Characteristics of Patients with COPD Initiating Budesonide/Glycopyrronium/Formoterol or Other Triple Therapies in Japan: A Real-World Healthcare Claims Database Study (MITOS-AURA)

Abstract

Introduction

In Japan, patients with chronic obstructive pulmonary disease (COPD) can be escalated to treatment with inhaled triple therapy. Two single-inhaler triple therapies combining an inhaled corticosteroid/long-acting muscarinic antagonist/long-acting β₂-agonist (ICS/LAMA/LABA) are approved maintenance therapies for patients with COPD, and multiple-inhaler triple therapies (MITTs) are also available. There is limited evidence regarding real-life treatment patterns and characteristics of patients with COPD initiating triple therapies.

Methods

This observational, retrospective cohort study identified patients with COPD in Japan from an administrative claims database (May 2018–December 2021). Demographics, clinical characteristics, and healthcare resource utilization (HCRU) were assessed in four cohorts initiating a triple therapy: budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF) early adopters (initiated ≤ 12 months after market approval [September 1, 2019]), contemporary BGF users (initiated > 12 months after market approval), fluticasone furoate/umeclidinium/vilanterol (FF/UMEC/VI) users, and any MITT users.

Results

A total of 636 patients were BGF early adopters, 2558 were contemporary BGF users, 11,187 used FF/UMEC/VI, and 5931 used MITT. The percentage of patients with concomitant asthma in each cohort was 73.0%, BGF early adopter; 74.2%, contemporary BGF; 75.7%, FF/UMEC/VI; and 84.5%, MITT. During the 12-month baseline period, the frequency of patients with ≥ 1 moderate/severe exacerbation was 18.2%, BGF early adopter; 14.3%, contemporary BGF; 13.1%, FF/UMEC/VI; and 14.0%, MITT. ICS/LABA treatment during baseline was the most frequent pathway to triple therapy, ranging from 38.2% to 51.7% across cohorts. HCRU was relatively high across cohorts (range of hospital outpatient visits/patient during the 12-month baseline period, 11.0–14.1). Multimorbidity was observed in > 80% of patients in all cohorts; cardiovascular diseases were among the most common.

Conclusion

Many patients initiating triple therapy for COPD had concomitant asthma and had previously received ICS/LABA maintenance therapy. Patients prescribed BGF in the initial post-launch period were more likely to have a previous exacerbation history versus other cohorts, indicating more severe disease.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12325-024-02994-8.

Plain Language Summary

In Japan, single inhalers containing three medicines, known as single-inhaler triple therapies, are approved for people with chronic obstructive pulmonary disease (COPD). These inhalers include one medicine that reduces inflammation (a corticosteroid) and two medicines that expand the airways in the lung (bronchodilators). There are two approved single-inhaler triple therapies for long-term COPD treatment in Japan, consisting of active substances referred to as budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF) or fluticasone furoate/umeclidinium/vilanterol (FF/UMEC/VI). Although guidance is available to doctors on when to prescribe single-inhaler triple therapy, two important questions related to use in routine clinical practice remain: when are these medicines prescribed, and what types of patients receive them. To answer these questions, we assessed anonymized medical claims of people with COPD obtained from a large database in Japan. People with COPD who had used triple therapy were split into four groups: early BGF users (started treatment within 1-year of approval in Japan), contemporary BGF users (started treatment over 1-year after approval in Japan), FF/UMEC/VI users, and users of any multiple-inhaler triple therapy. Our findings showed that the majority of patients starting any triple therapy had diagnoses of both asthma and COPD, and multiple other medical conditions, including heart-related conditions. Healthcare use, including outpatient hospital clinic visits, was high across all groups before starting triple therapy. As early BGF users were more likely to have previous flare-ups (acute worsening of COPD symptoms) than the other groups, upon approval in Japan, doctors may have initially prescribed BGF to people with more severe COPD.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s12325-024-02994-8.

Key Summary Points

Why carry out this study?

There is limited evidence regarding the characteristics and real-life treatment patterns of patients with chronic obstructive pulmonary disease (COPD) who initiate inhaled corticosteroid/long-acting muscarinic antagonist/long-acting β2-agonist (ICS/LAMA/LABA) triple therapies in Japan, where there is high prevalence of COPD among people aged ≥ 40 years.

This study aimed to use real-world data to describe the demographics and clinical characteristics of patients with COPD in Japan, from the Medical Data Vision claims database, initiating an inhaled triple therapy, including budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF), fluticasone furoate/umeclidinium/vilanterol, or multiple-inhaler triple therapy.

Many patients initiating triple therapy for COPD in Japan had concurrent asthma and COPD at baseline, and had previously received ICS/LABA dual therapy as maintenance treatment.

What was learned from the study?

Multimorbidity was observed in > 80% of patients in all cohorts; cardiovascular diseases were among the most common medical conditions, occurring with a similar frequency across cohorts.

Patients prescribed BGF in Japan during the initial post-launch period were more likely to have a history of previous exacerbations than the other cohorts, potentially indicative of more severe disease at baseline; all patients initiating triple therapies had relatively high prior healthcare resource utilization.

Introduction

In Japan, a high prevalence of chronic obstructive pulmonary disease (COPD) among people aged ≥ 40 years had been estimated (8.6%), but COPD is considered to be underdiagnosed in routine clinical practice [1–3]. Inhaled corticosteroid/long-acting muscarinic antagonist/long-acting β₂-agonist (ICS/LAMA/LABA) triple therapies are used as maintenance treatment for COPD [4], and two single-inhaler ICS/LAMA/LABA triple therapies were approved for use in Japan as maintenance treatments for COPD in 2019: budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF) and fluticasone furoate/umeclidinium/vilanterol (FF/UMEC/VI) [5].

In the Japanese Respiratory Society (JRS) treatment guidelines, triple therapy is recommended for two groups of patients with COPD, the first being patients with no diagnosis of asthma who experience exacerbations despite LAMA/LABA treatment and have blood eosinophil (EOS) count ≥ 300 cells/mm³, and the second being patients with a concurrent asthma diagnosis who have continuing or worsening symptoms despite ICS/LABA (or ICS + LAMA) dual therapy (via a single or separate inhalers) [6]. However, there is a lack of evidence regarding the characteristics and real-life treatment patterns prior to initiation of triple therapies in Japan.

Many factors may influence treatment patterns in COPD, and it is possible that newly available triple therapies are preferentially initiated by patients with more severe or uncontrolled COPD. These patients could have a poorer prognosis and therefore could need new treatment options. Moreover, different restrictions after a product launch may also impact treatment patterns for triple therapies. For example, a standard regulation in Japan, known as Ryotanki, restricted prescriptions to a 2-week supply in the first year of launch after market approval of BGF in Japan.

This study aimed to use real-world data to describe the demographics, clinical characteristics, and healthcare resource utilization (HCRU) of patients with COPD in Japan initiating an inhaled triple therapy, including BGF, FF/UMEC/VI, or multiple-inhaler triple therapy (MITT), to help to understand when these medicines are prescribed and the types of patients that receive them.

Methods

Study Design and Data Source

This was an observational, retrospective cohort study of patients with COPD using data from the Medical Data Vision Co., Ltd. (MDV; Tokyo, Japan) claims database between May 2018 and December 2021.

The MDV is a large-scale administrative claims database that collects reimbursement data from over 480 hospitals covering approximately 45 million insured patients, which includes a representative population with a significant share of patients aged > 65 years [7, 8]. The database has previously been described and used in several studies of patients with respiratory disease, and includes details on hospital care and outpatient care provided by the participating hospital systems [9–13]. The MDV supports Diagnosis Procedure Combination (DPC) hospitals to collect and report administrative data and provide anonymized data that are processed for use by over 100 organizations [8].

The study design is presented in Fig. 1. The selection period varied by cohort and was defined as the time period during which all patients satisfying the selection criteria were enrolled in the study. The index date was defined as the date of COPD triple therapy (MITT or single-inhaler triple therapy [SITT]) initiation during the selection period and satisfying all the selection criteria. If a patient used more than one triple therapy (i.e., if they switched triple therapy regimens), they would have more than one index date and therefore a patient could appear in more than one cohort. The pre-index period included all the person-time before (at least 12 months), excluding the index date. When a patient had multiple index dates, one pre-index period was defined for each index date.

Fig. 1.

Study design. ^aFor first 12 months on market, BGF prescriptions limited to 14-day supply; after that, prescribers could issue a supply for up to 3 months per prescription; ^bTime when all records satisfying selection criteria were included; ^c≥ 30 days of overlap of supply for ICS + LABA + LAMA, ICS + LABA/LAMA, or ICS/LABA + LAMA; ^dDate of first triple inhaler prescription claim. If > 1 triple therapies used during observation period, multiple index dates were defined (patient can appear in > 1 cohort with different index dates). BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, MITT multiple-inhaler triple therapy, y year

Study Population

The study population comprised patients with COPD treated with triple therapy (SITT or MITT) selected from all participants registered in the MDV claims database.

Patients were assigned to one of four cohorts: early BGF adopters (who initiated BGF during the first 12 months following market approval [September 1, 2019–August 31, 2020]), contemporary BGF users (who initiated BGF after the end of the 1-year launch period [September 1, 2020–September 30, 2021]), FF/UMEC/VI users (who initiated FF/UMEC/VI following market approval [May 1, 2019–September 30, 2021]), and MITT users (who initiated a MITT [including ICS + LAMA + LABA, ICS + LABA/LAMA, ICS/LABA + LAMA during September 1, 2019–September 30, 2021]). MITT was defined by having ≥ 30 days of overlap (in days) of supply for ICS + LAMA + LABA, ICS + LABA/LAMA, or ICS/LABA + LAMA, where “/” denotes multiple treatments provided in one inhaler and “+” denotes separate inhalers. The early BGF adopter and contemporary BGF user groups were further stratified into those who had BGF as the first triple therapy and those who switched to BGF, with patients who initiated BGF as the first triple therapy further divided into treatment-naive and step-up cohorts. Data from these sub-cohorts are briefly described in the Supplementary Material.

Patients included in the study had ≥ 1 inpatient record (hospitalization) or ≥ 2 outpatient records of a COPD diagnosis (International Classification of Disease, 10th revision [ICD-10] code J41–J44) during the selection period. Eligible patients were aged ≥ 40 years at the time of first COPD diagnosis and had ≥ 1 prescription record of BGF (between September 1, 2019–September 30, 2021) and/or FF/UMEC/VI (between May 1, 2019–September 30, 2021), and/or MITT (between September 1, 2019–September 30, 2021). Patients included also had ≥ 12 months of historical data before each index date. Patients were excluded from the FF/UMEC/VI and MITT user cohorts if they had evidence of the respective treatment during the 12-month pre-index period. For the BGF user subgroup cohorts, patients were excluded from the BGF treatment-naive cohort if they had evidence of monotherapy, dual therapy, or triple therapy during the 12-month pre-index period, and were excluded from the BGF step-up cohort if they had evidence of any triple therapy during the 12-month pre-index period.

The study protocol was approved by the NPO-MINS Institutional Review Board (Approval No. 220216) in July 2022. This observational study was performed in accordance with ethical principles consistent with the Ethical Guidelines for Biomedical Research Involving Human Subjects, the Declaration of Helsinki, and the Japanese Act on the Protection of Personal Information. Permission was obtained to access and use the data from the MDV database. Informed consent was waived because the data were anonymized.

Study Measures

Demographic and clinical characteristics were evaluated during the 12-month pre-index baseline period. This included smoking status, based on the self-reported Brinkman index [14], calculated as the number of cigarettes smoked per day times by smoking years; if the Brinkman Index was > 0, the patient was considered as a smoker, irrespective of their current smoking status; when the Brinkman Index was 0, the patient was considered as a non-smoker. Specific COPD diagnosis among patients with ≥ 1 COPD ICD-10 diagnosis code (J41.x, J42, J43x, J44.x) during the pre-index period was evaluated, and asthma + COPD was determined on the basis of the entire pre-index period using ICD-10 codes J45 and J46, and MDV code 8850253. Multimorbidity (i.e., the presence of multiple medical conditions) was assessed using the Elixhauser Comorbidity Index (ECI) score (excluding Elixhauser concept of chronic pulmonary disease, which includes asthma and COPD) [15].

During the baseline pre-index period, patients with evidence of COPD exacerbations were assessed, and exacerbations were classified as moderate or severe. Moderate exacerbations were defined as those with a hospital clinic outpatient visit with a diagnosis code for COPD (J41–J44) or for acute bronchitis (J20–J22), and either an outpatient dispensation claim for injectable steroids or oral corticosteroids (OCS) within 5 days of the visit and for a maximum duration of 15 days, or for an antibiotic commonly used in treating respiratory tract infections within 5 days of the visit (see Supplementary Material, Table S1 for a list of defined respiratory antibiotics). Severe exacerbations were defined as those with a hospitalization of ≥ 1 night with a record of COPD diagnosis (J41–J44) as main disease name or disease name behind hospitalization. All exacerbations of any severity starting within a 14-day period were counted as the same exacerbation event and the exacerbation with highest severity was assigned to the event.

Treatment history was assessed during the 12-month pre-index period to determine the usage of COPD-related medications. HCRU was also assessed during the 12-month pre-index period. HCRU assessments included the number of hospital outpatient clinic visits, the number of hospitalizations, average length of stay, and type of admission wards.

Statistical Analyses

Analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) and descriptive statistics were produced for all study outcomes. As the study was descriptive, no formal statistical testing was conducted.

Missing values in prescription duration were imputed by calculating the median gap between two consecutive prescriptions by individual drug (receipt code) when the number of observations was ≥ 30. In all other variables, no imputation was performed, statistics were reported on the basis of the available valid values.

Results

Patient Disposition

In total, 17,958 patients initiating at least one triple therapy were identified. The study included 3194 BGF initiators, 11,179 patients in the FF/UMEC/VI user cohort, and 5931 patients in the MITT user cohort. Out of the BGF initiators, 636 were in the BGF early adopter cohort, 401 of whom had BGF as the first triple therapy and 235 switched from another triple therapy; 2558 were in the contemporary BGF user cohort, 1466 of whom had BGF as the first triple therapy and 1092 switched from another triple therapy. Patient disposition is depicted in Fig. 2.

Fig. 2.

Flowchart of study population. BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium, MITT multiple-inhaler triple therapy, TT triple therapy

Patient Demographic and Clinical Characteristics

Patient demographics were generally similar across cohorts; mean (SD) age at index date was 73.7 (10.0) years in the early BGF adopter cohort, 73.1 (10.1) years in the contemporary BGF user cohort, 72.9 (9.9) years in the FF/UMEC/VI user cohort, and 72.0 (11.3) years in the MITT user cohort (Table 1). Use of the MDV database provided good coverage of the older population, especially those with severe health status [7]. There was a high proportion of male patients in each cohort; 74.4% in the early BGF adopters, 72.5% in the contemporary BGF users, 77.1% in the FF/UMEC/VI users, and 66.5% in the MITT users (Table 1). Consistent with the main cohorts, patient demographics were generally balanced across the BGF cohort subgroups (Supplementary Material, Table S2).

Table 1.

Baseline demographic characteristics by cohort

Characteristic	Early BGF adopters (N = 636)	Contemporary BGF users (N = 2558)	FF/UMEC/VI users (N = 11,179)	MITT users (N = 5931)
Age at index, mean (SD), years	73.7 (10.0)	73.1 (10.1)	72.9 (9.9)	72.0 (11.3)
Age category at index, n (%)
40–59 years	56 (8.8)	262 (10.2)	1108 (9.9)	835 (14.1)
60–64 years	40 (6.3)	167 (6.5)	782 (7.0)	406 (6.8)
65–69 years	85 (13.4)	316 (12.4)	1448 (13.0)	736 (12.4)
70–74 years	121 (19.0)	543 (21.2)	2364 (21.1)	1143 (19.3)
≥ 75 years	334 (52.5)	1270 (49.6)	5477 (49.0)	2811 (47.4)
Sex, n (%)
Female patients	163 (25.6)	704 (27.5)	2563 (22.9)	1986 (33.5)
Male patients	473 (74.4)	1854 (72.5)	8616 (77.1)	3945 (66.5)
Index year, n (%)
2019	274 (43.1)	0 (0.0)	3466 (31.0)	1566 (26.4)
2020	362 (56.9)	930 (36.4)	4192 (37.5)	3078 (51.9)
2021	0 (0.0)	1628 (63.6)	3521 (31.5)	1287 (21.7)

BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, MITT multiple-inhaler triple therapy, SD standard deviation

While clinical information is not readily obtainable in the MDV, data presented in Table 2 show that baseline clinical characteristics were similar across cohorts. Unspecified COPD was the most frequently occurring diagnosis code for all patients before initiating triple therapy (60.7–75.6% across cohorts), but 36.0–40.5% had diagnoses for chronic bronchitis and 29.4–32.7% had diagnoses of emphysema, though not mutually exclusive. Asthma + COPD was reported in 464 (73.0%) patients in the early BGF adopter cohort (221 [94.0%] of whom had switched to BGF from another triple therapy), 1898 (74.2%) in the contemporary BGF user cohort (975 [89.3%] of whom had switched to BGF from another triple therapy), 8457 (75.7%) in the FF/UMEC/VI user cohort, and 5013 (84.5%) in the MITT user cohort (Table 2 and Supplementary Material, Table S3). Of note, only 1063 (5.9%) had EOS data available. The mean (SD) EOS percentage per patient among the early BGF adopter and contemporary BGF user cohorts was 5.3% (6.9) and 3.3% (3.3), respectively, compared with 4.2% (5.1) and 3.7% (3.8) in the FF/UMEC/VI and MITT cohorts, respectively. Likewise, while data related to spirometry was limited, approximately 52.0% of patients had an indication that a test was performed.

Table 2.

Baseline clinical characteristics by cohort

Characteristic	Early BGF adopters (N = 636)	Contemporary BGF users (N = 2558)	FF/UMEC/VI users (N = 11,179)	MITT users (N = 5931)
Body mass index, at index, mean (SD), kg/m2	21.9 (4.1)	22.3 (4.3)	22.5 (4.3)	22.6 (4.7)
Smoking status, n (%)
Unknown/missing	405 (63.7)	1762 (68.9)	7711 (69.0)	3728 (62.9)
Non-smoker	69 (10.8)	230 (9.0)	1027 (9.2)	762 (12.8)
Smoker	162 (25.5)	566 (22.1)	2441 (21.8)	1441 (24.3)
Specific COPD diagnosis (pre-index period), n (%)
J41.x (Simple and mucopurulent chronic bronchitis)	0 (0.0)	8 (0.3)	10 (0.1)	12 (0.2)
J42 (Unspecified chronic bronchitis)	229 (36.0)	999 (39.1)	4474 (40.0)	2402 (40.5)
J43.x (Emphysema)	187 (29.4)	793 (31.0)	3655 (32.7)	1804 (30.4)
J44.x (Other COPD)	501 (78.8)	1962 (76.7)	8304 (74.3)	3847 (64.9)
J44.0 (COPD with [acute] lower respiratory infection)	8 (1.3)	24 (0.9)	72 (0.6)	58 (1.0)
J44.1 (COPD with [acute] exacerbation)	75 (11.8)	277 (10.8)	1023 (9.2)	641 (10.8)
J44.9 (COPD, unspecified)	481 (75.6)	1902 (74.4)	7990 (71.5)	3600 (60.7)
Asthma + COPD overlap (pre-index period), n (%)a	464 (73.0)	1898 (74.2)	8457 (75.7)	5013 (84.5)

BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, COPD chronic obstructive pulmonary disease, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, MITT multiple-inhaler triple therapy, SD standard deviation

^aUsing ICD-10 codes J45 and J46 and MDV code 8850253

Multimorbidity

In the 12-month pre-index period, the frequency of multimorbidity was comparable across cohorts, with over half of patients in all cohorts (52.4–56.9%) having a weighted ECI ≥ 5, and > 80% of patients having ≥ 1 ECI-related condition (Table 3). Cardiovascular diseases (including uncomplicated hypertension, congestive heart failure, cardiac arrythmias, and peripheral vascular disease) were among the most frequently observed, with similar rates across cohorts (Fig. 3). In the 12-month pre-index period, the frequency of multimorbidity was similar across BGF cohort subgroups compared with the main cohorts (Supplementary Material, Table S4).

Table 3.

Baseline multimorbidity (reported in > 5% of any cohort) during the 12-month pre-index period by cohort

	Early BGF adopters (N = 636)	Contemporary BGF users (N = 2558)	FF/UMEC/VI users (N = 11,179)	MITT users (N = 5931)
Conditiona, n (%)
Any	547 (86.0)	2090 (81.7)	9029 (80.8)	4912 (82.8)
Cardiac arrhythmias	117 (18.4)	449 (19.5)	2127 (19.0)	1193 (20.1)
Congestive heart failure	181 (28.5)	762 (29.8)	3031 (27.1)	1638 (27.6)
Deficiency anemia	76 (11.9)	253 (9.9)	1034 (9.2)	617 (10.4)
Depression	45 (7.1)	175 (6.8)	545 (4.9)	364 (6.1)
Diabetes, complicated	29 (4.6)	167 (6.5)	675 (6.0%)	369 (6.2%)
Diabetes, uncomplicated	127 (20.0)	341 (13.3)	1213 (10.9)	734 (12.4)
Fluid and electrolyte disorders	103 (16.2)	427 (16.7)	1551 (13.9)	1046 (17.6)
Hypertension, uncomplicated	311 (48.9)	1268 (49.6)	5548 (49.6)	2974 (50.1)
Liver disease	101 (15.9)	406 (15.9)	1784 (16.0)	945 (15.9)
Metastatic cancer	34 (5.3)	117 (4.6)	429 (3.8)	250 (4.2)
Other neurological disorders	34 (5.3)	92 (3.6)	372 (3.3)	247 (4.2)
Peptic ulcer disease, excluding bleeding	149 (23.4)	583 (22.8)	2505 (22.4)	1377 (23.2)
Peripheral vascular disorders	78 (12.3)	326 (12.7)	1258 (11.3)	618 (10.4)
Renal failure	44 (6.9)	195 (7.6)	719 (6.4)	429 (7.2)
Rheumatoid arthritis/collagen vascular diseases	60 (9.4)	180 (7.0)	777 (7.0)	450 (7.6)
Solid tumor without metastasis	169 (26.6)	668 (26.1)	2717 (24.3)	1442 (24.3)
Valvular disease	43 (6.8)	185 (7.2)	802 (7.2)	404 (6.8)
Solid tumor without metastasis (lung cancer)	82 (12.9)	287 (11.2)	1108 (9.9)	534 (9.0)
Weighted ECI ≥ 5, n (%)	362 (56.9)	1408 (55.0)	5853 (52.4)	3297 (55.6)
Patients with ≥ 1 ECI comorbidity	547 (86.0)	2090 (81.7)	9029 (80.8)	4912 (82.8)

BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, ECI Elixhauser Comorbidity Index, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, MITT multiple-inhaler triple therapy

^aECI comorbidities

Fig. 3.

Most frequently observed cardiovascular diseases during the 12-month pre-index period by cohort. BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, MITT multiple-inhaler triple therapy

Prior Therapy (Maintenance and Non-maintenance) Treatment Patterns

Almost three quarters of patients across cohorts had some baseline maintenance treatment in the 12-month pre-index period. ICS/LABA dual therapy was the most common maintenance therapy across cohorts and was used by more than 38.0% of patients in every cohort (Fig. 4). Slightly more patients in the FF/UMEC/VI and MITT cohorts received ICS/LABA treatment (48.3% and 51.7%, respectively) compared with the BGF cohorts (38.2% and 40.9%, respectively), and more patients in the BGF cohorts had received LABA/LAMA treatment (31.1% and 26.0%, respectively) than in the MITT cohort (17.3%; Fig. 4). Approximately one-quarter of patients across cohorts had not received any maintenance treatment in the 12-month pre-index period.

Fig. 4.

Maintenance therapy treatment patterns^a during the 12-month pre-index period by cohort. ^aPatients within each cohort could have received more than one type of maintenance therapy in the 12-month index period and therefore percentages may exceed 100%. BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, ICS inhaled corticosteroid, LABA long-acting β₂-agonist, LAMA long-acting muscarinic antagonist, MITT multiple-inhaler triple therapy

In the early BGF adopter and contemporary BGF user cohorts, the most common COPD maintenance medication classes ≤ 90 days before initiation of BGF were ICS/LABA single-inhaler dual therapy (20.9% and 24.9%, respectively) and LABA/LAMA single-inhaler dual therapy (19.5% and 16.1%, respectively). At ≤ 90 days before initiation of BGF, over half of patients in the early BGF adopter and contemporary BGF user cohorts were using ≥ 1 COPD maintenance medication (56.0% and 55.6%, respectively). However, the number of days from the end of the previous maintenance treatment to BGF initiation was higher in the contemporary user cohort (mean, 41.7 [SD 25.2]; median, 35 [IQR 25; 56]) compared with the early BGF user cohort (mean, 34.7 [SD 24.7]; median, 28 [IQR 14; 56]).

Differences were observed in supplemental oxygen use for patients in the BGF cohorts versus the other cohorts. Oxygen therapy was dispensed in 12.9% of the early BGF adopter cohort, 12.7% of the contemporary BGF user cohort, 9.3% of the FF/UMEC/VI user cohort, and 6.5% of the MITT user cohort.

Treatment patterns with any respiratory antibiotic in the 12-month pre-index period were also similar across cohorts, with a slight trend for more antibiotic use in the early BGF adopter cohort; respiratory antibiotics were received by 54.7% of the early BGF adopter cohort, 47.1% of the contemporary BGF user cohort, 44.2% of the FF/UMEC/VI user cohort, and 50.2% of the MITT user cohort.

Non-maintenance therapy treatment patterns in the 12-month pre-index period were generally similar across BGF and MITT cohorts, with slightly lower proportions observed for the FF/UMEC/VI user cohort. Any rescue inhaler was used by 31.3% of the early BGF adopter cohort, 28.6% of the contemporary BGF user cohort, 27.6% of the FF/UMEC/VI user cohort, and 33.8% of the MITT user cohort. A short-acting β₂-agonist (SABA) was used by 31.1%, 28.5%, 27.6%, and 33.7% of patients in the BGF early adopter, contemporary BGF, FF/UMEC/VI, and MITT cohorts, respectively, and a short-acting muscarinic antagonist (SAMA) was used by 0.5%, 0.3%, 0.2%, and 0.4% of patients in the BGF early adopter, contemporary BGF, FF/UMEC/VI, and MITT cohorts, respectively. There was no use of a SAMA/SABA combination in one device in any cohort. At least one prescription of OCS in the prior 12 months occurred among 26.6% of the early BGF adopter cohort, 24.1% of the contemporary BGF user cohort, 21.1% of the FF/UMEC/VI user cohort, and 26.2% of the MITT user cohort.

Exacerbations

There was a trend towards a higher proportion of patients having experienced moderate/severe exacerbations, moderate exacerbations, or severe exacerbations during the 12-month baseline period in the BGF early adopter cohort compared with the other cohorts (Fig. 5). The mean (SD) number of moderate/severe COPD exacerbations before initiation of triple therapy was 0.3 (0.7) in the early BGF adopter cohort, 0.2 (0.7) in the contemporary BGF user cohort, and 0.2 (0.6) in both the FF/UMEC/VI and MITT user cohorts. The mean (SD) number of moderate COPD exacerbations or severe COPD exacerbations before initiation of triple therapy, respectively, was 0.2 (0.5) and 0.1 (0.5) in the early BGF adopter cohort, 0.1 (0.6) and 0.1 (0.3) in the contemporary BGF user cohort, and 0.1 (0.5) and 0.1 (0.3) in both the FF/UMEC/VI and MITT user cohorts. The proportion of patients with prior severe exacerbations across cohorts ranged from 5.3% to 8.8%. The percentage of patients with ≥ 1 moderate/severe exacerbation was 18.2% in the BGF early adopter cohort, 14.3% in the contemporary BGF user cohort, 13.1% in the FF/UMEC/VI user cohort, and 14.0% in the MITT user cohort. The distribution of the number of prior exacerbations during the 12-month pre-index baseline period by severity are presented in the Supplementary Materials (Table S5).

Fig. 5.

Proportions of patients with moderate/severe exacerbations, moderate exacerbations, or severe exacerbations during the 12-month pre-index period by cohort. BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium/vilanterol, MITT multiple-inhaler triple therapy

Approximately twice as many patients who switched to BGF from another triple therapy (27.2% for the BGF early adopter cohort; 19.0% for the contemporary BGF user cohort) had a moderate or severe exacerbation than those who received BGF as their first triple therapy (13.0% for the BGF early adopter cohort; 10.8% for the contemporary BGF user cohort) in both the early adopters and contemporary cohorts, respectively (Supplementary Material, Table S3).

Healthcare Resource Utilization During the 12-Month Pre-index Period

Hospital Clinic Outpatient Visits

The mean (SD) number of outpatient visits before initiation of triple therapy was 14.1 (19.1), 11.5 (11.1), 11.1 (11.7), and 11.0 (13.3) per patient for the BGF early adopter, contemporary BGF, FF/UMEC/VI, and MITT cohorts, respectively (Table 4). When grouped across specialties, most outpatient visits were observed for Respiratory/Pulmonology/Allergy specialties (64.2%, 67.8%, 64.7%, and 55.9% of patients in the BGF early adopter, contemporary BGF user, FF/UMEC/VI user, and MITT user cohorts, respectively). Cardiovascular-related visits were the third most frequently observed specialty outpatient visits (occurring in 18.7–22.8% of patients across cohorts) after Internal Medicine-related visits (occurring in 31.6–36.7% of patients across cohorts; Table 4).

Table 4.

HCRU during the 12-month pre-index baseline period by cohort

	Early BGF adopters (N = 636)	Contemporary BGF users (N = 2558)	FF/UMEC/VI users (N = 11,179)	MITT users (N = 5931)
Number of outpatient visits per patient, mean (SD)	14.1 (19.1)	11.5 (11.1)	11.1 (11.7)	11.0 (13.3)
0	57 (9.0)	146 (5.7)	688 (6.2)	668 (11.3)
1	26 (4.1)	78 (3.0)	368 (3.3)	298 (5.0)
2	22 (3.5)	70 (2.7)	365 (3.3)	244 (4.1)
3	25 (3.9)	87 (3.4)	404 (3.6)	231 (3.9)
≥ 4	506 (79.6)	2177 (85.1)	9354 (83.7)	4490 (75.7)
Number of all-cause hospitalizations per patient/year, mean (SD)	0.4 (0.6)	0.4 (0.6)	0.3 (0.6)	0.3 (0.6)
Average length of hospital stay, mean (SD)	12.1 (12.8)	12.8 (16.0)	10.9 (12.9)	10.8 (12.5)
Hospital outpatient clinic specialty type, n (%)
Respiratory/Pulmonology/Allergy	408 (64.2)	1734 (67.8)	7233 (64.7)	3315 (55.9)
Internal Medicine	201 (31.6)	809 (31.6)	3924 (35.1)	2178 (36.7)
Cardiovascular	119 (18.7)	583 (22.8)	2406 (21.5)	1154 (19.5)
Orthopedic/Rheumatology	116 (18.2)	457 (17.9)	1804 (16.1)	1081 (18.2)
Urology and Genitourinary Medicine	105 (16.5)	449 (17.6)	1899 (17.0)	1002 (16.9)
Gastroenterology	100 (15.7)	407 (15.9)	1645 (14.7)	847 (14.3)
Neurology/Psychiatry	93 (14.6)	313 (12.2)	1204 (10.8)	663 (11.2)
Ophthalmology	69 (10.8)	234 (9.1)	1170 (10.5)	622 (10.5)
Dermatology	65 (10.2)	234 (9.1)	1083 (9.7)	586 (9.9)
General Surgery	52 (8.2)	226 (8.8)	1014 (9.1)	597 (10.1)
Endocrinology and Metabolism	38 (6)	134 (5.2)	568 (5.1)	283 (4.8)
Other/unknown	100 (15.7)	421 (16.5)	1687 (15.1)	940 (15.8)
≥ 1 pneumonia-specific hospitalization, n (%)	41 (6.4)	149 (5.8)	678 (6.1)	441 (7.4)

BGF budesonide/glycopyrronium/formoterol fumarate dihydrate, FF/UMEC/VI fluticasone furoate/umeclidinium, HCRU healthcare resource utilization, MITT multiple-inhaler triple therapy, SD standard deviation

Hospitalization

The mean (SD) number of all-cause hospitalizations was 0.8 (1.2) per patient in the BGF early adopter cohort, 0.6 (1.1) per patient in the contemporary BGF user cohort, 0.6 (1.1) per patient in the FF/UMEC/VI user cohort, and 0.7 (1.1) per patient in the MITT user cohort. The mean (SD) length of stay was 12.1 (12.8), 12.8 (16.0), 10.9 (12.9), and 10.8 (12.5) days for the BGF early adopter, contemporary BGF user, FF/UMEC/VI user, and MITT user cohorts, respectively (Table 4). Overall 56–66% of hospital admissions were missing specific data on the ward of admission. Among the remaining 34–44% of patients with ≥ 1 hospitalization and available ward data, the most frequent wards of admission across cohorts included Internal Medicine (23.4–33.2%), Pulmonary Medicine (17.1–26.5%), Pulmonology (7.5–10.6%), and Cardiovascular Medicine (6.8–8.9%) (Supplementary Material, Table S6). A higher proportion of patients in the MITT user cohort reported ≥ 1 pneumonia-specific hospitalization compared to other cohorts (6.4%, 5.8%, 6.1%, and 7.4% in BGF early adopter, contemporary BGF user, FF/UMEC/VI user, and MITT user cohorts, respectively; Table 4).

Discussion

In this study, patients with COPD were predominately male, aged ≥ 70 years, and had substantial multimorbidity, often including interrelated cardiovascular diseases. These findings are consistent with previous observational studies in Japan [16–18]. There was also a high prevalence of asthma + COPD overlap and prior ICS/LABA maintenance therapy among all cohorts, with the latter being consistent with JRS guidelines on the use of triple therapy for COPD [6].

The proportion of BGF users in this analysis who had a previous moderate exacerbation (9.0–10.7%) was lower than that observed in a similar study in the USA (41.4%) [19]. Exacerbations have historically been lower or underreported in Japan, in comparison with other countries [17, 20, 21], which could explain the low moderate-to-severe exacerbation rates observed in this study. This difference could also be related to the use of hospital claims data in this study, where a hospital visit (outpatient or inpatient) was required for a patient’s exacerbation to be recorded. Exacerbations occurring outside a hospital visit would not be captured in this analysis, thus reducing the overall proportion of recorded exacerbations. However, the frequency of prior hospitalized (i.e., severe) exacerbations was relatively high across cohorts, ranging from 5.3% to 8.8%. HCRU was also relatively high across cohorts, with a mean number of 11.0–14.1 outpatient visits during the pre-index baseline period.

Similar to observations in a USA-focused study, which reported ICS/LABA as the most common maintenance therapy in patients with COPD (35.8%) [19], ICS/LABA was also the most common maintenance treatment during the baseline period for BGF users (38.2–40.9%); this is consistent with JRS guidelines, given the high frequency of asthma + COPD overlap in the study population, as guidelines recommend that these patients are prioritized for ICS/LABA. It was observed that slightly more patients in the FF/UMEC/VI and MITT user cohorts received ICS/LABA treatment compared with the BGF cohorts. This could in part be related to the slightly higher presence of asthma + COPD in the FF/UMEC/VI and MITT user cohorts. Treatment with ICS-containing therapy (including ICS/LABA) in patients with asthma + COPD would be in accordance with guidance in Japan [6]. Furthermore, FF/UMEC/VI is indicated for the treatment of asthma whereas BGF does not currently have an asthma indication [22]. As such, the higher percentage of patients with prior ICS/LABA could also be related to a step-up to FF/UMEC/VI in patients with inadequate response to ICS/LABA. Patterns of short-term OCS use and respiratory antibiotics in the 12-month pre-index period could also be an indicator of exacerbation history.

While a high level of cardiovascular multimorbidity was observed in this study population, only approximately one-fifth of hospital outpatient visits were related to a cardiovascular specialty, indicating that patients may not be receiving specialist cardiovascular care, when it could be of benefit. This finding is important because cardiovascular-related deaths have been reported in a substantial proportion of patients with COPD [23, 24].

Differences in the exacerbation rate between the BGF early adopters and the other triple therapy cohorts could be expected, as the study was designed to compare patients initiating new (BGF) and established (FF/UMEC/VI and MITT) therapies. Known as channeling bias, physicians may have prescribed BGF preferentially for more severe, symptomatic patients upon initial approval, potentially as a result of a lack of familiarity or experience with the treatment, or because these patients are more likely to be in therapeutic failure, and a new treatment in the market could be seen as an opportunity. Results showing that more patients who switched to BGF had a history of exacerbations than those receiving BGF as their first triple therapy seem to support the existence of channeling bias. This could also be related to the 12-month launch period, during which prescriptions of BGF were limited to a 14-day supply, as physicians may have preferred to prescribe long-term treatments for more stable patients, therefore reserving the 14-day supply of BGF for those with more severe disease. Along these lines, the lower percentage of patients with an exacerbation history in the BGF as first triple therapy subgroups appears to be driven by treatment-naive patients, in whom a very low (< 5%) proportion had evidence of exacerbations at baseline. The proportion of patients with prior exacerbations was similar among those who stepped up from dual therapies to BGF, compared with those who switched to BGF from another triple therapy.

It is also possible that the 12-month launch period limiting BGF to a 14-day supply may have resulted in differences in prescription patterns compared with triple therapies that were not subject to such a restriction. Furthermore, the time points for initiation of triple therapy in each cohort means that the pre-index baseline period for some patients may have occurred during the coronavirus disease 2019 (COVID-19) pandemic. Therefore, it is possible that COVID-19 restrictions may have impacted prescribing patterns, especially if some medications could not be prescribed for long durations.

A key study strength is that it is the first real-world study to assess baseline characteristics, patterns of COPD medication use, exacerbation history, and HCRU in patients with COPD initiating triple therapy in Japan. Furthermore, the study was large and utilized the well-established MDV claims database, which includes approximately 45 million insured patients from over 480 hospitals [8].

Limitations of the study must be considered when interpreting these data. First, although some findings from the current study align with previous reports [16–19], the generalizability of the findings outside of Japan may be limited owing to differences in healthcare systems, genetic backgrounds, and environmental factors. Also, the MDV data used are obtained predominantly from large hospitals responsible for acute care and the patients analyzed may be in poorer health than the general population, resulting in an underreporting of less severe events. For example, patients might have attended their general practitioner with a moderate exacerbation, which is not recorded in this database. In addition, it is not possible to identify care provided in different hospitals to the same patient during a given period, and only dispensing data are available whereas actual inhaler use is not. Further, while EOS values can be obtained via routine blood tests, the MDV data used in this study indicate that they may not be as readily reported (with only approximately 6.0% having EOS data). Therefore, it may not always be possible to rely on this measure for decision-making or insurer coverage. Similarly, results indicate that spirometry may not be used as regularly in patients with COPD as the Global Initiative for Obstructive Lung Disease report recommends [4]. As the current analyses were based on data from a claims database, detailed information on the clinical outcomes associated with exacerbations was not available. However, it is acknowledged that understanding the outcomes associated with exacerbations and with the initiation of inhaled triple therapy will be important for future analyses to examine. Additionally, as this study population comprised patients that may have been selected during the COVID-19 pandemic, the associated disruptions in healthcare and changes in behavior during that time [25] may have led to decreased exposure to viral respiratory infections, potentially impacting the generalizability of these results. Lastly, it should also be noted that the MDV database has limited contemporary lung function data so validation of a COPD diagnosis or a more detailed assessment of COPD severity could not be conducted.

Conclusions

Among patients in Japan initiating triple therapy for COPD, most had concomitant asthma at baseline and substantial multimorbidity, including a high level of cardiovascular disease, many had previously received ICS/LABA dual therapy as maintenance treatment, and there was relatively high HCRU. Patients prescribed BGF in the initial period after launch were more likely to have a history of previous exacerbations than the other cohorts, potentially indicative of more severe disease at baseline.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Author contributions to development of this report are as follows: conceptualization (Koichiro Takahashi, Naoyuki Makita, Seham Issa, Isao Matsumoto, Hana Müllerová); methodology (Naoyuki Makita, Johann Castañeda-Sanabria, Ramzi Argoubi, Seham Issa, Hana Müllerová); data curation (Ramzi Argoubi, Gregoire Nowacki); supervision (Koichiro Takahashi, Johann Castañeda-Sanabria, Ramzi Argoubi, Hana Müllerová); formal analysis (Johann Castañeda-Sanabria, Ramzi Argoubi, Gregoire Nowacki); software (Gregoire Nowacki); visualization (Ramzi Argoubi, Gregoire Nowacki, Yuri Yoshida); investigation (Johann Castañeda-Sanabria, Ramzi Argoubi); project administration (Johann Castañeda-Sanabria, Ramzi Argoubi, Seham Issa); funding acquisition (Hana Müllerová); writing – original draft (Yuri Yoshida, Hana Müllerová); writing – reviewing & editing (Koichiro Takahashi, Naoyuki Makita, Johann Castañeda-Sanabria, Seham Issa, Isao Matsumoto, Yuri Yoshida, Hana Müllerová).

Funding

This study was sponsored by AstraZeneca. AstraZeneca also provided funding for development of the graphical abstract and for the journal’s Rapid Service fee and Open Access fee.

Data Availability

Data underlying the findings described in this manuscript may be obtained in accordance with AstraZeneca’s data sharing policy described at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure. Data for studies directly listed on Vivli can be requested through Vivli at www.vivli.org. Data for studies not listed on Vivli could be requested through Vivli at https://vivli.org/members/enquiries-about-studies-not-listed-on-the-vivli-platform/. The AstraZeneca Vivli member page is also available outlining further details: https://vivli.org/ourmember/astrazeneca/.

Declarations

Conflict of Interest

Koichiro Takahashi has received honoraria (lecture fees) from AstraZeneca. Naoyuki Makita, Isao Matsumoto, Yuri Yoshida, and Hana Müllerová are employees of AstraZeneca and may hold stock and/or stock options in the company. Johann Castañeda-Sanabria is a former employee of AstraZeneca and a current employee of IQVIA Real World Solutions, Global Epidemiology and Database Studies. Ramzi Argoubi, Grégoire Nowacki, and Seham Issa are employees of Oracle Life Sciences (formerly Cerner Enviza), which was contracted by AstraZeneca to conduct the study.

Ethical Approval

The study protocol was approved by the NPO-MINS Institutional Review Board (Approval No. 220216) in July 2022. This observational study was performed in accordance with ethical principles consistent with the Ethical Guidelines for Biomedical Research Involving Human Subjects, the Declaration of Helsinki, and the Japanese Act on the Protection of Personal Information. Permission was obtained to access and use the data from the MDV database. Informed consent was waived because the data were anonymized.