It has been suggested that BEC may change in response to ICS treatment and higher BEC measured while patients are receiving ICS (BEC on ICS) may not be associated with treatment response to ICS.
We explored the association of BEC on ICS and BEC off ICS with treatment response to ICS in data from the FLAME trial, a 52-week double-blind trial comparing a combination of LABA+LAMA with LABA+ICS.
We assessed the three-way interaction between outcomes of interest, treatment and the BEC variables. We used generalised linear models (negative binomial distribution) for assessing exacerbation frequency and Cox proportional hazards model for time-to-event outcomes. We adjusted for main demographics and historical treatments.
Our analysis was based on 1,380 (41%) participants that were receiving ICS prior to recruitment and had BEC measured on and off ICS. In line with the main analysis, LABA+LAMA was superior to LABA+ICS in most participants. However, higher BEC off ICS was associated with improved treatment response to LABA+ICS with regards to the frequency of and time-to-first exacerbation (moderate or severe, any exacerbation, exacerbations treated with systemic steroids only, p<0.05) and with time-to-first pneumonia. On the contrary, BEC on ICS was only significantly associated with treatment response to exacerbations treated with systemic steroids only. Importantly, among patients that received ICS within 2 days prior to BEC on ICS measurement, higher BEC on ICS was associated with less favourable response to LABA+ICS.
Our results suggest use of BEC (on ICS) to guide ICS discontinuation should be further evaluated.