Abstract Rationale People with COPD are at risk of adverse events such as exacerbations and major adverse cardiovascular events(MACE), both of which increase mortality. Whether different COPD inhaler treatment regimens reduce cause-specific mortality after adverse events is unknown. Amongst a nationally representative COPD population, comprising people with an exacerbation or MACE history, we compared cause-specific mortality by treatment regimen, i.e., inhaled corticosteroid(ICS)-containing treatments versus long-acting bronchodilators without ICS. Methods We used routinely collected electronic healthcare records (Clinical Practice Research Datalink Aurum primary care, Hospital Episode Statistics secondary care, and Office of National Statistics death certification). Inclusion criteria were a COPD diagnosis, age≥40 years, ≥one year continual registration at primary care practice, smoking history (current or ex), eligibility for data linkage, and data on record between January 1st, 2010 and March 31st, 2023. Start of follow-up for people without exacerbation or MACE events was the latest date of meeting basic inclusion criteria. Start of follow-up for the exacerbation group was the date of first exacerbation after meeting criteria, and for the MACE group was the date of first lifetime MACE, after meeting follow-up criteria. The exposure was inhaler treatment determined in the year before start of follow-up. ICS-containing treatment was defined as ≥one ICS prescription (mono or combination), and the control group comprised individuals on long-acting bronchodilators (mono or combination) in the absence of ICS. The outcome was one-year all-cause mortality and one-year cause-specific mortality (respiratory, cardiovascular, cancer, or other), with consideration of competing risk of the other causes. We implemented Cox Proportional Hazard models, and Fine and Gray models to obtain sub-distribution hazard ratios, adjusted for age, sex, socioeconomic status, smoking status, body mass index category, current asthma, exacerbation frequency and severity in the previous year, COPD severity, dyspnoea score, diabetes, hypertension, depression, anxiety, gastro-oesophageal reflux disease(GORD/GERD), lung cancer, short-acting bronchodilators, and cardiovascular medications. Results We derived four adverse outcome-related COPD patient phenotypes (not mutually exclusive), namely cardiovascular-naïve exacerbator(Nphenotype%ICS=87,81871.2%), cardiovascular-naïve non-exacerbator(Nphenotype%ICS=32,32462.3%), MACE(Nphenotype%ICS=21,95373.6%) and non-MACE(Nphenotype%ICS=90,66270.2%). There was no association between treatment regimen and all-cause mortality in exacerbation, MACE, or non-MACE phenotypes, but ICS was positively associated with all-cause mortality only in cardiovascular-naïve non-exacerbators(aHR95%CI=1.391.15-1.69), driven by respiratory-specific mortality(aSHR95%CI=2.361.44-3.86)(Tab.1). Treatment regimen was not associated with cardiovascular-, cancer-, or other-specific cause of death(Tab.1). Conclusions ICS-containing treatment was not associated with subsequent mortality across these phenotypes in COPD, except for a positive association between ICS and respiratory-specific mortality amongst cardiovascular-naïve non-exacerbators. This abstract is funded by: National Institute for Health and Care Research (NIHR)
Ioannides et al. (Fri,) studied this question.