BGF initiation after an exacerbation in COPD: Does timing matter?

Findings from the MITOS EROS+CP (US) study underscore the importance of prompt initiation of budesonide/glycopyrrolate/formoterol fumarate (BGF) following an exacerbation in patients with chronic obstructive pulmonary disease (COPD).

Patients with COPD are at an increased risk of acute exacerbation of respiratory symptoms and acute cardiovascular (CV) events, particularly following acute COPD exacerbations (AECOPD). [J Thorac Dis 2020;12:2791-2802; Circulation 2014;129:971-980; Int J Chron Obstruct Pulmon Dis 2024;19:225-241; Am J Respir Crit Care Med 2024;209:A1865]

“Timely initiation of triple therapy is associated with a lower risk of AECOPD,” said Michael Pollack, Respiratory and Immunology Franchise Lead at AstraZeneca, Wilmington, Delaware, US, at ATS 2025. “[Hence, we] investigated whether prompt initiation of single-inhaler triple therapy with BGF after a COPD exacerbation is associated with reduced subsequent cardiopulmonary events (severe AECOPD and CV events) compared with delayed or very-delayed initiation strategies.”

This retrospective, observational database analysis comprised 25,603 COPD patients (mean age 60.3 years, 64.3 percent women) who had a 12-month baseline history and initiated BGF therapy following ≥1 exacerbation. About 96 percent of participants had moderate index exacerbation. The study also evaluated a subgroup of dual escalators, ie, patients who escalated from prior dual therapy* (n=15,023). [ATS 2025, session D14]

Nearly half (47.5 percent) of the participants were very-delayed BGF initiators (181–365 days), 37.7 percent were delayed (31–180 days), and only 14.8 percent were prompt (≤30 days). According to Pollack, this reflects the lack of timely BGF initiation amongst COPD patients.

Dual escalators had more very-delayed initiators (51.9 percent). The median time to BGF initiation following an AECOPD was 169 days. The mean follow-up was 462.2 days.

The proportions of participants with ≥1 CV-related comorbidity at baseline did not differ across the prompt, delayed, and very-delayed BGF initiation cohorts (79, 81.6, and 83.8 percent, respectively).

Among prompt initiators, crude event rates of any AECOPD following an index exacerbation and cardiopulmonary events were 1.29 and 0.23 per patient-year (PY), respectively. These were lower than those reported for delayed (1.91 and 0.31 per PY) and very-delayed initiators (2.45 and 0.32 per PY).

Prompt initiators also had lower incidences of subsequent exacerbation events than delayed initiators in both overall (adjusted incident rate ratio [aIRR], 0.74) and dual-escalation (aIRR, 0.80) cohorts. A consistent pattern was observed in the comparison between prompt and very-delayed initiators (aIRRs, 0.69 and 0.75, respectively).

This effect favouring prompt over delayed and very-delayed initiation was similarly seen for cardiopulmonary events. The aIRRs for the prompt vs delayed comparison were 0.84 (overall) and 0.81 (dual escalation); the corresponding aIRRs in the prompt vs very-delayed comparison were 0.83 and 0.81.

In summary, prompt BGF use reduced exacerbation and cardiopulmonary events by 26–31 percent and about 17 percent, respectively, in the overall cohort. These effects were mirrored in the dual-escalation cohort (20–25 percent and 19 percent, respectively).

Pollack shared that studies looking at the post-discharge setting are in the works, as this is a very vulnerable time for patients.

Why the delay in initiation?

He also suggested conducting qualitative analyses to understand why physicians are not initiating. “[We could take] patient profiles, discuss with advisors and physicians, and look through some cases to determine why they did not initiate earlier. It could be patient preference or product access – there could be a whole host of things.”

Pollack expounded on this when asked why the study had a high percentage of delayed initiators. “Looking at the broad comorbidity profile in our study, very-delayed initiators tend to be sicker. Hence, there may have been a prioritization on the patient’s or physician’s part to address [other comorbidities first]. This is the third study, yet we are still left scratching our heads about why they are getting delayed therapy.”

This may have also been driven by a lack of urgency because most of the very-delayed initiators had background dual therapy, he continued. “Perhaps patients or physicians felt that [the symptoms] were being managed appropriately with the background therapy.”

“As for the prompt initiators, they have more severe index exacerbations. This probably gives them a sense of urgency to initiate BGF promptly,” Pollack noted.

“Nonetheless, I think [the current results are] consistent with what we have seen before, introducing some new results relating to cardiopulmonary events,” he said.