FormalPara Key Points

It is established that in patients with asthma, symptom control is a primary goal, as it leads to reduced burden to the patient and the healthcare system.

The first commercially available fixed-dose combination of a long-acting beta-agonist, an inhaled corticosteroid, and a long-acting muscarinic antagonist, a single inhaler of indacaterol/glycopyrronium/mometasone furoate, has been compared to both inhaled corticosteroid/long-acting beta-agonist fixed combinations and to the use of an inhaled corticosteroid/long-acting beta-agonist plus long-acting muscarinic antagonist (separate inhalers).

On the basis of the clinical results, a Markov model has been fed with local Italian economic inputs.

Under the assumptions of the model, it is predicted that the new option is cost effective against existing alternatives.

1 Introduction

According to the Global Initiative for Asthma strategy document, asthma is a heterogeneous disease, normally characterised by a chronic airflow limitation. It is defined by a history of respiratory symptoms such as wheeze, shortness of breath, chest tightness and cough that varies over time and in intensity, along with a variable limitation of expiratory airflow. Restriction of airflow may subsequently become persistent [1].

Asthma can affect patients of any age, but is more common in children than in adults [1]. More than 358 million people worldwide suffer from asthma, with 400,000 reported deaths due to the disease in 2015. The prevalence of asthma varies between 3 and 20% in different countries [2], in Italy, the prevalence is around 6%, thus affecting more than 3 million people [3,4,5].

The primary goal of asthma guidelines is to diagnose asthma correctly, achieve asthma control by reducing exacerbations, improve overall quality of life and educate patients on self-management of asthma. The long-term goals for asthma management are risk reduction and symptom control, leading to reductions in the burden to the patient and the risk of asthma-related death, exacerbation, airway damage, medication side effects and associated healthcare costs. Every guideline recommends a treatment step-up either by increasing the doses of inhaled corticosteroid (ICS) and/or the addition of controllers until patients with moderate-to-severe asthma achieve optimal disease control. Subsequently, patients should be frequently monitored to ensure that their disease is controlled with minimal drug therapy; once disease control is achieved, step-down treatment is recommended [1, 6,7,8].

In fact, according to Global Initiative for Asthma guidelines, the severity of asthma (mild, moderate or severe) is assessed retrospectively in relation to the level of treatment required to control symptoms and exacerbations and may change over months or years. Mild asthma is asthma controlled by step 1 or 2 of treatment, for example, with only medications as needed or with low-dose ICS or antileukotrienes. Moderate asthma is well controlled with step 3 treatment, for example, low doses of ICS/long‑acting beta‑agonist (LABA). Severe asthma requires step 4 or 5 treatment, for example, high doses of ICS/LABA, to prevent it from becoming “uncontrolled”, or asthma that remains uncontrolled despite treatment.

Some recent Italian studies report that, despite current treatment, more than 50% of patients in Global Initiative for Asthma Steps 4 and 5 are not adequately controlled and have experienced one or more exacerbations in the previous year [9, 10]. Patients with uncontrolled asthma may downplay or underestimate the severity of their disease and are at a higher risk of exacerbation, hospitalisation or death [11,12,13].

It should also be noted that the cost of asthma increases gradually with the worsening of disease control. Almost half (46.2%) of asthma expenditure in Italy is attributable to 25% of patients with the worst control [14]. In fact, according to the study carried out by Marcellusi and colleagues, the cost per patient ranges from €126 for intermittent asthma to €2169 for persistent severe asthma [15]. The once-daily Enerzair Breezhaler is the first LABA/long-acting muscarinic antagonist/inhaled corticosteroid fixed-dose combination available in Europe as a maintenance treatment for adult patients with asthma not adequately controlled with a maintenance combination of a LABA and a high dose of an ICS who experienced one or more asthma exacerbations in the previous year.

Two phase III studies, ARGON and IRIDIUM, investigated the effectiveness and safety of Enerzair Breezhaler. Regarding ARGON, in patients with uncontrolled asthma, single inhaler fixed-dose combinations of high-dose and medium-dose indacaterol/glycopyrronium/mometasone furoate (IND/GLY/MF) once daily (o.d.) were non-inferior to salmeterol/fluticasone twice daily (bid) plus tiotropium o.d. (SAL/FLU plus TIO). High-dose IND/GLY/MF o.d. demonstrated greater improvements in quality of life, lung function, asthma control and health status, and reduced moderate exacerbations [16].

With regard to IRIDIUM, the once-daily combination therapy of medium-dose and high-dose MF/IND/GLY, from a single inhaler, significantly improved lung function vs the respective once-daily MF/IND and high-dose twice-daily FLU/SAL, a well-established ICS/LABA combination. Both doses of MF/IND/GLY showed similarly large improvements in asthma control from baseline, with no difference between any of the treatments. The annualised rate of exacerbations was numerically lower with MF/IND/GLY vs the ICS/LABA comparators [17]. Through data derived from ARGON and IRIDIUM, published literature and regulatory disposition, we performed an economic analysis using a Markov model aiming to evaluate the cost effectiveness of Enerzair in treating patients with asthma who are uncontrolled despite treatment with LABA and a high dose of an ICS.

2 Methods

2.1 Model Description

Asthmatic patients’ clinical experience is simulated at a cohort level using a Markov model with two mutually exclusive states of health (“day-to-day symptoms”, “death”) over a lifetime horizon (50 years), until complete death of the cohort of patients. The simulated time is divided into equal cycles of 4 weeks to reflect the average duration of an asthmatic exacerbation: this choice is in line with recent economic assessments on asthma [18, 19]. The structure, shown in Fig. 1, is an adapted version of a cost-effectiveness model previously presented at the National Institute for Health and Care Excellence [20] in order to economically assess omalizumab in asthmatic patients with severe and persistent symptoms.

Fig. 1
figure 1

Markov model structure. ED emergency department, OCS oral corticosteroid

Full size image

Patients enter the model in the “day-to-day symptoms” state, which takes into account the costs and quality of life associated with the treatments in question. During their stay in the “day-to-day symptoms”, patients may experience one of three types of clinically significant severe exacerbations, which require treatment with oral corticosteroids, access to the emergency department or hospitalisation. Patients may also experience moderate exacerbations, defined when two or more of the following situations occur: (i) symptoms such as wheeze, cough, shortness of breath and tightness in the chest for at least 2 consecutive days; (ii) a 50% increase in the use of short-acting beta-agonists compared with the reference value and (iii) a 20% decrease in the forced expiratory volume in 1 s compared with the reference value.

The rates of exacerbations, which depend on the therapy the patient is receiving, are extrapolated from the clinical trials of Enerzair, IRIDIUM and ARGON. The impact of each type of exacerbation is considered by applying a decrease in utility and a cost for the treatment of this exacerbation.

Death is an absorbing state of health, which includes both death from asthma due to exacerbations and general mortality by age and sex. However, because asthma deaths have not been reported in Enerzair’s studies, nor, or only extremely rarely in the comparators’ phase III studies (one death in QUARTZ and PALLADIUM studies for Atectura, no deaths in AUSTRI and VESTRI studies for SAL/FLU, no deaths in the PrimoTinA Asma study for TIO), asthma mortality has not been considered. A half-cycle correction is not applied, as the cycle length is considered short compared with the total simulation time.

2.2 Assumptions

  • The model assumes that the clinical course of the disease is constant over the simulated period.

  • Day-to-day symptoms are assumed to include both symptom-free periods and periods with non-clinically significant exacerbations.

  • The model assumes that patients, regardless of treatment, do not discontinue treatment. This assumption is based on evidence that in clinical trials the probability of treatment discontinuation was negligible.

  • Mortality from asthma exacerbations is assumed to be low and negligible and is not included—no asthma-related deaths were observed in the trials.

  • As the trials did not measure it, the model assumes complete therapeutic adherence with Enerzair (i.e. 100%) and all other treatments compared.

2.3 Population

The target population considered in the model is adult patients with asthma aged 18 years and older who are inadequately controlled despite treatment with high-dose inhaled corticosteroids and LABAs. The main characteristics of the population considered in the model are assigned on the basis of clinical trial data. The mean age was 52 and 53 years and the female proportion 62% and 63% in the IRIDIUM and ARGON studies, respectively.

2.4 Time Horizon

The model simulates the clinical experience of patients over a lifetime horizon (50 years); both costs and clinical effectiveness are discounted at a rate of 3% per year, after the first year.

2.5 Treatments

The model compares the costs and consequences of Enerzair, a fixed-dose combination of MF, IND and GLY. The model considers the following dose: IND/GLY/MF 150/50/320 μg.

The following comparators are included in the model:

  • Atectura® Breezhaler® 150/320 µg o.d. (MF/IND);

  • Seretide® Diskus® 50/500 µg bid (SAL/FLU);

  • Salmeterol/fluticasone 50/500 µg twice daily plus Spiriva Respimat® (TIO) 5 µg o.d. (SAL/FLU plus TIO)

2.6 Clinical Inputs

2.6.1 Exacerbation Rate

Annual exacerbation rates (both severe and moderate exacerbations, Table 1) were extrapolated from the clinical studies. Among severe exacerbations, 90% are assumed to be treated with oral corticosteroids, 5% to require an emergency room and the remaining five to lead to hospitalisation, basing on previous economic research [21].

Table 1 Exacerbation rate
Full size table

2.6.2 Utility Values

The utility values associated with the “day-to-day symptoms” status were extrapolated from the data of the EQ-5D questionnaires of the ARGON and IRIDIUM clinical studies and are shown in Table 2. Disutilities associated with the most clinically severe exacerbations (oral corticosteroid burst, emergency department visit or hospital admission) were derived from the literature (Table 3). These disutilities are then applied to the utility associated with day-to-day symptoms in cycles where an exacerbation is experienced, to account for the decrease in quality of life.

Table 2 Utility values associated with the “day-to-day symptoms” status
Full size table
Table 3 Duration and disutility of exacerbations
Full size table

2.7 Economic Inputs

The cost categories considered in the model include pharmaceutical and exacerbation management costs.

2.7.1 Pharmaceutical Costs

The daily cost of drugs is calculated as the product of the unit cost and the dose consumed per day (Table 4). The unit cost of drugs is derived from the ex-factory price after mandatory discounts.

Table 4 Drug cost and dosage
Full size table

2.7.2 Exacerbation Management Costs

The cost of management of exacerbations has been quantified based on the type of exacerbation (Table 5):

  • Exacerbations requiring hospitalisation: the cost of admission was valued as the average between the national tariff for DRG 96 and DRG 97 weighted for their relative frequency in Italy in 2018, according to data collected from hospital discharge records [23, 24].

  • Exacerbations requiring emergency room access: the intensive short-term observation tariff, already used in the literature for analyses focusing on Italy, was used [25].

  • Exacerbations requiring treatment with oral corticosteroids: literature data were used that quantified the cost per consumption of drugs for the management of exacerbations as €21.09 [26]; furthermore, it was assumed that the patient makes a visit to the general practitioner whose cost was valued at €15.40 on the basis of the DYSCO study [27] and is derived from the average of the cost of the outpatient visit and the cost of the home visit weighted for the respective frequency, updated to 2020 Euros with the ISTAT consumer price index [28].

  • Moderate exacerbations: a similar cost was assumed as for severe exacerbations requiring oral corticosteroid treatment only.

Table 5 Cost per type of exacerbation
Full size table

2.8 Sensitivity Analysis

Uncertainties of input parameters, and their effect on estimated results, are evaluated through a probabilistic sensitivity analysis (PSA), conducted through 1000 simulations. A unique combination of parameters, randomly sampled from distributions within the confidence intervals for exacerbations and with a standard deviation of ± 10% of the mean value for utilities and costs, in the absence of a confidence interval, is used for each simulation. The PSA results are presented in the incremental cost-effectiveness plane and are used to estimate the cost-effectiveness acceptability curve (CEAC). The incremental cost-effectiveness plane shows the dispersion of the thousand results of the iterations, expressed as the respective incrementals of benefit [quality-adjusted life-year (QALY)] and cost of Enerzair compared to MF/IND or SAL/FLU or SAL/FLU plus TIO. The acceptability curve, based on the 1000 iterations, indicates the frequencies (percentage) with which the ICER is lower than a certain threshold value; that is, it provides the probability that, for a given threshold value, Enerzair is more cost effective than MF/IND or SAL/FLU or SAL/FLU plus TIO.

3 Results

3.1 IND/GLY/MF vs SAL/FLU Plus TIO

IND/GLY/MF in asthma maintenance therapy is associated with an incremental efficacy of 0.25 QALY and a cost saving of €3213.90 compared with therapy with SAL/FLU in combination with TIO.

3.2 IND/GLY/MF vs SAL/FLU

IND/GLY/MF in asthma maintenance therapy is associated with an incremental efficacy of 0.21 QALY and a cost increase of €2547.76 compared with therapy with SAL/FLU, with a resulting ICER of 11,897.36 €/QALY gained.

3.3 IND/GLY/MF vs MF/IND

IND/GLY/MF in asthma maintenance therapy is associated with an incremental efficacy of 0.34 QALY and a cost increase of €4745.91 compared with therapy with MF/IND, with a resulting ICER of 14,088.45 €/QALY gained.

3.4 Sensitivity Analysis

3.4.1 IND/GLY/MF vs SAL/FLU Plus TIO

The PSA, represented by the scatter plot in Fig. 2, shows robustness of the result to parameter uncertainty, as shown by the dense and symmetrical point cloud; the CEAC is not shown, as virtually all simulations indicate dominance.

Fig. 2
figure 2

Probabilistic sensitivity analysis scatterplot, indacaterol/glycopyrronium/mometasone furoate vs salmeterol/fluticasone plus tiotropium. QALY quality-adjusted life-year

Full size image

3.4.2 IND/GLY/MF vs SAL/FLU

The PSA, represented as scatter plot in Fig. 3A, has consistent findings with the base case, with moderate dispersion. When shown as a CEAC (Fig. 3B), IND/GLY/MF has about 80% probability of being the most cost-effective choice at the willingness-to-pay threshold of €15,000/QALY.

Fig. 3
figure 3

Probabilistic sensitivity analysis for indacaterol/glycopyrronium/mometasone furoate vs salmeterol/fluticasone. A Scatterplot, B cost-effectiveness acceptability curve. QALY quality-adjusted life-year

Full size image

3.4.3 IND/GLY/MF vs MF/IND

Furthermore, for this comparison, the PSA, represented as a scatter plot in Fig. 4A, has consistent findings with the base case, with moderate dispersion. When represented as a CEAC, it indicates that the treatment with IND/GLY/MF has about a 65% probability of being the most cost-effective choice at the willingness-to-pay threshold of €15,000/QALY (Fig. 4B).

Fig. 4
figure 4

Probabilistic sensitivity analysis for indacaterol/glycopyrronium/mometasone furoate vs mometasone furoate/indacaterol. A scatterplot, B cost-effectiveness acceptability curve. QALY quality-adjusted life-year

Full size image

4 Discussion

Given the limited resources available to the healthcare system, decision making ought to be based on both evidence and rational economic analyses. Modeling the problem provides a deeper insight into the consequences of treatment. First, it enables a more accurate assessment of the differences between the drugs considered by attributing uncertainty intervals to predicted health outcomes and costs. Second, a model permits an increase in the time horizon beyond the duration of the trial, hence assessing the long-term implications. Last, a sensitivity analysis can be carried out to investigate possible thresholds, which might invert results on expected outcomes. With the help of such a model, we compared IND/GLY/MF and SAL/FLU plus TIO or SAL/FLU or MF/IND and investigated uncertainty regarding costs and effectiveness through a PSA. Our analysis shows that the fixed-dose combination of high-dose IND/GLY/MF, when compared to SAL/FLU plus TIO, SAL/FLU, IND/MF or IND/GLY/MF is expected to be more effective and either cost saving or with an ICER well below conventional willingness-to-pay thresholds [29].

Potential limitations of this study are worth mentioning. First, effectiveness simulations are based on the short 24-week observation period set in the ARGON study, and are carried forward well beyond the time frame, being extrapolated to the rest of the analysis time frame. Given current data availability, it is not possible to compare these predictions with observed data, and the results should be interpreted keeping this in mind. Exacerbations were assessed as an exploratory endpoint and did not have the 52-week minimum follow-up, although greater reductions in moderate exacerbations with high-dose IND/GLY/MF vs high-dose SAL/FLU plus TIO were observed.

Another potential limitation is the missing input value available for compliance and adherence to the prescribed regimes; given the clinical relevance of the parameter, actual differences among treatment groups in the daily behaviour of the patients in this regard would change the results. Given the introduction of the new digital delivery device, we expect adherence to be favoured in the fixed-dose combination of the high-dose IND/GLY/MF group, thus, setting it as maximal and equal for all comparators is very likely a conservative assumption introduced in the model. Despite the limitation described, we believe that this study may offer a guide for authorities to make more informed decisions regarding the control of national pharmaceutical expenditure.

5 Conclusions

Our results indicate that high-dose IND/GLY/MF is cost effective compared with SAL/FLU plus TIO or SAL/FLU or IND/MF, as maintenance treatment in adults who have had at least one asthma attack (exacerbation) in the last year and whose asthma is not controlled well enough with an inhaled LABA together with a high dose of an inhaled corticosteroid.