Abstract
Background The efficacy and safety of gefapixant in adults with chronic cough remain unclear. Our objective was to assess the efficacy and safety of gefapixant using updated evidence.
Methods MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL) and Embase databases were searched from inception through September 2022. Subgroup analysis based on dose of gefapixant (i.e. ≤20, 45–50 and ≥100 mg twice daily for low, moderate and high doses, respectively) was performed to explore a potential dose-dependent effect.
Results Five studies involving seven trials showed the efficacy of moderate- or high-dose gefapixant for reducing objective 24-h cough frequency (estimated relative reduction 30.9% and 58.5%, respectively) (i.e. primary outcome) and awake cough frequency (estimated relative reduction 47.3% and 62.8%, respectively). Night-time cough frequency was only reduced with high-dose gefapixant. Consistently, the use of moderate- or high-dose gefapixant significantly alleviated cough severity and improved cough-related quality of life, but increased the risk of all-cause adverse events (AEs), treatment-related AEs and ageusia/dysgeusia/hypogeusia. Subgroup analysis showed dose dependency in both efficacy and AEs with a cut-off dose being ≥45 mg twice daily.
Conclusions This meta-analysis revealed dose-dependent efficacy and adverse effects of gefapixant against chronic cough. Further studies are required to investigate the feasibility of moderate-dose (i.e. 45–50 mg twice daily) gefapixant in clinical practice.
Abstract
Gefapixant improved 24-h awake cough frequency, severity and quality of life. Gefapixant increased the risk of all-cause and treatment-related adverse events. Gefapixant showed dose-dependent efficacy with a cut-off dose ≥45 mg twice daily. https://bit.ly/3ZtTUtN
Introduction
Chronic cough, clinically defined as coughing that lasts for >8 weeks [1, 2], has an estimated overall global prevalence of ∼6.22–10%, with higher rates seen in Europe, America and Oceania (11–18%) [3, 4]. On the other hand, the prevalence of refractory chronic cough, which refers to a condition in which cough persists despite extensive guideline-based evaluation and treatments, was reportedly 20–46% among patients visiting specialist clinics [5, 6]. Although the causes can be identified in some of the patients, the aetiology remains unclear in many other cases (i.e. unexplained chronic cough) [2, 5, 7]. Not only did chronic cough pose a substantial impact on quality of life [6, 8, 9], but it was also associated with an increased risk of anxiety and depression [10–12].
P2X3 receptors are ATP-gated ion channels expressed in afferent neurons including vagal C- and Aδ-fibres involved in cough sensitisation [13, 14]. Recent evidence has shown that gefapixant (previously known as MK-7264 and AF-219), which is a selective P2X3 and P2X2/3 receptor antagonist, was effective for the treatment of chronic cough [14]. An earlier proof-of-concept trial demonstrated a significant reduction in objective cough frequency in patients treated with 600 mg gefapixant twice daily compared with placebo [15], while subsequent trials found that its efficacy could be retained at a dose as low as 50 mg twice daily [16]. Nevertheless, data were mixed regarding the statistical significance of gefapixant treatment efficacy when the dose was further reduced or when the outcomes were assessed with subjective tools [2, 17–19]. Although a previous systematic review attempted to evaluate the efficacy and safety of gefapixant for the treatment of chronic cough, it included only phase 2 studies and was limited by the small number of randomised controlled trials (RCTs) (i.e. four RCTs) [20]. Although the overall estimate of that study comprised a wide range of doses (i.e. from 7.5 to 600 mg twice daily), the inclusion of only a small number of trials in each dose-stratified subgroup limited the clinical implication and application of its findings [20].
Therefore, we conducted this systematic review and meta-analysis aiming at pooling up-to-date evidence from published RCTs to evaluate the efficacy and safety of gefapixant for the treatment of chronic cough at different doses evaluated with both subjective and objective outcome measures.
Methods
We registered the protocol of the current meta-analysis with the International Prospective Register of Systematic Reviews (PROSPERO; CRD42022358694). This research adhered to the Preferred Reporting Items Systematic Reviews and Meta-Analysis (PRISMA) guidelines (www.prisma-statement.org).
Data sources and search strategies
Eligible articles were searched from inception to 8 September 2022 in three electronic databases without language restriction: MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL) and Embase. Only RCTs that investigated the efficacy of gefapixant against chronic cough were included. The search terms were: (“AF-219” OR “gefapixant” OR “MK-7264”) and (cough OR cough (MeSH or Emtree)). In addition, manual searches were also conducted to identify potentially eligible articles from the reference lists of the retrieved articles. The syntax search strategies are available in supplementary table S1.
Inclusion and exclusion criteria
The following screening criteria were used to identify eligible RCTs: 1) population: adult participants (age ≥18 years) with refractory/unexplained cough regardless of its severity; 2) intervention: administration of gefapixant as an antitussive agent irrespective of its dosage or treatment duration; 3) comparison: use of placebo or no treatment as a control group; and 4) outcomes: frequency and severity of cough, cough-related quality of life and treatment-associated adverse events (AEs).
Exclusion criteria included: 1) paediatric population; 2) studies without a control group; 3) those adopting other antitussive agents as a control group; 4) those without details about outcomes of interest; 5) those reporting results from post-hoc analysis of previous RCTs; 6) studies focusing on animal experiments; 7) those published as abstracts only or adopting research designs other than RCTs (e.g. reviews and case reports); 8) those published without peer review; and 9) those focusing on patients with acute cough or chronic cough from specific diseases (e.g. obstructive or interstitial lung diseases).
Study selection and data extraction
The records obtained from the database search were surveyed as follows. To determine the eligibility of the retrieved studies, two reviewers (M-H.C. and I-W.C.) independently screened the titles and abstracts. The studies were subsequently read in full text independently by the same two reviewers to check for eligibility. Relevant data were extracted from the eligible RCTs: study details (e.g. country, author, publication year and study design), participants (e.g. age, gender and sample size), pharmacological therapies (e.g. dosage and treatment duration), efficacy (e.g. cough frequency and severity) and rate of AEs. Should missing information be needed, we e-mailed the corresponding authors in an attempt to acquire the data of interest.
Definitions
Primary outcome
The primary outcome was 24-h cough frequency defined as number of coughs per hour.
Secondary outcomes
Secondary outcomes included: awake cough frequency (coughs per hour); night-time cough frequency (coughs per hour); cough severity based on a 100-mm visual analogue scale (VAS) (scores 0 and 100 represent no symptom and the worst symptoms, respectively); a cough severity diary (CSD) that comprises seven items in total, each of which is assigned a score from 0 (best) to 10 (worst), giving a total daily CSD score between 0 to 70 (the mean score is acquired by dividing the sum of the seven item scores by 7); cough-related quality of life by using validated scores (e.g. Leicester Cough Questionnaire (LCQ) or Cough Quality of Life Questionnaire (CQLQ)); and safety outcomes including risk of all-cause AEs, study discontinuation due to AEs, serious AEs, AEs related to treatment, dysgeusia, hypogeusia, ageusia, oral paraesthesia or hypoesthesia and renal/urological AEs.
Risk of bias and certainty of evidence assessment
Using the Cochrane Collaboration's tool for assessing the risk of bias (RoB version 1.0), two authors independently assessed possible biases in selection, performance, detection, attrition and reporting, as well as other bias among the included RCTs [21]. Each domain was graded as low, unclear or high risk of bias after resolving disagreements between the two authors (M-H.C. and I-W.C.) by discussion with the corresponding author (K.C.-H.). For the risk of attrition bias, we regarded the extent of missing data <5% as “low”, 5–10% as “unclear” and >10% as “high”. Regarding the potential risk of bias arising from pharmaceutical sponsorship, we considered the risk of bias to be “unclear”.
Using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology, two independent authors investigated the certainty of evidence as well as summary of findings on the primary (i.e. 24-h cough frequency) and secondary outcomes (i.e. awake/night-time cough frequency, cough severity, CSD, cough-related quality of life, all-cause AEs, treatment-related AEs, serious AEs and discontinuation rate) based on risk of bias, inconsistency, indirectness, imprecision and publication bias. The certainty of evidence was classified into one of four categories: very low, low, moderate or high. All disagreements were discussed among the two authors (M-H.C. and I-W.C.) and resolved through consulting the corresponding author (K-C.H.) if required.
Strategies for data analysis
For dose-escalation studies, we only analysed the data with the highest dosage. For studies with multiple intervention arms, event number and total number in the control group were divided as previously reported [22] to avoid sample size overlapping for dichotomous data. On the other hand, the total number of participants in the placebo group was divided without changing the means and standard deviations for continuous outcomes.
By adopting a random effects model, continuous data are reported as mean difference (MD) or standardised mean difference (SMD) with 95% confidence interval and dichotomous data are presented as risk ratio with 95% confidence interval. With the conversion equations previously described [23], the reported arithmetic means were converted into geometric means. The natural log-transformed scale was used for the analysis of cough frequencies initially presented as coughs per hour. The natural log-transformed results of MD in cough frequencies were exponentially back-transformed to provide the point estimates and 95% confidence intervals for relative reduction of geometric means that facilitates interpretation. Subgroup analysis was prespecified and performed based on the dosage of gefapixant (i.e. low dose, ≤20 mg; moderate dose, 45–50 mg; high dose, ≥100 mg) to investigate the impact of dosage on the efficacy and safety of gefapixant. For the current study, the selection of cut-off ranges was based on two reported phase 2 trials [16, 19] in which the efficacy of gefapixant at different dosages was investigated. The current study set the reference dose at ∼50 mg twice daily because it was the reported potentially effective dose [16, 19]. Accordingly, the cut-off ranges of the other two subgroups were taken roughly at two-fold intervals (i.e. doses ≥twice or ≤half of the reference value). In this study, heterogeneity was evaluated by I2-statistics along with inspection of the forest plots; significant heterogeneity was considered when I2>50% or when the point estimates in the forest plots were significantly different with little overlapping of the confidence intervals. The potential publication bias for a specific outcome mentioned in 10 or more trials was evaluated through visual inspection of a funnel plot. Using “leave-one-out” sensitivity analysis, we examined how the findings from each trial could affect the overall results if I2>50% as previously reported [24, 25]. The statistical significance was set at a p-value <0.05 in all comparisons and two-tailed tests were conducted for all comparisons. With respect to continuous as well as dichotomous data, the Cochrane Collaboration's Review Manager software (RevMan version 5.3) was used for the analysis.
Results
Selection and characteristics of studies
A total of 215 records were identified through electronic and manual searching. After removing duplicates (n=104) and reports deemed unsuitable after screening their titles and abstracts (n=84), 27 articles were considered eligible. After full-text reading, 22 studies were further excluded due to various reasons (gefapixant not used, n=1; patients with acute cough, n=1 [26]; conference abstract, n=18; no placebo group, n=1 [27]; patients with idiopathic pulmonary fibrosis, n=1 [17]). Finally, five studies focusing on seven trials published between 2015 and 2022 were included in the current meta-analysis [15, 16, 18, 19, 28]. A flowchart regarding the selection process is presented in figure 1.
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram of study selection for the current meta-analysis. CENTRAL: Cochrane Central Register of Controlled Trials; IPF: idiopathic pulmonary fibrosis.
Among the five studies included, one included two randomised, double-blind, placebo-controlled, crossover, dose-escalation trials [19] and another study included two double-blind, randomised, parallel-group, placebo-controlled, phase 3 trials (i.e. COUGH-1 and COUGH-2) [18]. For the current meta-analysis, these trials are reported separately (i.e. Smith et al., 2020a and Smith et al., 2020b [19]; McGarvey et al., 2022a (COUGH-1) and McGarvey et al., 2022b (COUGH-2) [18]).
The characteristics of the seven trials are presented in table 1. Three studies involving four phase 2 trials adopted a crossover design [15, 19, 28], one was a parallel-group phase 2 trial [16] and the other study with two phase 3 trials utilised a parallel-group design [18]. The dosage of gefapixant used varied widely among the included studies, ranging from 7.5 to 600 mg twice daily. The mean or median age of participants was 55–63 years. All studies involving seven trials included a higher proportion of female participants (74.1–88%) [15, 16, 18, 19, 28]. The mean duration of cough ranged from 10.7 to 17.1 years (three studies) [16, 18, 28], while the median duration was between 9 and 15.4 years (two studies) [15, 19]. All studies investigated the efficacy of gefapixant in patients with chronic cough [15, 16, 18, 19, 28]. The treatment duration and follow-up duration are shown in table 1.
Characteristics of the seven trials from five studies
Quality assessment of the included studies
We found a low risk of bias in the selection, performance, detection and reporting bias domains in all trials. The attrition bias was considered to be unclear in two trials in which the percentage of missing data ranged from 6.8% to 9.5%, while it was deemed high in one trial in which the percentage of missing data was up to 10.7% (figure 2). Sponsoring by pharmaceutical companies in all trials contributed to their being deemed at unclear risk of bias in the “Other bias” domain. The risk of bias summary is depicted in figure 2.
A summary of the risk of bias assessment for the included randomised controlled trials. #: trial from the same study; ¶: trial from the same study.
Outcomes
Primary outcome: 24-h cough frequency
For the current study, only the data derived from a higher dosage were used in one study that included two dose-escalation trials [19]. Merged results demonstrated a lower 24-h cough frequency in patients with gefapixant treatment compared with those without (MD= −0.38, 95% CI −0.58– −0.18; p=0.0002; I2=77%; estimated relative reduction 31.6%; 2335 participants; sensitivity analysis: consistent) (figure 3). Visual inspection of the forest plot showed significant heterogeneity in effect sizes which might be attributed to different dosages. Subgroup analysis based on dosage (i.e. ≤20, 45–50 and ≥100 mg) revealed that the efficacy of gefapixant was only significant in those receiving a dosage of 45–50 mg (estimated relative reduction 30.9%; p=0.007) or ≥100 mg (estimated relative reduction 58.5%; p<0.00001) compared with those in the placebo group, suggesting a dose-dependent effect (figure 3). The funnel plot indicated a potential risk of publication bias (supplementary figure S1).
Forest plot showing difference in 24-h cough frequency between gefapixant and placebo groups. #: trial from the same study; ¶: trial from the same study. df: degrees of freedom; IV: inverse variance.
Meta-analysis of data on awake cough frequency (MD= −0.6, 95% CI −0.89– −0.31; p<0.0001; I2=56%; estimated relative reduction 45.1%; 382 participants; sensitivity analysis: consistent) showed a finding similar to that for 24-h cough frequency (supplementary figure S2). Inspection of the forest plot suggested substantial heterogeneity in effect sizes that might be explained by the dosage. However, for night-time cough frequency (MD= −0.29, 95% CI −0.7–0.12; p=0.17; I2=45%; estimated relative reduction 25.2%; 378 participants; sensitivity analysis: consistent), only high-dose gefapixant was associated with a favourable outcome (estimated relative reduction 54.2%; p=0.002) (supplementary figure S3). Inspection of the forest plot showed that heterogeneity was not substantial. The estimated relative reduction in cough frequency for all the subgroups based on dosage is summarised in supplementary table S2. These findings suggested that the efficacy of gefapixant against chronic cough was dose dependent and a dosage ≥45 mg may be required for reducing cough frequency.
Secondary outcomes: cough severity and CSD
Our results revealed a lower cough severity assessed based on the VAS in the gefapixant group compared with that in the placebo group (MD= −8.52, 95% CI −12.48– −4.57; p<0.0001; I2=66%; 2427 participants; sensitivity analysis: consistent) (figure 4a). Examination of the forest plot showed substantial heterogeneity that might be attributable to the dosage used. Subgroup analysis demonstrated that only moderate-dose (MD= −6.45, 95% CI −9.24– −3.66; p<0.00001; I2=1%) and high-dose (MD= −24.5, 95% CI −34.58– −14.42; p<0.00001; I2=27%) gefapixant were effective for reducing cough severity (all p<0.05) (figure 4a). The funnel plot demonstrated a potential risk of bias (supplementary figure S4). Analysis of the CSD showed similar findings (supplementary figure S5) with low risk of publication bias detected (supplementary figure S6). These findings supported a dose-dependent reduction in cough severity with the use of gefapixant.
Forest plots showing a comparison for a) cough severity and b) cough-related quality of life between the gefapixant and placebo groups. #: trial from the same study; ¶: trial from the same study. df: degrees of freedom; IV: inverse variance; SMD: standardised mean difference.
Secondary outcomes: cough-related quality of life
Cough-related quality of life was investigated with two scales including the CQLQ score and the LCQ score in the included studies of the current meta-analysis. Despite the favourable overall effect estimates (SMD= −0.35, 95% CI −0.53– −0.18; p<0.0001; I2=65%; 2252 participants; sensitivity analysis: consistent), subgroup analysis showed a significant improvement in quality of life with the use of moderate- or high-dose gefapixant (all p<0.05), while no significant therapeutic effect was noted with low-dose gefapixant (p=0.06) (figure 4b). Visual inspection of the forest plot showed substantial heterogeneity in effect sizes that might be linked to the dosage. Potential risk of publication bias was detected in this outcome (supplementary figure S7).
Secondary outcomes: safety issues
The forest plot indicated a higher risk of all-cause AEs in patients receiving gefapixant compared with those administered placebo (risk ratio 1.30, 95% CI 1.15–1.49; p<0.0001; I2=81%; 2505 patients; sensitivity analysis: consistent) (supplementary figure S8). A potential risk of publication bias was detected for this outcome (supplementary figure S9). There was a similar finding for treatment-related AEs (risk ratio 2.31, 95% CI 1.54–3.46; p<0.0001; I2=84%; 2411 patients; sensitivity analysis: consistent) (supplementary figure S10) with a low risk of publication bias (supplementary figure S11). For all-cause and treatment-related AEs, subgroup analysis showed a dose-dependent association between the use of gefapixant and AEs. An examination of the forest plots also demonstrated substantial heterogeneity that might be attributed to the differences in dosage (supplementary figures S8 and S10). Nevertheless, further analysis revealed no significant difference in study discontinuation rate between gefapixant and placebo groups in pooled analysis (risk ratio 2.19, 95% CI 0.77–6.21; p=0.14; I2=0%; 461 patients) or subgroup analysis (all p>0.05) (supplementary figure S12). Similarly, the serious AE rate was also comparable in both groups in pooled analysis (risk ratio 1.01, 95% CI 0.70–1.46; p=0.95; I2=0%; 2459 patients) or subgroup analysis (all p>0.05) (supplementary figure S13).
The risk of ageusia (supplementary figure S14), dysgeusia (supplementary figure S15) and hypogeusia (supplementary figure S16) was higher in the gefapixant group compared with the placebo group. For ageusia and dysgeusia, subgroup analysis showed a dose-dependent association of gefapixant with these AEs. In contrast, the use of gefapixant did not increase the risk of oral hypoesthesia (supplementary figure S17), oral paraesthesia (supplementary figure S18) and renal/urological AEs (supplementary figure S19).
Certainty of evidence
Using GRADE, the overall certainty for most of the included outcomes was considered moderate (i.e. 24-h cough frequency, awake cough frequency, cough severity, cough-related quality of life, all-cause AEs, treatment-related AEs and discontinuation due to AEs). The certainty of evidence was deemed low for two secondary outcomes, i.e. night-time cough frequency and serious AEs, while it was regarded as high for the CSD score. The certainty of evidence on the efficacy and safety of gefapixant for chronic cough is summarised in supplementary table S3.
Discussion
This systematic review and meta-analysis of five studies with seven RCTs showed the dose-dependent efficacy of gefapixant for reducing objective 24-h and awake cough frequency, but without a positive impact on night-time (asleep) cough frequency. Consistently, the use of gefapixant alleviated cough severity and improved cough-related quality of life in a dose-dependent manner. There was also a dose-dependent increase in the risk of all-cause AEs, treatment-related AEs, ageusia and dysgeusia, while the elevated risk of hypogeusia was not dose dependent. No significant differences were observed in oral hypoesthesia or paraesthesia, renal/urological AEs, discontinuation of treatment due to AEs and serious AEs.
Chronic cough, defined as cough that lasts for >8 weeks, is commonly encountered in both primary and specialty care [29]. Neuronal injury from inflammation, infection or allergy that results in altered cough regulation and cough hypersensitivity has been proposed as a potential mechanism of chronic cough [30, 31]. “Cough hypersensitivity syndrome”, which is characterised by troublesome coughing often triggered by otherwise innocuous stimuli with or without specific phenotypes (e.g. eosinophilic coughs, reflux coughs or nasal symptoms), may coexist with other pulmonary or extrapulmonary diseases [32, 33]. The use of neuromodulators targeting the cough pathway (e.g. opiates, pregabalin, gabapentin and tricyclics) in current clinical practice was based on low- to moderate-quality evidence [2].
Many studies have explored the therapeutic use of neuromodulators including opioids, gabapentinoids and tricyclics to target the cough pathway of chronic cough. Despite the observed efficacy, adverse side-effects including dizziness and sedative effects were common, and the use of these medications is still considered off-label under current regulations [34–38]. Clinical research on novel therapeutic agents targeting neuroreceptors involved in the cough pathway including P2X3, transient receptor potential (TRP) vanilloid-1 (TRPV1), TRP ankyrin-1 (TRPA1), TRP vanilloid-4 (TRPV4) and neurokinin-1 (NK-1) receptors has made significant progress, particularly with P2X3 receptor antagonists (e.g. gefapixant) [39] whose efficacy and safety have been evaluated in one study with two phase 3 trials [18].
Our findings revealed lower 24-h and daytime (awake) cough frequencies as measured objectively with ambulatory audio recording devices in the gefapixant group than in the placebo group. Subgroup analysis showed that gefapixant at doses ≥45 mg twice daily was effective, while gefapixant at doses ≤20 mg twice daily did not significantly reduce 24-h and awake cough frequencies compared with placebos. A larger effect size was observed in the high-dose (i.e. ≥100 mg twice daily) subgroup without statistical heterogeneity. Nevertheless, the number of patients in each of the included trials was relatively small (i.e. <60 participants); therefore, whether the observed difference in effect size reflected a dose-dependent effect or was due to research bias from small study effects remains to be elucidated. On the other hand, our finding of a lack of positive impact of gefapixant on night-time (asleep) cough frequency may be attributed to the markedly reduced cough frequency during sleep when the low frequency and high variability of cough at baseline may preclude any sizeable benefit from treatment in the chronic cough population.
A previous study has reported that cough frequency per se may not be fully representative of cough severity since other factors (e.g. the intensity and effort of coughs) are also important contributors [40]. A combined evaluation with cough-related quality of life indicators was therefore recommended [40]. In our meta-analysis, the beneficial effect of gefapixant against chronic cough at doses ≥45 mg (but not ≤20 mg) twice daily was supported by subjective patient-reported outcomes that included questionnaires (CQLQ and LCQ scores), a CSD and a 100-mm VAS.
Focusing on AEs, our study demonstrated that treatment with gefapixant was associated with a dose-dependent increased risk of all-cause and treatment-related AEs at doses ≥45 mg twice daily compared with placebos. However, there was no significant difference in serious AEs or treatment discontinuation due to AEs in all subgroups. Incidences of taste-related AEs including ageusia, dysgeusia and hypogeusia were also higher in the gefapixant group, while no significant differences were observed in oral hypoesthesia or paraesthesia and renal/urological AEs. The mechanism underlying our finding of taste-related AEs as the most common AEs of gefapixant has been proposed to be caused by co-inhibition of the P2X2/3 receptor based on animal studies [41, 42]. Such taste-related AEs were reported to be mild to moderate in intensity and were reversible, with resolution observed in most patients during or soon after discontinuation of gefapixant treatment [18]. The safety profile demonstrated in the present study was in line with that reported in another RCT evaluating the AEs of gefapixant as its primary outcomes [27]. Some recently developed P2X3 receptor antagonists such as BLU-5937, sivopixant (S-600918), eliapixant (BAY1817080) and filapixant (BAY1902607) are currently under investigation [43–46]. Although a higher P2X3 selectivity over P2X2/3 receptors among these newer agents may confer a reduction in taste-related adverse effects compared with gefapixant, this theoretical benefit needs to be confirmed in future large-scale and comparative trials.
There were some limitations in the present investigation that should be considered. First, because the current study mainly recruited patients with refractory or unexplained chronic cough, the efficacy of gefapixant against chronic cough from specific diseases (e.g. obstructive or interstitial lung diseases) remains to be addressed. Second, because the follow-up period for all the included RCTs was <1 year, we were unable to evaluate the long-term efficacy and potential adverse effects of gefapixant. This is especially relevant since chronic cough can persist for several years [8, 47], with treatment fitting into the same timeframe. Further research is warranted to address this issue. Third, numbers of included trials and sample sizes for some of the outcomes, especially in the high-dose subgroup, were relatively small and caution is needed when interpreting the results. Finally, it is noteworthy that all of our included studies were funded by two companies, i.e. Afferent Pharmaceuticals and Merck Sharp & Dohme, in which the latter amalgamated the former in 2016, thereby raising the concern of a conflict of interest for which all the included trials were judged to be at unclear risk of bias in the “Other bias” domain. In addition, the presence of potential publication bias in several outcomes warrants a cautious interpretation of the findings.
In conclusion, this systematic review and meta-analysis showed that gefapixant at doses ≥45 mg twice daily was effective for reducing 24-h and daytime (awake) cough frequencies as well as subjective cough severity, but not night-time cough frequency. Patients treated with gefapixant had an increased risk of all-cause, treatment-related and taste-related AEs compared with placebos, but no significant difference was observed in oral hypoesthesia or paraesthesia, renal/urological AEs, discontinuation of treatment due to AEs and serious AEs. Further results from large-scale studies are warranted to support our findings.
Points for clinical practice
Gefapixant improved 24-h/awake cough frequency, subjective cough severity and quality of life. Nevertheless, treatment with gefapixant was associated with an increased risk of all-cause, treatment-related and taste-related AEs. Both therapeutic and adverse effects were dose dependent with a cut-off dose of 45–50 mg twice daily.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material ERR-0219-2022.SUPPLEMENT
Footnotes
Provenance: Submitted article, peer reviewed.
Availability of data and materials: All data related to the present systematic review and meta-analysis are available from the corresponding author on reasonable request.
Author contributions: Conceptualisation: M-H. Chuang and I-W. Chen. Methodology: J-Y. Chen. Software: F-C. Kang. Validation: K-C. Hung and C-N. Ho. Formal analysis: K-C. Hung. Investigation: S-C. Wu and M. Yew. Resources: K-M. Lan. Data curation: I-W. Chen and K-C. Hung. Writing (original draft preparation): M-H. Chuang and K-C. Hung. Writing (review and editing): K-C. Hung and I-W. Chen. Visualisation: K-C. Hung. Supervision: K-C. Hung. All authors have read and agreed to the published version of the manuscript
Conflict of interest: The authors declare no conflicts of interest.
- Received November 8, 2022.
- Accepted February 27, 2023.
- Copyright ©The authors 2023
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org
References
