Role of Long-Acting Bronchodilators in Patients with Clinical Asthma Remission

Up to 1/3 of patients with severe asthma achieve clinical remission under biologics defined as good symptom control and stable lung function in the absence of exacerbations and oral corticosteroid use for at least 1 year [1, 2]. After achieving remission, the question arises whether the inhaled background therapy may be reduced or even stopped to reduce the therapeutic burden, but also potential side effects.

Recently, the SHAMAL trial, a randomized controlled open-label trial, showed that reduction in inhaled corticosteroid (ICS)/long-acting beta-agonist (LABA) maintenance therapy from high-dose to lower doses or even anti-inflammatory reliever therapy only was feasible in patients with well-controlled severe eosinophilic asthma during benralizumab treatment without losing symptom control [3]. However, increases in the airway inflammation marker fractional exhaled nitric oxide (FeNO) and reduction in forced expiratory volume in 1 s (FEV1) occurred with ICS/LABA reduction in a dose-dependent manner. At the end of the 48-week-study, the patients who used ICS/LABA solely as anti-inflammatory reliever therapy had a mean FEV1 decrease of 150 mL and FeNO increase of 30 ppb compared to baseline [3].

Asthma being a type-2 inflammatory disease, ICSs are the basis of treatment, but high-dose ICSs are associated with systemic side effects [4]. Thus, with good asthma control reached by the addition of benralizumab, the focus of the trial was on reducing ICS. While it seems probable that increases in FeNO are mainly linked to the reduction in ICS, reduction in LABA may additionally play a role in the deterioration of FEV1 even though trough pre-bronchodilator measurements were performed. The SHAMAL study does not allow unambiguous differentiation of the effects of LABA and ICS as both treatments were used and reduced simultaneously from a single-inhaler device.

Pulmonary function is an important outcome parameter and a mean deterioration of 150 mL might be clinically relevant with regard to long-term outcome. For comparison, the increase in FEV1 compared to placebo in phase-3 trials of benralizumab as an add-on treatment to ICS/LABA in severe asthma was also in the range of 150 mL [5], even though the mechanism leading to improved FEV1 is different and independent for anti-inflammatory treatments compared to bronchodilators [6, 7].

The question arises as to what the future role of long-acting bronchodilators will be in the treatment of severe asthma once clinical remission is achieved during biological therapy. Long-acting muscarinic antagonists (LAMAs) are another class of bronchodilators that additionally reduce mucus secretion and have some anti-inflammatory effects reducing asthma exacerbations when added to ICS/LABA treatments [8]. Similar to LABAs, LAMAs were previously classified as “controllers” in asthma [9] but are not considered among the disease-modifying anti-asthmatic drugs in the new concept of remission [10] because the anti-inflammatory effect is limited and they should only be used in combination with other anti-inflammatory treatments such as ICS, not alone.

In Germany, even though not a prerequisite for prescribing biologics, LAMAs are widely used in patients with severe asthma: 80% of patients who newly start a biologic in the German Asthma Net (GAN) severe asthma registry are treated with a LAMA in addition to ICS/LABA [1]. In the SHAMAL study, previous LAMA treatment was not an exclusion criterion but had to be stopped before entering the study [3].

To explain this open question, I exemplarily use retrospective data from patients stopping tiotropium Respimat 2.5 µg × 2 hubs daily therapy before entering the SHAMAL trial (n = 11) and were included in the GAN registry at LMU Munich (with written informed consent, IRB approval 21-0436). Notably, this was clinical routine before inclusion into the trial. The patients in this sample mostly had late-onset, intrinsic asthma with a strong eosinophilic signature and frequently comorbid nasal polyps (see Table 1). Comparing pre-BD measurements before and after stopping tiotropium, FEV1 decrease (mean difference −158 mL, median difference −95 mL) occurred while high-dose ICS/LABA and benralizumab were continued unchanged (Table 1). There were large individual differences with some individuals displaying larger losses (3 patients lost >10% of initial FEV1 in l) and others remaining stable (Fig. 1), while FeNO and asthma control were unchanged (Table 1) and all patients subsequently entered the SHAMAL trial. Interestingly, there was a correlation between loss in FEV1 in l after LAMA withdrawal and higher FeNO during inhaled triple therapy plus benralizumab (r Spearman = −0.828, p = 0.009, n = 10), but no correlation to the highest blood eosinophil count before benralizumab, age at asthma onset, or asthma duration. It may be hypothesized that patients with persisting IL13-driven airway inflammation under benralizumab as signified by high FeNO, still benefit from LAMA as an additional bronchodilator.

Even though clinical remission had been achieved in all these patients, biological remission still seems several steps away in the portion of patients who had relevant worsening of obstructive ventilatory defect upon stopping LAMA despite ongoing biological therapy and high-dose ICS/LABA. Similar to findings in the SHAMAL study, it is surprising that the relevant loss of FEV1 in some patients was not associated with reporting of worsening symptoms. Here, poor perception of bronchoconstriction might play a role and has been reported as a risk factor for adverse outcomes [11].

This is a small retrospective, real-life dataset with inherent limitations and should only be regarded as incentive for further studies. Effects of LAMAs (and LABAs) need to be considered and differentiated from ICS to inform the current discussion of remission and reduction in inhaled treatments. LAMA are well-tolerated and safe, thus stopping them is not an imposing treatment goal per se. The role of lung function in the concept of remission merits further thought and long-term data. Currently, at least “stable” lung function is proposed as a clinical remission goal [2], but a more detailed analysis taking into account bronchial hyperresponsiveness and bronchodilator response will be necessary to characterize biological remission.