Effects of Adding Tiotropium or Aclidinium as Triple Therapy Using Impulse Oscillometry in COPD

The results of this proof of concept study showed no significant differences between randomized treatments in any IOS or spirometry outcomes measured at trough after chronic dosing with TIO and ACL when used as triple therapy in patients with COPD. We found no significant difference between treatments in the primary outcome of R5, although neither drug produced any significant improvements in R5 from baseline. Within the power constraints of the sample size, we cannot exclude the possibility that we may have missed a difference in R5 smaller than 0.1 kPa L⁻¹ s, which we considered to be a clinically important difference. Other IOS outcomes including R20, X5, and AX were not significantly improved from baseline with either drug from baseline. Nonetheless, we showed that both treatments produced comparable significant improvements from baseline in RF, while ACL also produced a significant improvement in R5–R20. However, the clinical relevance of small changes in IOS outcomes in COPD is uncertain as there are currently no published data with regard to minimal important differences. In one cross-sectional study in COPD patients, peripheral airway resistance correlated with FEV₁ and FEF_25–75 but not with the Medical Research Council (MRC) dyspnea score [14]. In a baseline cross-sectional analysis of the ECLIPSE study in COPD, IOS was found to be reproducible and was able to define the severity of disease according to global initiative for chronic obstructive lung disease (GOLD) status [15].

We are aware of only one other clinical trial evaluating triple therapy in COPD using IOS. In that study, Williamson et al. enrolled patients with severe COPD (FEV₁ 42 %) comparing 2 weeks of TIO and placebo as add-on to ICS/LABA [9]. IOS outcomes including R5, R20, and X5 were not significantly different after 2 weeks of treatment despite a significant improvement in FEV₁. This disconnect between improvements in FEV₁ but not R5 in response to TIO after 2 weeks is consistent with our current findings. Moreover, the mean R5 at baseline on ICS/LABA prior to TIO reported by Williamson et al. was 0.7 kPa L⁻¹ s which is similar to baseline R5 values in our patients. Perhaps performing a further evaluation using a full 24-h profile with serial IOS as an area under the curve (AUC) might be able to detect more subtle differences in airway caliber which we did not observe on a single trough measurement.

Whole-body plethysmography is alternative effort-independent test using a panting maneuver which can measure airway resistance (or its reciprocal as conductance) but is more difficult and time consuming to perform than IOS. Singh et al. reported on a comparison of fluticasone/salmeterol versus fluticasone/salmeterol/TIO for 2 weeks in patients with COPD, where the primary outcome of peak specific airway conductance (as AUC 0–4 h) showed a 27 % significant difference in favor of triple therapy, while for trough FEV₁, there was a significant mean difference of 110 ml [3]. In our study, we showed significant within treatment improvements in mean trough FEV₁ amounting to 110 ml with ACL and 150 ml with TIO, as compared to the minimal important difference of 100–140 ml [16] .

For visit-based spirometry outcomes including FEV₁, FVC, and RVC we observed significant within treatment effects compared to baseline in response to TIO and ACL, but no difference between treatments. Moreover, there were no significant differences between treatments for domiciliary diurnal profiles of FEV₁ and FEV₆. In a previous comparison of the same doses of TIO and ACL used as monotherapy, there were no differences between drugs in terms of trough pre-dose FEV₁ measurements after 2 weeks, although the AUC 12–24 h for the visit-based diurnal FEV₁ profile was significantly better with ACL compared to TIO after chronic dosing [11]. This observed difference in the FEV₁ AUC 12–24 h corresponds to the 12-h period after taking the evening dose of ACL when the effect of TIO is beginning to wane throughout the night time prior to the next morning dose. Perhaps on reflection, we might have also shown such differences in domiciliary FEV₁ if we had performed diurnal evening measurements between the 14- and 24-h time point for TIO, although we considered that this was not practical for elderly patients to perform at home. Another possibility is that in the present study, there may be less room for further improvement conferred by LAMA when given as add-on to ICS/LABA compared to its use as monotherapy. In a previous comparison of TIO and GLYC as add-on to ICS/LABA, the mean difference in the primary outcome of trough FEV₁ amounted to 7 ml [10], as compared to a 40 ml difference between TIO and ACL in the present study.

We used dry powder formulations of TIO and ACL which emit coarse particles >2 µm, which could also explain the somewhat limited improvements in pulmonary function. In this regard, we are not aware of any head to head lung deposition studies comparing TIO and ACL dry powder formulations in COPD. Perhaps using smaller particle formulations such as the fine-mist TIO Respimat inhaler might result in greater improvements in regional lung deposition.

With the development of single inhaler triple therapy, it will be important to know what the impact is on more clinically relevant outcomes. In a retrospective cohort study of 2853 patients with moderate to severe COPD followed up over 4.65 years, there were 996 receiving ICS/LABA (FEV₁ of 63 %) and 1857 receiving ICS/LABA/LAMA (FEV₁ of 51 %) [17]. Comparing triple versus dual therapy, there was a 15 % reduction in hospital admissions, 29 % fewer oral corticosteroid bursts, and 26 % lower all-cause mortality, while triple therapy was associated with a fall in serial FEV₁ of 30 ml over 4 years.

Aside from pulmonary function outcomes, it is also important to consider the impact of treatment upon functional status (6-min walk distance and dyspnoea index) and health-related quality of life (SGRQ) [16]. While we observed no difference between treatments in 6-min walking distance, there was a 36 m mean improvement from baseline with ACL, which exceeds the minimal important difference of 25 m [18, 19]. There was a significant improvement post-walk fatigue with ACL and in post-walk dyspnoea with TIO, but no differences were seen in either outcome between treatments. Furthermore post-walk heart rate and oxygen saturation did not change with either treatment. In terms of quality of life, although the SGRQ symptoms score was significantly improved with both drugs, there was no difference between them. The transition dyspnoea index was significantly improved by ACL versus TIO by a mean difference of 1.54 which exceeded the minimal important difference of 1.0 [20]. The relevance of this finding in the absence of any differences in objective pulmonary function outcomes is uncertain.

In summary, we observed comparable bronchodilator efficacy at trough with TIO and ACL when used as triple therapy in COPD, while IOS was no more sensitive than spirometry. The utility of IOS measurements in COPD requires further evaluation perhaps using a diurnal IOS profile to compare with spirometry or whole-body plethysmography.