Efficacy and safety of bronchial thermoplasty in clinical practice: A prospective, longitudinal, cohort study using evidence from the UK Severe Asthma Registry

Objectives Use data from the UK Severe Asthma Registry (UKSAR) to assess the efficacy and safety of bronchial thermoplasty (BT) in routine UK clinical practice and to identify characteristics of ' responders'. Design Prospective, longitudinal, cohort, multicentre registry study. Setting All (11) UK centres performing BT. Participants and intervention Patients receiving BT in the UK between 01/06/2011 and 30/09/2016 who had consented to data entry into UKSAR (n=133). Efficacy data were available for 86 patients with a BT baseline and at least one follow-up record. Safety data were available for 131 patients with at least one BT procedure record. Primary and secondary outcome measures Efficacy: AQLQ, ACQ, EuroQol, HADS anxiety and HADS depression scores, FEV 1 (% predicted), rescue steroid courses, unscheduled healthcare visits (A&E/Asthma clinic/GP), hospital admissions and days lost from work/school. Safety: peri-procedural events, device problems and any other safety-related findings. Responder analysis: differences in baseline characteristics of ' responders' (≥0.5 increase in AQLQ at 12 months) and ' non-responders'. Results Following Bonferroni correction for paired comparisons, mean improvement in AQLQ at 12 months follow-up compared with BT baseline was statistically and clinically significant (0.75, n=28, p=0.0003). Median reduction in hospital admissions/year after 24 months follow-up was also significant (-1.0, n=26, p<0.0001). No deterioration in FEV 1 was observed. From 28 patients with AQLQ data at BTBL and 12-month follow-up, there was some evidence that lower age may predict AQLQ improvement. 18.9% (70/370) of procedures and 44.5% (57/128) of patients were affected by an adverse event; only a minority were considered serious. Conclusions Improvement in AQLQ is consistent with similar findings from clinical trials. Other efficacy outcomes demonstrated improving trends without reaching statistical significance. Missing follow-up data impacted this study but multiple imputation confirmed observed AQLQ improvement. The safety review suggested BT is being performed safely in the UK.