Efficacy and safety of two doses of budesonide/formoterol fumarate metered dose inhaler in COPD

Inhaled corticosteroid/long-acting β2-agonist combination therapy is a recommended treatment option for patients with chronic obstructive pulmonary disease (COPD) and increased exacerbation risk, particularly those with elevated blood eosinophil levels. SOPHOS (NCT02727660) evaluated the efficacy and safety of two doses of budesonide/formoterol fumarate dihydrate metered dose inhaler (BFF MDI) versus formoterol fumarate dihydrate (FF) MDI, each delivered using co-suspension delivery technology, in patients with moderate-to-very severe COPD and a history of exacerbations. In this phase 3, randomised, double-blind, parallel-group, 12–52-week, variable length study, patients received twice-daily BFF MDI 320/10 µg or 160/10 µg, or FF MDI 10 µg. The primary endpoint was change from baseline in morning pre-dose trough forced expiratory volume in 1 s (FEV1) at week 12. Secondary and other endpoints included assessments of moderate/severe COPD exacerbations and safety. The primary analysis (modified intent-to-treat) population included 1843 patients (BFF MDI 320/10 µg, n=619; BFF MDI 160/10 µg, n=617; and FF MDI, n=607). BFF MDI 320/10 µg and 160/10 µg improved morning pre-dose trough FEV1 at week 12 versus FF MDI (least squares mean differences 34 mL [p=0.0081] and 32 mL [p=0.0134], respectively), increased time to first exacerbation (hazard ratios 0.827 [p=0.0441] and 0.803 [p=0.0198], respectively) and reduced exacerbation rate (rate ratios 0.67 [p=0.0001] and 0.71 [p=0.0010], respectively). Lung function and exacerbation benefits were driven by patients with blood eosinophil counts ≥150 cells·mm−3. The incidence of adverse events was similar, and pneumonia rates were low (≤2.4%) across treatments. SOPHOS demonstrated the efficacy and tolerability of BFF MDI 320/10 µg and 160/10 µg in patients with moderate-to-very severe COPD at increased risk of exacerbations.

The budesonide/formoterol fumarate dihydrate metered dose inhaler (BFF MDI) is an ICS/LABA fixed-dose combination that is formulated using innovative co-suspension delivery technology. Unlike drug crystal-only suspensions, co-suspension delivery technology enables reliable aerosol performance and consistent drug delivery, even in the presence of simulated patient-handling errors [12,13].
The recently completed phase 3 TELOS study (NCT02766608), conducted in patients with moderate-to-very severe COPD who were not required to have an exacerbation history, demonstrated that BFF MDI 320/10 µg and 160/10 µg improved lung function over 24 weeks compared with ICS monotherapy (which is not a recommended therapy for COPD) and reduced COPD exacerbations compared with LABA monotherapy. There was a numerical dose response favouring the BFF MDI 320/10 µg dose for all lung function and exacerbation endpoints [11].
The current SOPHOS study (NCT02727660) evaluated lung function ( primary objective), COPD exacerbations (secondary objectives) and safety, following treatment with BFF MDI 320/10 µg, BFF MDI 160/10 µg or formoterol fumarate dihydrate (FF) MDI monotherapy for at least 12, and up to 52, weeks in patients with moderate-to-very severe COPD who had a documented history of at least one moderate/ severe COPD exacerbation in the previous 12 months.

Material and methods
Patients Patients (40-80 years of age) were current/former smokers (⩾10 pack-years) with a history of COPD (defined by American Thoracic Society/European Respiratory Society or locally applicable guidelines) [14] who remained symptomatic (COPD Assessment Test score ⩾10) despite receiving one or more inhaled maintenance bronchodilators. Patients were required to have a post-bronchodilator forced expiratory volume in 1 s (FEV 1 )/forced vital capacity ratio <0.70, and post-bronchodilator FEV 1 ⩾25% to <80% predicted normal (second screening visit) according to the National Health and Nutrition Examination Survey III reference equations [15,16]. Patients had a documented history of at least one moderate/severe COPD exacerbation in the previous 12 months. Patients with a current diagnosis of asthma were excluded.
The primary endpoint (US registration approach) was the change from baseline in morning pre-dose trough FEV 1 at week 12, analysed using the efficacy estimand. Analysis of the same outcome using an attributable estimand was the first secondary endpoint.
Further secondary endpoints were: time to first moderate/severe COPD exacerbation; change from baseline in average daily rescue medication use over 12 weeks; and St George's Respiratory Questionnaire (SGRQ) responder rate ( percentage of patients achieving a minimal clinically important difference ⩾4 units in total score) at week 12. The rates of moderate/severe exacerbations were also assessed.
Pre-specified subgroup analyses of the primary endpoint and exacerbation rates were performed according to baseline blood eosinophil count (<150 versus ⩾150 cells·mm −3 ) and according to number of moderate/ severe COPD exacerbations in the previous 12 months (1 versus ⩾2). Additional post hoc analyses of exacerbation rates by ICS use at baseline and by 4-week intervals were also performed.

Safety analyses
Treatment-emergent adverse events (TEAEs) were monitored in the safety population (all randomised patients who received study drug, even if <1 full dose). Potential major adverse cardiovascular events (MACE) and pneumonia cases were adjudicated by an external Clinical Endpoint Committee.  figure E2). In the 12 months before screening, 38.5% of patients had two or more moderate/severe exacerbations.

Efficacy analyses
For the primary endpoint of change from baseline in morning pre-dose trough FEV 1 at week 12, analysed using the efficacy estimand, BFF MDI 320/10 µg and 160/10 µg demonstrated statistically significant  Pre-specified subgroup analyses suggested that the improvements in morning pre-dose trough FEV 1 at week 12 and reductions in the rate of moderate/severe COPD exacerbations with both doses of BFF MDI compared with FF MDI were driven by patients with a baseline blood eosinophil count ⩾150 cells·mm −3 ( fig. 4). Further analysis of lung function response by continuous baseline eosinophil count with locally weighted scatterplot smoothing (LOESS) curves showed that improvements in morning pre-dose trough FEV 1 for BFF MDI 320/10 µg versus FF MDI at week 12 began at blood eosinophil levels of approximately 100-150 cells·mm −3 . Results for BFF MDI 160/10 µg versus FF MDI were consistent with these findings up to eosinophil counts of approximately 350 cells·mm −3 , after which improvements to lung function with BFF MDI 160/10 µg began to decrease ( fig. 5a and b). LOESS curves of BFF MDI 320/10 µg and BFF MDI 160/10 µg showed that the two doses began to diverge from each other at blood eosinophil levels of approximately 350-400 cells·mm −3 , with the treatment difference between the two doses exceeding 20 mL from approximately 450 cells·mm −3 ( fig. 5c). Analysis of the rate of moderate/severe exacerbations by continuous baseline eosinophil count showed numerical improvements with both doses of BFF MDI versus FF MDI began at blood eosinophil levels of approximately 100-150 cells·mm −3 and were maintained with increasing eosinophil levels ( fig. 5d and e). With both BFF MDI doses, improvements in the rate of moderate or severe COPD exacerbations increased with increasing blood eosinophil levels.
When plotted together, the BFF MDI doses began to diverge from each other at blood eosinophil levels of approximately 150-200 cells·mm −3 , with a treatment difference between the two doses exceeding 15% at blood eosinophil levels above approximately 250-300 cells·mm −3 ( fig. 5f ).  Subgroup analyses based on the number of moderate/severe exacerbations in the year before the study showed that improvements in lung function and exacerbation rates were similar in patients with a history of one moderate/severe exacerbation and those with at least two moderate/severe exacerbations ( figure E3).
Post hoc analyses of the rate of moderate/severe exacerbations by ICS use prior to study entry demonstrated the benefit of ICS in all patients regardless of ICS usage prior to study entry. In an analysis that estimated the rate of exacerbations for each 4-week period, both BFF MDI doses separated from FF MDI during the first 4 months and the last 5 months of the treatment period, with some overlap during the mid-portion of the treatment period. Further details are provided in the online data supplement.
There were 22 deaths in the study (15 on-treatment, i.e. due to adverse events that started during the treatment period, and seven post-treatment). The highest number of on-treatment deaths occurred in the FF MDI group (n=8), followed by BFF MDI 160/10 µg (n=4) and BFF MDI 320/10 µg (n=3). No deaths were considered to be drug related. Details of the causes of death are provided in the online data supplement.

Discussion
The SOPHOS study aimed to evaluate the efficacy and safety of BFF MDI 320/10 µg and 160/10 µg in symptomatic patients with moderate-to-very severe COPD who had a prior history of COPD Baseline eosinophil count cells·mm -3   250  300  350  450  400  500  550  700  750  800  650  600  0  50  100  150  200  250  300  350  450  400  500  550  700  750  800  650  600  0  50  100  150  200  250  300  350  450  400  500  550  700  750  800  exacerbations. Both doses of BFF MDI resulted in statistically significant improvements in lung function ( primary efficacy endpoint) compared with FF MDI. Also, both doses of BFF MDI resulted in significant improvements in all secondary efficacy endpoints, including time to first moderate/severe exacerbation, and both doses reduced the rate of moderate/severe exacerbations, versus FF MDI. The small numerical dose response favouring BFF MDI 320/10 µg for improvements to lung function (trough FEV 1 ) and for reducing exacerbation rates in the overall study population appeared to be more pronounced as blood eosinophil levels increased, with the doses diverging above approximately 350-400 cells and 150-200 cells for trough FEV 1 and exacerbation rates, respectively. Results for the Ex-US analysis approach that are provided in the online data supplement were similar to the results for the US analysis approach that are presented in this manuscript.
Our findings were well aligned with the results of TELOS (NCT02766608), a 24-week study investigating the therapeutic effects of BFF MDI 320/10 µg and 160/10 µg relative to treatment with monocomponents in patients with moderate-to-very severe COPD [11]. In TELOS, both doses of BFF MDI also showed superiority to FF MDI monotherapy, with a dose-related response for lung function and exacerbation endpoints. However, while >70% of patients in TELOS did not have a moderate/severe COPD exacerbation in the year before the study, SOPHOS specifically enrolled patients with a documented history of moderate/ severe exacerbations in the year prior to the study. As expected given the differences in the patient population, the exacerbation rates in SOPHOS were considerably higher than those in TELOS, but the treatment effect of BFF MDI compared with FF MDI, as assessed by the rate ratio for COPD exacerbations, was very similar in both studies. For patients who continued treatment beyond 24 weeks and up to 52 weeks, the effects of BFF MDI on exacerbation risk seen in the first 24 weeks of treatment were sustained over the entire treatment period. Hence, SOPHOS confirmed the efficacy of BFF MDI in a population known to be at risk of exacerbations (i.e. in a population for which ICS/LABA is a recommended treatment option to prevent further exacerbations [1]) over a treatment period of 12 to 52 weeks.
A pre-defined subgroup analysis according to baseline blood eosinophil count showed that improvements in lung function and reductions in exacerbation rates were driven by patients with baseline blood eosinophil counts ⩾150 cells·mm −3 . These findings were generally comparable to the TELOS study [11], which included a similar eosinophil subgroup analysis. The relationship between blood eosinophil levels and response to ICS-containing therapies, with regards to reducing exacerbation risk, has also been described in other recent studies [17][18][19][20][21], suggesting that patients with higher eosinophil counts may In addition, our analysis by continuous baseline eosinophil count showed that a dose response favouring BFF MDI 320/10 µg was evident from blood eosinophil levels of approximately 450 cells·mm −3 for lung function endpoints and approximately 250 cells·mm −3 for exacerbation rates. These data suggest that there is the potential for patients with COPD to derive benefit from more than one ICS dose and that overall ICS dose response is influenced appreciably by eosinophil levels. These hypotheses will be prospectively evaluated in the recently completed ETHOS study [22]. Finally, subgroup analyses based on the number of moderate/severe COPD exacerbations (1 or ⩾2) in the prior year showed that the benefits on lung function and exacerbation rates for BFF MDI versus FF MDI were not affected by the number of exacerbations in the year before the study.
To minimise the potential for a run-in bias [23] patients in our study who were using ICS prior to the study entry continued the ICS during the screening period until randomisation. Additional analysis of exacerbation rates for each 4-week period of treatment indicated that the benefit of BFF MDI on exacerbations was not driven solely by acute ICS withdrawal in the first month after randomisation, as suggested for recent trials comparing triple therapy with LAMA/LABA dual therapies [24].
BFF MDI 320/10 μg and 160/10 μg were well tolerated in this 12-52 week study, with no new or unexpected safety findings or dose-related effects on the pattern or frequency of TEAEs, which confirmed the findings from the 24-week TELOS study [11]. ICS use in patients with COPD has been associated with an increased risk of pneumonia [1, 25,26]. However, there is evidence to suggest that the pneumonia risk with treatments containing budesonide may be lower compared with other ICS components [27][28][29], and a recent pooled analysis found no significant increase in the risk of pneumonia when comparing budesonide-containing with non-budesonide-containing treatments [30]. In the present study, the percentage of patients with confirmed pneumonia events was 2.1% overall, and lowest in the BFF MDI 320/10 µg group (1.6%); this was similar to the results of TELOS and also the KRONOS study, neither of which showed a higher incidence of pneumonia in their budesonide-containing treatment arms relative to non-budesonide-containing treatment arms [11,31]. As in the TELOS study, no ICS dose-related effect on the incidence of pneumonia was observed in this present study [11].
A potential limitation of SOPHOS was the variable length study design and the correspondingly small proportion of patients (10.2%) who completed 52 weeks of treatment. However, the effects of BFF MDI on exacerbation risk observed in the first 24 weeks of treatment were sustained after 24 weeks, and were comparable to those observed with a different formulation of BFF versus FF over 6 months [4]. Since only a small portion of patients completed 12 months of treatment, the study may not be optimal for assessing risk of exacerbations over the long term. However, patients in this study were recruited over an 18-month period and all seasons were well represented, reducing the potential impact of seasonal variation on exacerbation rates.
In conclusion, the SOPHOS study extended the findings from the TELOS study by demonstrating the efficacy and safety of BFF MDI, formulated using co-suspension delivery technology, in the key target population for ICS/LABA treatment; i.e. patients with a history of COPD exacerbations, over a treatment period of 12 to 52 weeks. BFF MDI 320/10 µg and 160/10 µg both improved lung function and reduced exacerbation risk compared to FF MDI monotherapy, with a dose response favouring the BFF MDI 320/ 9.6 µg dose at elevated blood eosinophil levels. Both doses of BFF MDI were well tolerated, and there was no ICS dose-related effect on safety outcomes, including the incidence of confirmed pneumonia events.
The study findings confirm that BFF MDI was an effective treatment for the management of COPD in patients who had a prior exacerbation history.