A systematic review of comparative studies of tiotropium Respimat® and tiotropium HandiHaler® in patients with chronic obstructive pulmonary disease: does inhaler choice matter?

Background In many countries worldwide, the long-acting anticholinergic drug tiotropium is available as a dry powder formulation delivered by means of the HandiHaler® inhalation device and as an aqueous solution delivered via the Respimat® Soft Mist™ Inhaler. Tiotropium HandiHaler® is a single-dose, dry powder, breath-actuated inhaler that provides delivered doses and lung deposition of tiotropium that are, over a wide range, not influenced by the severity of chronic obstructive pulmonary disease (COPD). Tiotropium Respimat® is a propellant-free, multi-dose inhaler that delivers a metered dose of medication as a fine, slow-moving, long-lasting soft mist, independently of patient inspiratory effort. The high fine-particle fraction of droplets produced by the Respimat® inhaler optimizes the efficiency of drug delivery to the lungs. Methods To help inform the choice of tiotropium inhaler for prescribers and patients, this systematic review summarizes the available pharmacokinetic, efficacy and safety data from comparative studies of tiotropium Respimat® and tiotropium HandiHaler® in COPD, focusing on the licensed once-daily doses of 5 and 18 μg, respectively. Data sources reviewed include publications and abstracts identified from database searches. Results Published evidence from comparative studies suggests that tiotropium Respimat® 5 μg and tiotropium HandiHaler® 18 μg provide similar clinical outcomes in patients with COPD. Conclusions The findings indicate that physicians can base their decision about an inhaler for tiotropium on factors other than efficacy or safety. These could be patient preference for a particular inhaler, ease of use and the efficiency of drug delivery, with the aim of optimizing adherence and clinical outcomes with long-term tiotropium maintenance therapy. Video (MP4 314368 kb) Electronic supplementary material The online version of this article (doi:10.1186/s12890-016-0291-4) contains supplementary material, which is available to authorized users.


Background
Current guidelines for the management of chronic obstructive pulmonary disease (COPD) recommend maintenance treatment with inhaled bronchodilator therapy and a variety of inhaler devices are currently available, with different technical properties, levels of drug deposition within the lungs and modes of operation. The choice of inhalation device is important because it can influence patients' adherence to therapy, which can potentially affect long-term outcomes in a chronic disease such as COPD [1].
The long-acting anticholinergic drug tiotropium is available in many countries as a dry powder formulation delivered by means of the HandiHaler® inhaler device (Boehringer Ingelheim International GmbH, Ingelheim, Germany) [2,3] and as an aqueous solution delivered via the Respimat® Soft Mist™ Inhaler (Boehringer Ingelheim International GmbH) [4,5]. The efficacy and safety profile of tiotropium HandiHaler® in patients with COPD is well established, based on numerous clinical studies and also extensive post-marketing experience since its approval in Europe in 2002 and in the United States in 2004 [2][3][4][5]. Tiotropium Respimat® was approved as a COPD maintenance bronchodilator in 2007 in Europe and in 2014 in the United States and Canada [4][5][6][7][8].
Tiotropium HandiHaler® is a single-dose, dry powder, breath-actuated inhaler that provides consistent rates of delivered doses and lung deposition of tiotropium that are, over a wide range, not influenced by the severity of COPD [9]. In-vitro data have shown that the delivered dose of tiotropium was consistent at flow rates ranging from 20 to 60 L/min [10], and fine particle dose and fine particle fraction (defined as the mass fraction of particles with an aerodynamic diameter <5.8 μm) [11] were consistent at flow rates between 28.3 and 60 L/min, with a decline in fine particle dose of approximately 20% observed when flow rates decreased from 28.3 to 20 L/min [10]. In-vivo study data confirmed that COPD patients across a wide range of severities were able to generate sufficient inspiratory flow rates to activate the tiotropium HandiHaler® [10]. These findings indicate that the large majority of patients, irrespective of stage of COPD, can achieve acceptable delivery of medication through the tiotropium HandiHaler®.
Tiotropium Respimat® is a propellant-free multidose inhaler that uses mechanical power from a spring to deliver a metered dose of medication as a fine, slow-moving, long-lasting soft mist [11]. The inhaler was developed as an active system with a constant energy source, and the quality of dose and particle size distribution is uniquely independent of the patient's inspiratory flow rate [12][13][14][15]. The tiotropium Respimat® inhaler aerosolizes the majority of each metered dose in the form of droplets of >1 μm (to avoid loss of small droplets during subsequent exhalation) and <5.8 μm (to facilitate efficient lung deposition through the mechanism of sedimentation); particles that are too large (≥6 μm) deposit in the oropharynx and large conducting airways, therefore having no clinical effect [11,16,17]. The fine particle fraction (defined as the proportion of drug mass in aerosolized particles that is carried by particles with an aerodynamic diameter of not more than 5.8 μm) is 65-80% [15,18]. This high fine particle fraction, combined with the low velocity and long duration of the aerosol, results in a high level of drug deposition in the lungs and reduced oropharyngeal deposition [12,14,19]. This allows a more than 3-fold lower nominal dose of tiotropium to be administered compared with tiotropium HandiHaler® [7,12,20]; a quantitatively higher fraction of the inhaled dose is delivered to the bronchial system, with qualitatively higher distribution throughout the lung compared with other devices [21].
Both the tiotropium HandiHaler® and tiotropium Respimat® inhalers are available in many countries worldwide for the delivery of tiotropium as a maintenance treatment for COPD. Consequently, a detailed evaluation of their respective effects on clinical outcomes is warranted to help inform the choice of inhaler for prescribers and patients [7,22]. The objective of this review was to summarize and evaluate the available pharmacokinetic, efficacy and safety data that have been published and presented to date from comparative studies of tiotropium Respimat® and tiotropium HandiHaler® in patients with COPD, focusing on the once-daily licensed doses of 5 and 18 μg, respectively.

Methods
A systematic literature search was conducted for all interventional and non-interventional study publications containing the terms "tiotropium" AND "Respimat" AND "HandiHaler" AND "COPD", using the following online sources: US National Library of Medicine, National Institutes of Health PubMed database; American Thoracic Society (ATS) and European Respiratory Society (ERS) congress abstracts; British Thoracic Society (BTS) congress abstracts (via the Thorax journal website) as well as the Chest journal website. Limits were not placed upon the language of the publication. The period searched was 2006-30 September 2015 for congress proceedings and any time up to 30 September 2015 for PubMed and the Thorax and Chest journal websites. Clinical trials were also searched, using the terms "tiotropium" AND "Respimat" AND "HandiHaler" AND "COPD" at www.clinicaltrials.gov.
The total hits from the search were assessed for their relevance (based on titles/abstracts), and those publications that were deemed potentially relevant (i.e. including comparative data for tiotropium Respi-mat® and tiotropium HandiHaler® in patients with COPD) were obtained in full for analysis. The researcher then examined the Methods and Results sections of the publications to extract and summarize the data for tiotropium Respimat® 5 μg and tiotropium HandiHaler® 18 μg.
Duplicate publications, republished papers, studies not comparing tiotropium Respimat® and tiotropium HandiHaler® efficacy, safety or pharmacokinetic data (at the licensed doses), studies in non-COPD patients and secondary/review publications that did not report original data were excluded from the analysis. The trial list from www.clinicaltrials.gov was compared against the literature search results to exclude trials with data already covered by the publications.

Summary of search results
A total of 89 hits resulted from database searches (ATS abstracts = 10; ERS abstracts = 18; Thorax journal = 18 [including seven BTS abstracts and 11 other publications]; Chest journal = 13 [including nine meeting abstracts and four other publications]; PubMed = 30 publications). The number of records identified, included and excluded, and the reasons for exclusions are summarized in Fig. 1.
The total number of manuscript publications from this search that contained tiotropium Respimat® and tiotropium HandiHaler® data was 45. Of the 45 records, 35 publications were excluded for the following reasons: one reported trial design/rationale only (no results); one was a duplicate publication (included in both Thorax and PubMed search results); one was a republished paper (both publications were listed in the PubMed search results); seven studies did not include comparative efficacy, safety or pharmacokinetic data on tiotropium Respimat® and tiotropium HandiHaler® at the licensed doses; one study was not conducted in COPD patients; 24 publications did not include original data/ analyses (13 review/commentary/opinion articles, eight correspondence articles, one editorial, one health care institute report; one treatment guidelines document). An additional manuscript submitted for publication was included in the analysis, as the authors considered it to be relevant (providing further evidence to address key questions posed in this review).
The number of congress presentation (ATS, ERS, BTS, Chest) abstracts from this search that contained tiotropium Respimat® and tiotropium HandiHaler® data was 44. Of the 44 records, 33 abstracts were excluded for the following reasons: 19 did not include comparative Fig. 1 Flow diagram to show number of records identified, together with the numbers of records included and excluded, with reasons for exclusion efficacy or safety data on tiotropium Respimat® and tiotropium HandiHaler® at the licensed doses; four studies were not in COPD; nine studies with data also published in full papers (full publications were listed in search results); and 1 was a duplicate abstract (presented at both BTS and ATS). Therefore, 22 publications in total were included in this review (11 manuscripts and 11 congress presentation abstracts) (for a complete list, see Additional file 1: Table S1) . The characteristics of the clinical trials assessing tiotropium Respimat® and tiotropium HandiHaler® at the licensed doses (and reported as primary publications) and the pooled, combined and database studies covered by this review are summarized in Table 1.
The search of www.clinicaltrials.gov provided 13 records and all 13 were excluded from the systematic review for the following reasons: seven trials were completed with data published in manuscripts or abstracts already selected for the current analysis; six trials did not include comparative data on tiotropium Respimat® and tiotropium HandiHaler® at the licensed doses (five trials compared tiotropium Respimat® or tiotropium HandiHaler® with other therapies [olodaterol or indacaterol]; one observational study showed only combined results for tiotropium Respimat® together with tiotropium HandiHaler®).
Previously, it had been suggested that systemic exposure with tiotropium Respimat® might be greater than with tiotropium HandiHaler®, with associated potential for increased risk of toxicity [45,46]. However, a recent extensive study comparing the pharmacokinetic properties of tiotropium administered via the two inhalers showed that systemic exposure to tiotropium (as shown by mean plasma concentration profile at steady state) was lower in patients with COPD treated with tiotropium Respimat® 5 μg compared with patients treated with tiotropium HandiHaler® 18 μg [31]. The crossover design study included five 4-week treatment periods of placebo and once-daily doses of tiotropium Respimat® 1.25, 2.5 and 5 μg, and tiotropium HandiHaler® 18 μg. Based on the findings of earlier studies in COPD patients, which showed that pharmacokinetic steady state was achieved after 2-3 weeks of once-daily dosing with tiotropium, with no further accumulation after this time [37,47], 4 weeks was considered to be sufficient to reach pharmacokinetic and pharmacodynamic steady state. Figure 2 shows mean plasma concentrations of tiotropium from 2 min to 6 h postdosing with tiotropium HandiHaler® 18 μg and tiotropium Respimat® 5 μg.

Sleep quality study
Patients with COPD can be affected by disordered gas exchange and poor sleep quality. A study was performed to compare the effect of tiotropium Respimat® and tiotropium HandiHaler® on sleeping arterial oxygen saturation (SaO 2 )  Safety analysis in patients with renal impairment included in placebo-controlled trials of once-daily tiotropium Respimat® 5 μg (7 trials    and sleep quality in 200 patients with COPD, 6 months after the start of treatment [25,44]. At the end of treatment (n = 188), both treatment groups showed significant improvements in minimum SaO 2 (p <0.001) and percentage of sleep spent below 90% of SaO 2 (TST90) (tiotropium Respimat®, p <0.001; tiotropium HandiHaler®, p = 0.002) compared with baseline (Respimat® vs. HandiHaler® for SaO 2 and TST90 at 6 months: p = 0.83 and p = 0.04, respectively). Patients treated with tiotropium Respimat® had significantly better TST90 than did the patients treated with tiotropium HandiHaler®. Sleep disturbance was highly variable, but the durations of sleep stages (and therefore overall sleep quality) were significantly improved in the tiotropium Respimat® group compared with the tiotropium HandiHaler® group (p ≤0.01).

Pooled and combined analyses of safety
A pre-specified pooled analysis of two 30-week crossover trials reported the following findings [37]: the most common adverse events were COPD exacerbations (9.6% with tiotropium Respimat® 5 μg, 11.2% with tiotropium HandiHaler® 18 μg and 13% with placebo) and nasopharyngitis (7.5% with tiotropium Respimat® 5 μg, 5.9% with tiotropium HandiHaler® 18 μg and 8.2% with placebo) [37]. COPD exacerbation, dry mouth and nasopharyngitis were also the most common adverse events in the study in Japanese patients, and the number of adverse events reported in patients receiving tiotropium Respimat®  The safety of tiotropium delivered by Respimat® and HandiHaler® was recently reviewed in a pooled analysis of adverse event data from 28 HandiHaler® and seven Respimat® studies involving 12,929 patients treated with tiotropium and 11,626 patients treated with placebo [30]. Patients were eligible for inclusion in these studies if they had a diagnosis of COPD with FEV 1 ≤ 70% of FVC, were aged ≥40 years and had ≥10 pack-years of smoking history. Patients were excluded if they had significant disease other than COPD. Other exclusion criteria in earlier studies were heart failure leading to hospitalization in the previous 3 years, cardiac arrhythmia requiring drug treatment or myocardial infarction (MI) within the past year. More recent trials only excluded life-threatening cardiac arrhythmia or arrhythmia that needed a change in medication or heart failure resulting in hospitalization in the past year, and/ or MI within the previous 6 months. These relatively broad inclusion and exclusion criteria mean that the patient population included in the analysis reflected real-world heterogeneity of populations and phenotypes of COPD patients, as far as is possible in randomized clinical trials. The risk of adverse events (rate ratio 0.90; 95% CI 0.87-0.93) and serious adverse events (rate ratio (0.94; 95% CI 0.89-0.99) was significantly lower in the tiotropium group than in the placebo group, and the risk of fatal adverse events (rate ratio 0.90; 95% CI 0.79-1.01) and cardiac adverse events (rate ratio 0.93; 95% CI 0.85-1.02) was numerically lower in the tiotropium group. Similar results were obtained when tiotropium HandiHaler® 18 μg and tiotropium Respimat® 5 μg groups were analysed separately, and no increased risk of cardiac, vascular, and respiratory, thoracic and mediastinal disorders, or stroke, were observed in the tiotropium groups, except for a higher risk of ischaemic heart disease for tiotropium versus placebo in the tiotropium Respimat® 5 μg group (rate ratio 1.6 [95% CI 1.04-2.49]) but not in the tiotropium HandiHaler® 18 μg group. However, the incidence rates were lower in the placebo Respimat® group than in the placebo HandiHaler® group (1.25 vs. 1.89), and there was no evidence of increased risk of major adverse cardiovascular events (MACE) or fatal MACE in the tiotropium group compared with placebo, or in tiotropium HandiHaler® 18 μg and tiotropium Respimat® 5 μg groups separately. These results do not indicate an increased overall risk for fatal or cardiovascular events in COPD patients during tiotropium treatment, and support the findings from the TIOSPIR® (Tiotropium Safety and Performance in Respimat) trial-the largest randomized, double-blind, parallel-group study of patients with COPD, which did not show any relevant differences between tiotropium HandiHaler® 18 μg and tiotropium Respimat® 5 μg [48]. The findings of TIOSPIR® with regard to the safety of tiotropium HandiHaler® 18 μg and tiotropium Respimat® 5 μg are discussed in more detail later in this review.
The issue of whether inhaled anticholinergics and, in particular, tiotropium administered by Respimat®, may induce cardiac arrhythmias in a vulnerable subpopulation with cardiovascular morbidity has been discussed in the literature [39,46,49]. In this context, the results of a combined analysis of all tiotropium (HandiHaler® 18 μg and/or Respimat® 1.25-10 μg) trials in COPD involving Holter electrocardiogram (ECG) monitoring, and conducted between 2003 and 2012 [32], are important. In the four trials that were included in the analysis, patients were required to have a diagnosis of COPD with FEV 1 ≤ 70% of FVC, were aged ≥40 years and had ≥10 pack-years of smoking history. Holter ECGs were evaluated for heart rate, pauses (absence of a heart beat for more than 3 seconds), supraventricular premature beats and ventricular premature beats. Maintenance therapy with either tiotropium Respimat® 5 μg or tiotropium HandiHaler® 18 μg was not associated with changes in any of these variables ( Table 2). The authors commented that the results are in line with those of TIOSPIR®, which found no evidence that tiotropium Respimat® is associated with an increased risk of mortality, especially in patients with cardiac disease, or specifically arrhythmias at baseline. An analysis of safety in patients with renal impairment (n = 10,753 evaluable patients) included in placebocontrolled trials of once-daily tiotropium Respimat® 5 μg (seven trials) or tiotropium HandiHaler® 18 μg (15 trials) has been conducted [35]. The incidence of adverse events, serious adverse events or fatal adverse events with either tiotropium Respimat® 5 μg or tiotropium HandiHaler® 18 μg showed no association with mild to moderately impaired renal function (Fig. 4). Results for severe renal impairment were limited due to the low number of patients (n = 52).

Database analysis of mortality
A report from a Dutch primary care database (source population 11,287, including 24,522 episodes of tiotropium use) found that the use of tiotropium Respimat® was associated with an almost 30% increase of mortality compared with tiotropium HandiHaler® [38,39]. The association was strongest for cardiovascular/cerebrovascular death. These findings, however, are not supported by those of the large prospective TIOSPIR® trial, which showed no difference in mortality between patients using tiotropium Respimat® or tiotropium HandiHaler®, as described below.

The TIOSPIR® study
The TIOSPIR® study was a 2-3 year, randomized, doubleblind, parallel-group trial enrolling 17,135 patients with COPD [43,48]. The aim of the trial was to evaluate the safety and efficacy of once-daily tiotropium Respimat® 2.5 or 5 μg and tiotropium HandiHaler® 18 μg in a large COPD population. Patients were permitted to continue their usual respiratory therapy (with the exception of other inhaled anticholinergics). Patients with cardiovascular diseases were allowed to participate, except for patients with heart failure resulting in hospitalization or cardiac arrhythmia requiring new drug treatment during the previous year, or experiencing MI within the past 6 months.
The primary safety endpoint for TIOSPIR® was time to all-cause mortality and the primary efficacy endpoint was time to first COPD exacerbation; secondary outcome measures included the number of exacerbations and time to the first MACE. For the primary endpoint of all-cause mortality, tiotropium Respimat® 5 μg was non-inferior to tiotropium HandiHaler® 18 μg (hazard ratio 0.96; 95% CI 0.84−1.09). Analysis of causes of death as assigned by TIOSPIR® investigators compared with those assigned by a mortality adjudication committee (MAC) found that fewer deaths were assigned by the MAC to cardiac disorders in the tiotropium HandiHaler® 18 μg group than in the tiotropium Respimat® 5 μg group, although this was not a significant effect (the CI of the rate ratio was overlapping 1) [42].
The spirometry sub-study of TIOSPIR® (n = 1370) found that tiotropium Respimat® 5 μg was noninferior to tiotropium HandiHaler® 18 μg for adjusted mean trough FEV 1 (averaged over 24-120 weeks: difference vs. HandiHaler® −10 mL; 95% CI −38−18) [23,24,48]. Adjusted mean trough FVC was also similar between treatment groups. BPM beats per minute, ECG electrocardiogram, FAS full analysis set, N number of patients with non-missing data; n, number of patients with event, SD standard deviation, SVPB supraventricular premature beat, VPB ventricular premature beat. A pause was defined as absence of a heart beat for >3 s Fig. 4 Forest plot of incidence rate ratios (95% CI) of on-treatment AEs by renal function at baseline: post-hoc analysis of tiotropium trials [35]. Renal function classification using National Institute for Health and Clinical Excellence (NICE) criteria: normal ≥90 mL/min, mild ≥60 to <90 mL/min, moderate ≥30 to <60 mL/min, severe <30 mL/min creatinine. Incidence rate ratios could not be calculated for severe renal impairment due to low patient numbers. Where there were no events in the placebo or tiotropium group, incidence rate ratios could not be calculated (division by zero) or are equal to zero, respectively, and are not graphically displayed. AE, adverse event; CI, confidence interval; FAE, fatal adverse event; IRR, incidence rate ratio; SAE, serious adverse event; SOC, System Organ Class (Medical Dictionary for Regulatory Activities) Post-hoc subgroup analyses including data from the TIOSPIR® study have further supported the clinical equivalence of tiotropium Respimat® 5 μg and tiotropium HandiHaler® 18 μg. Analyses of the 4-year placebo-controlled Understanding Potential Long-term Impacts on Function with Tiotropium (UPLIFT®) trial of tiotropium HandiHaler® and TIOSPIR® found that in patients who had experienced a cardiac event (for which they would have been excluded at baseline) during the trials, the risk of serious (including fatal) cardiac or MACE was not increased by tiotropium (tiotropium HandiHaler® 18 μg or tiotropium Respimat® 5 μg) [36]. Similar findings were obtained in a separate analysis of patients experiencing cardiac events during TIOSPIR® (Fig. 6) [41].
An analysis of data from patients from TIOSPIR® who were naïve to anticholinergic treatment at baseline (n = 6966) found that, as in the primary analysis [48], these patients had similar safety and exacerbation efficacy profiles when treated with tiotropium Respimat® 5 μg or tiotropium HandiHaler® 18 μg [40]. The HR was 0.93 (95% CI 0.75−1.17) for risk of death (measured as time to death) and 0.99 (95% CI 0.90−1.08) for exacerbations (measured as time to first exacerbation).

Discussion
This systematic review evaluated pharmacokinetic, efficacy and safety results from published studies of tiotropium Respimat® and tiotropium HandiHaler® at the licensed doses (5 and 18 μg), respectively, with the aim of summarizing evidence that might inform the choice of tiotropium inhaler in clinical practice.
The results of several randomized dose-finding studies and the TIOSPIR® study have demonstrated that tiotropium Respimat® 5 μg has a pharmacokinetic, efficacy and safety profile that is comparable with that of tiotropium HandiHaler® 18 μg [24,26,31,33,37,48]. Results from post-hoc and pooled analyses provide further confirmation that overall lung function, exacerbation, quality of life and safety outcomes are equivalent for the two tiotropium inhalers across a range of patient subtypes. The efficacy and safety of tiotropium, when administered by either Respi-mat® or HandiHaler®, is supported by previously published systematic reviews [7,22,49,50].
There has been debate about the safety of inhaled anticholinergics and, in particular, the cardiac safety of tiotropium administered by Respimat®, which was triggered by meta-analysis [46] and database analysis [39] that reported an increase in mortality in patients treated with tiotropium Respimat®. The meta-analysis examined data from five randomized, controlled trials of tiotropium Respimat® [46] and found an increased risk of mortality compared with placebo. However, there was no direct  [41]. Events were counted from the day following the initial cardiac event through drug stop +30 days. FAEs, fatal adverse events; MACE, major adverse cardiovascular events; NS, not significant; SAEs, serious adverse events; TIOSPIR®, Tiotropium Safety and Performance in Respimat® comparison with tiotropium HandiHaler®, and the analysis was limited by differences in the populations studied, the tiotropium dose used, and length of follow-up. The investigators also noted that low event rates precluded precise estimates of risk [46]. The authors of the Dutch database analysis commented that it was unclear whether the apparent association between the use of tiotropium Respimat® and an increased risk of death was causal or due to residual confounding by COPD severity [39]. One source of confounding could be the substantial differences in the population treated with tiotropium HandiHaler® and tiotropium Respimat®. Although the analysis was adjusted for several factors, the adjustment was incomplete, and a channelling effect towards more severe patients being treated with tiotropium Respimat® was described by the same group [51].
The tiotropium safety data from the meta-and database analyses are in contrast to the results of TIOSPIR® [48], in-cluding~17,000 patients, which provided the most robust data available to date regarding the comparative safety and efficacy of the two tiotropium formulations, and particularly the licensed doses. Key findings of TIOSPIR® were that tiotropium Respimat® 5 μg was non-inferior to tiotropium HandiHaler® 18 μg in terms of all-cause mortality and that the risk of cardiovascular mortality or MACE did not differ significantly between the two treatment groups [41,43]. In addition, there was no increased risk of subsequent cardiac events with tiotropium Handi-Haler® 18 μg or tiotropium Respimat® 5 μg in patients experiencing a serious cardiac event during the trial (the risk of MACE was actually lower with tiotropium HandiHaler® 18 μg than with placebo in UPLIFT®) [36]. This is an important finding, as many patients with COPD in clinical practice are likely to have underlying cardiac disease, yet such patients are typically excluded from randomized clinical trials of maintenance COPD treatments. The TIOS-PIR® trial did not exclude most patients in routine care with cardiac diseases including stable coronary artery disease or stable arrhythmias, making it a study that was inclusive of the majority of patients that are typically seen in clinical practice. It is acknowledged that the data from TIOSPIR® are more robust than those arising from the meta-analyses or database studies [7,52], which had previously raised concerns about an increased mortality risk with the tiotropium Respimat® inhaler [38,39,46].
Limitations of this review are the descriptive presentation of the findings (not subject to statistical analysis) and the inclusion of secondary and post-hoc analyses (such as those conducted on sub-populations of patients in the TIOSPIR® trial). Generally, it is challenging to draw firm conclusions from the results obtained across numerous trials, owing to differences in study duration and design. However, the studies assessed here included COPD patients across a broad range of disease severity (from moderate to very severe), and the TIOSPIR® trial allowed patients to receive tiotropium HandiHaler® or tiotropium Respimat® while continuing with their usual COPD maintenance therapy (thus helping to reflect clinical practice) [43,48]. Overall, the review encompasses a large body of data on tiotropium HandiHaler® or tiotropium Respimat® from randomized trials, pooled analyses and database studies.
The studies reviewed here suggest that clinical efficacy appears equivalent between tiotropium Respimat® 5 μg and tiotropium HandiHaler® 18 μg, and as such, patient preferences and acceptance of different inhaler types become more important in the prescribing decision [53,54]. Patient preference for an inhaler is an important determinant of treatment adherence, which is a key consideration for treatment choices in chronic diseases [1,55,56]. In addition, mishandling of inhalers is a common issue that may result in reduced symptom control [57], and therefore for any individual patient, it is important to assess ability to use the different types of available device. The inhalation and handling characteristics of tiotropium Respimat® have been assessed by patients with COPD, and was preferred to alternative inhalers, including metereddose inhalers and dry powder inhalers [56,[58][59][60].
For physicians who may be considering whether patients are suitable candidates to switch from tiotropium Handi-Haler® to tiotropium Respimat®, TIOSPIR® data show that in patients who switched from tiotropium HandiHaler® to tiotropium Respimat®, mortality, cardiac safety and exacerbation outcomes were similar to those who remained on tiotropium HandiHaler® [27,29]. From the patient's perspective, studies have suggested that they find it easy to switch from tiotropium HandiHaler® to tiotropium Respi-mat®, and have reported high levels of preference for, and adherence to, tiotropium Respimat® [61][62][63][64][65].
It also appears from "real-world" experience that physicians are already confident to prescribe tiotropium Respimat® for their patients with more severe disease and/or comorbidities. A study of the Dutch Integrated Primary Care Information Database was performed to compare patient characteristics at the time of the first prescription of tiotropium Respimat® or tiotropium HandiHaler® (source population 501,474, including 11,753 tiotropium users) [51]. COPD was found to be more severe and underlying comorbidities were more prevalent for first-time users of tiotropium Respimat® compared with tiotropium HandiHaler®. Also, in Italy, a drug utilization study conducted in patients receiving tiotropium during 2011-2012 found that users of tiotropium Respimat® and tiotropium HandiHaler® (n = 4390) had similar characteristics [66], but in this study, the probability of switching to tiotropium Respimat® was greater in patients with severe respiratory disease. If tiotropium Respimat® is being selected for patients with more severe COPD in clinical practice, this might help to explain the increased mortality risk suggested by earlier database and meta-analyses. However, TIOSPIR® demonstrated that tiotropium Respimat® has a similar safety and exacerbation efficacy profile to tiotropium HandiHaler® in patients with moderate-to-very severe COPD [48], supporting its use across the disease spectrum.

Conclusions
The approval of tiotropium Respimat® in many countries has provided physicians with a choice of inhaler for the delivery of tiotropium maintenance therapy for their patients with COPD. Published evidence from comparative studies suggests that tiotropium Respimat® 5 μg and tiotropium HandiHaler® 18 μg provide similar clinical outcomes in COPD, indicating that physicians can choose between the two inhalers with confidence. Factors other than efficacy and safety, such as patient preference for a particular inhaler, ease of use and handling, and the efficiency of drug delivery (which has improved significantly for tiotropium with the Respimat® device) [20] should also be taken into account with the aim of optimizing adherence and clinical outcomes with long-term tiotropium maintenance therapy.

Additional file
Additional file 1: Table S1.