multiple days for each patient were assessed.Results: There was significant daily variability in all mediators, which was greater in usual COPD, despite IL8 and LTB4 concentrations being higher (p<0.01) in ATD (medians 11.29 vs 3.72 nM; and 12.16 vs 6.10 nM respectively). 3 or 5-day rolling means reduced the median CV in both groups compared to a single days' data (p<0.01) and reduced the number needed to show a 50% reduction in mediator as part of a POC study.Conclusion: There is greater variability in usual COPD than ATD; though mediator levels were higher in ATD. Sequential sampling reduced intra-patient variability in both groups. Averaging 3 consecutive samples per patient was optimal.