Which LABA/LAMA should be chosen in COPD patients in real life?

Abstract

BACKGROUND\nGiven COPD heterogeneity, we do not know if some LABA/LAMAs are more suitable for some COPD phenotypes. This real-life database study aimed to evaluate retrospectively the 4 LABA/LAMA effectiveness and highlight possible specificities that could better guide us in choosing the right LABA/LAMA to be used.\n\n\nMETHODS\nWe searched for subjects (1,779) adherent to umeclidinium/vilanterol (UM/VI), indacaterol/glycopyrronium (IND/GLY), aclidinium/formoterol (ACLI/FOR) and tiotropium/olodaterol (TIO/OLO) treatments in our prescribing/dispensing database. Prescriptions for systemic corticosteroids (SC), antibiotics and salbutamol during one year of LABA/LAMA treatment were analyzed.\n\n\nRESULTS\nA better adherence was found in individuals taking IND/GLY (10.42±1.86 packages/year) compared with UM/VI (10.09±1.9;p=0.008), ACLI/FOR (9.8±1.8;p=0.001) and TIO/OLO (10.1±2.1;p=0.047). The number of patients that were prescribed at least one package of SC/year and their package numbers/year were similar in males/females, across age groups and in non-frequent exacerbators with the 4 LABA/LAMAs. More SC were taken by frequent exacerbators, whereas fewer SC/antibiotic packages were prescribed to subjects aged >80 years with all treatments. Patients treated with ACLI/FOR or TIO/OLO showed lower risks to having antibiotic prescriptions compared with those taking UM/VI (0.698[0.516-0.945] and 0.696[0.491-0.985;p=0.020 and p=0.041) and IND/GLY (0.597[0.445-0.802] and 0.595[0.423-0.836];p=0.001 and p=0.003). Lower risks for salbutamol prescriptions were observed with UM/VI (0.678[0.480-0.958];p=0.027) and TIO/OLO (0.585[0.365-0.937];p=0.026) compared to ACLI/FOR.\n\n\nCONCLUSION\nAccording to our retrospective database study, each LABA/LAMA could have a specific efficacy profile in COPD that might be considered for personalized therapy. However, head-to-head targeted trials aimed to assess the impact of different LABA/LAMAs on COPD are needed to confirm/disprove such results.