Chronic obstructive pulmonary disease (COPD) is a common, preventable and treatable disease that is characterized by persistent respiratory symptoms and airflow limitation due to airway and/or alveolar abnormalities (GOLD., 2021). Significant exposure to noxious particles or gases and host factors including abnormal lung development usually contribute to the pathogenesis (GOLD., 2021). Based on Burden of Obstructive Lung Disease (BOLD) and other large scale epidemiological studies, a meta-analysis estimated that the number of COPD cases was 384 million in 2010, with a global prevalence of 11.7% (95% confidence interval (CI): 8.4–15.0%) (Adeloye et al., 2015). Around 3.2 million people died from COPD each year, making it the third leading cause of death worldwide (World Health Organization, 2007; Burney et al., 2015; Global Burden of Disease Study Collaborators, 2015; Halpin et al., 2019). In the latest Global Burden of Disease (GBD) analysis, COPD entered the top 10 causes of years of life lost (YLL), increasing from the 11th position in 2007 to seventh in 2017 (GBD 2017 Causes of Death Collaborators, 2018). Another GBD study also predicted that deaths from COPD would rise to 4.4 million per year in 2040 and by then, COPD would be the fourth most important cause of YLL (Foreman et al., 2018). With the increasing exposure to risk factors (e.g., smoking) and aging of the world’s population, the prevalence of COPD is expected to rise over the next 40 years and by 2060 there may be more than 5.4 million deaths from COPD and its related conditions annually (Lopez et al., 2006; GBD 2017 Causes of Death Collaborators, 2018; World Health Organization, 2020), which will induce a substantial and elevated economic burden (Lozano et al., 2012; Vos et al., 2012). In the European Union, COPD accounted for 56% (38.6 billion Euros) of the cost on respiratory disease which took up about 6% of the total annual healthcare budget (European Respiratory Society on behalf of the Forum of International Respiratory Societies (FIRS), 2017). In the United States, the estimated direct and indirect costs of COPD were $32 billion and $20.4 billion, respectively (Guarascio et al., 2013).

In absence of conclusive evidence supporting any existing medications which can modify the long-term decline in lung function for COPD (Anthonisen et al., 1994; Burge et al., 2000; Pauwels et al., 1999; Tashkin et al., 2008; Vestbo et al., 1999), the purpose of pharmacological therapy for COPD is to ameliorate symptoms, reduce the frequency and severity of exacerbations, and improve exercise tolerance and health status. As the first-line therapy to address COPD symptoms and prevent exacerbations (GOLD., 2021), long-acting muscarinic antagonists (LAMAs) can improve the effectiveness of pulmonary rehabilitation (Casaburi et al., 2005; Kesten et al., 2008) and reduce exacerbation and related hospitalization (Karner et al., 2014; Melani A.S., 2015) by durably blocking the bronchoconstrictor effects of acetylcholine on M₃ muscarinic receptors expressed in airway smooth muscle (Melani A.S., 2015). Revefenacin (REV), a novel once-daily LAMA for nebulization, was approved for the treatment of COPD by the United States Food and Drug Administration (FDA) in November 2018 (Highlights Of Prescribing Information, 2021). Several randomized trials (Donohue et al., 2019a; Donohue et al., 2019b; Donohue et al., 2019c; Ferguson et al., 2019; Krishna et al., 2017; Mahler et al., 2019; Quinn et al., 2018; Sethi et al., 2020; Siler et al., 2020; Theravance Biopharma, 2021a; Theravance Biopharma, 2021b) investigating the use of REV concluded that it was effective and safe in the treatment of COPD. Nevertheless, evidence has not been systematically assessed. To better understand and interpret available evidence, we conducted a systematic review incorporating a dose-response meta-analysis to evaluate the efficacy and safety of REV in patients with COPD.

We reported our study following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Supplementary Table S1). The study was prospectively registered on International Prospective Register of Systematic Review (PROSPERO, CRD42020182793).

This systematic review summarized the evidence of efficacy and safety of REV in patients with moderate to very severe COPD and used a novel meta-analysis method to account for the dose-response relationship of the trough FEV₁, AEs, and SAEs with REV dose. Low-quality evidence suggests that, compared to placebo, 175 μg/day REV might improve the lung function (increment of trough FEV₁ on average of 143.67 ml higher than placebo) without elevating the risk of AEs or SAEs. However, only very low-quality to low-quality evidence demonstrates that the safety profile of REV at a dose of 175 μg/day is similar to TIO and IPR but its long-term efficacy is inferior to TIO (decrease of trough FEV₁ on average of 39.2 ml lower than TIO). The effect of REV in increasing trough FEV₁ was correlated to the dose with a threshold value of 88 μg/day. Notably, the efficacy of REV would be weakened with the extension of therapeutical course.

Despite the serious risk of bias and inconsistence, the confidence rating of evidence regarding the efficacy of REV vs. placebo might be enhanced by the dose-response gradient which was consistent with the results in vitro (Pulido-Rios et al., 2013). The novel robust error meta-regression method had some merits of reducing the probability of type I error caused by repeated analyses, so it was utilized to investigate the dose-response relationship. This relationship was non-linear and included three phases based on REV dose: 0–88 μg/day, 88–175 μg/day, and 175–700 μg/day. The change from baseline in trough FEV₁ dramatically escalated with the increasing dose of REV from 0 to 88 μg/day. Thereafter, the growth rate started to slow down and achieved a plateau phase when the dose exceeded 175 μg/day due to a ceiling effect. Our finding is coincided with current suggestion where 88 and 175 ug/day REV are considered as appropriate doses for investigating longer-term safety and efficacy of REV (Krishna et al., 2017). To explore the heterogeneity of trough FEV₁ regarding 88 μg/day of REV vs. placebo at 12-weeks, we compared the baseline of participants in trial NCT02512510 with that in trial NCT02459080. Unfortunately, there was no significant difference in baseline characteristics. Hence, one possible reason for explaining the heterogeneity is that a dose of 88 μg/day was the threshold of the dose-response curve and some patients in the study might not receive the full benefits of the treatment, suggesting that a higher dose would be more optimal for all participants. Different from efficacy, there was no significant dose-response relationship between dose and the incidence of AEs or SAEs and the safety profile of REV was comparable to placebo. In addition, a previous study also reported that concurrent long-acting β-agonists (LABA) would slightly raise the incidence of AEs for patients receiving REV at a dose of 88 μg/day rather than those receiving REV at a dose of 175 μg/day (Donohue et al., 2019c). Thereby 175 μg/day has been approved as a standard dose by the United States FDA (Highlights Of Prescribing Information, 2021).

The effect of REV at a dose of 175 μg/day in improving the trough FEV₁ was superior to TIO within 26 weeks but then got inferior after 39 weeks. On one hand, the disproportionate number of poor performers who discontinued TIO during the final 3 months of treatment (Donohue et al., 2019b) could partially account for this phenomenon. On the other hand, the distinct mechanism of drug action should also be considered, as REV exhibits pharmacological effects through selective inhibition of M₃ receptor at the smooth muscle leading to bronchodilation, while TIO blocks both M₃ and M₁ receptors to take more prolonged effects (Li and Yang, 2019). Given that REV with novel biphenyl carbamate tertiary amine structure is different from TIO with quaternary ammonium feature (Donohue et al., 2019d; Montuschi and Ciabattoni, 2015), REV was supposed to have higher metabolic lability and more rapid systemic clearance than TIO (Babu and Morjaria, 2017; GlaxoSmithKline. Incruse, 2021) in terms of minimizing systemically mediated AEs. Nonetheless, this systematic review did not show any significant advantage of REV in reducing the risk of AEs or SAEs compared to TIO or IPR, which might be ascribed to underpowered sample size. Although present evidence showed that REV was not preferable to TIO both in efficacy and safety, certain COPD patients with chronic muscle weakness, or cognitive or visual impairment or diminished manual dexterity may still particularly benefit from the use of this once-daily nebulized delivery LAMA (Bonini and Usmani, 2015; Tashkin D. P., 2016). As the evidence about the efficacy and safety of REV vs. TIO was mainly from two trials (NCT02518139 and NCT03095456) with high risk of attribution, reporting, and other bias, its confidence rating was graded as very low to low quality.

This systematic review also found the therapeutical course would influence the efficacy in improving trough FEV₁, which could be explained by the progression of COPD with longer follow-up time. Considering the limited data from trials, we did not evaluate the association of reduced efficacy with treatment course. Furthermore, the proportion of ICS/LABA users varied a lot among all the included trials, which probably brought heterogeneity to the results of meta-analyses. Nevertheless, a subgroup analysis (Sethi et al., 2020) found that REV produced similar improvements from baseline in trough FEV₁ in the non-LABA and LABA groups despite more AEs reported in the LABA.

There are several limitations in this study. As we only included RCTs, the results may not have good generalizability for strict inclusion criteria and small sample size. Particularly, the representativeness of participants was compromised because all the trials were conducted in the United States, the United Kingdom, Northern Ireland, New Zealand, and South Africa where most of the participants were white. In addition, these trials were not sensitive to assess treatment-related rare AEs (incidence ≤0.01%) due to relatively lower power of test and shorter follow-up term. Furthermore, the quality of evidence was subpar for the high risk of attribution and reporting bias in primary studies. Moreover, the language restriction for English and Chinese could also reduce the generalizability of our results. Therefore, prospective, multicenter, RCTs with larger samples, different populations, and better methodological design are urgently needed in this field. Although the course of treatment would influence the efficacy of REV, we performed the dose-response meta-analysis without adjusting this confounder due to limited data from the trials, suggesting that the non-linearity relationship between dosage and improvements in the through FEV₁ of REV should be interpreted with caution. Finally, even though study design and concomitant medication such as formoterol in NCT03573817 would also be the possible source of heterogeneity, we did not assess the effect of these factors on the results due to small quantity of trials with the same outcomes.

To conclude, based on the findings of our systematic review and dose-response meta-analysis of RCTs, REV appears to be a promising option for the treatment of moderate to very severe COPD. Considering the low confidence rating of evidence, further studies are warranted to compare the efficacy, long-term safety and cost-effectiveness between REV and other LAMAs (TIO) in different populations. Although most studies used the FEV₁ to evaluate the efficacy of REV in treatment of COPD, but FEV₁ should just be set as a surrogate outcome. Therefore, the clinical benefit of REV in patients with COPD should be further evaluated. And researchers should increase focus on those important endpoints (e.g., death, exacerbations requiring antibiotics or oral steroids, hospitalizations due to exacerbation of COPD, exacerbations requiring a short course of an oral steroid or antibiotic, etc.) and patient-reported outcomes in the further research due to few trials reporting such related endpoints.